- MACHINE LEARNING -
General Index by Topic to AI in the news
AI Topics Home  
 

September 30, 2004: New Company Starts Up a Challenge to Google. By John Markoff. The New York Times (no fee reg. req'd.). "Google executives have long conceded that one of their great fears is to be overtaken by a more advanced Internet search technology. Vivisimo, a company founded by three former Carnegie Mellon University computer scientists, is hoping to prove that Google's worries are well founded. Four-year-old Vivisimo plans to start Clusty, a free, consumer search service based on results from Yahoo's Overture engine, Thursday. ... The service is meant to address the confusion that can be created when search engines return huge lists. Clustering is also intended to help users find related material they may overlook when they employ services that utilize page ranking methods. Such methods employ a variety of software algorithms to rank Web pages by their perceived relevance to a query. ... Vivisimo's co-founder and chief executive, Raul Valdes-Perez, was a protégé of Herbert A. Simon, a Nobel laureate who was a pioneer in artificial intelligence research. Before co-founding Vivisimo, Mr. Valdes-Perez was a computer scientist at Carnegie Mellon University. He professes that the way to deal with information overload is with information 'overlook' - techniques that strip away extraneous information."
>>> Information Retrieval, Machine Learning, Applications

September 27, 2004: The Grand Challenges of IT - Researchers are inventing new ways to tackle old problems. Emerging Technology by Thomas Hoffman. Computerworld. "Fundamental research on how to make computer hardware more powerful and software smarter goes back 50 years or more, but many of the traditional methods have nearly reached their limits. Now, researchers moving in bold new directions may be setting the course of IT for decades to come. There are literally dozens of grand challenges that scientists and economists are attacking, ranging from societal issues to technical advances. Here, we take a look at the challenges in three key areas of IT research: processor performance, chip miniaturization and artificial intelligence. ... AI, very broadly defined, comprises three primary disciplines: natural-language processing, machine-based learning and robotics. Recent advances in these areas have led to commercial technologies ranging from a robotic vacuum cleaner called Roomba, made by Burlington, Mass.-based iRobot Corp., to customer-service-oriented speech recognition systems from vendors such as Peabody, Mass.-based ScanSoft Inc. But despite these inroads, computer systems continue to have a tough time handling reasoning. 'The biggest challenges are figuring out how to organize computer programs to have more common sense,' says Tom Mitchell, the Fredkin professor of computer science at Carnegie Mellon University in Pittsburgh. ... The Defense Advanced Research Projects Agency is funding research to develop a computer-based 'executive assistant' that could handle administrative tasks like prioritizing e-mail requests for a military commander or a business executive. ... Using a grading scale of A to F, 'we would be thrilled if these systems could give us C-level performance over the next three to four years,' says Ron Brachman, director of the information processing technology office at DARPA. Computers also have trouble understanding context like humans do.... Systems that can handle more complicated human-to-computer interactions, like processing a request for movie tickets at a particular theatre via speech recognition, should be in use within five to 10 years, says Victor Zue, co-director of the MIT computer science and artificial intelligence laboratory."
>>> AI Overview, Applications, Commonsense, Agents, Machine Learning, Natural Language Processing, Speech, Robots, Reasoning

September 26, 2004: Crick's other goal - Unlocking riddle of the mind. Scientists continuing study of consciousness. By Bruce Lieberman. San Diego Union-Tribune & SignOnSanDiego.com. "Francis Crick focused on looking for an area of the brain that might be critical to human consciousness. As a young scientist in 1940s England, Francis Crick decided to devote his life to unraveling two mysteries: the foundation for all living things and how the brain gives rise to the mind. ... Tomorrow, when the Salk Institute in La Jolla hosts a public memorial for Crick, who died July 28 at 88, that unfinished business will most certainly be talked about. How billions of brain cells interpret sensations, draw on memory and association to make sense of them, and create conscious thoughts about the world is unknown. 'It's inconceivable to us, but somehow it happens,' said Terry Sejnowski, a computational neurobiologist at the Salk Institute who studies how computers can be used to understand the brain. 'Consciousness is elusive,' he said. 'It's hard to pin down.' ... Illuminating how the brain creates consciousness would profoundly change the way humans view themselves, scientists say. ... Engineers could build machines that truly think, bringing artificial intelligence out of science fiction and into the real world. ... [C]onsciousness is really about how all the parts come together to create the thinking mind. 'Being reductionist is a good way to start, but at some point you have to . . . put together the pieces and see how they work together,' Sejnowski said. He calls the effort to assemble the big picture of consciousness 'the Humpty Dumpty project.'"
>>> Cognitive Science, Philosophy, Neural Networks & Connectionist Systems, Events (@ Resources for Students), Machine Learning

September 26, 2004: Consumers must be mindful of credit theft - ID thieves ingenious, but card networks are making it tougher to defraud victims. By Paul Gores. Milwaukee Journal Sentinel / available from IndyStar.com. "[A]lthough illegal card purchases are the most common type of identity-theft crime, they are starting to decline because of increasingly sophisticated computer programs that spot credit-card fraud, she said. So-called 'neural' computer networks know who holds a credit card, the bank that issued it, how much the cardholder normally spends, the categories of things typically purchased, and the locations where the card usually is used. The computer is constantly learning and updating patterns. It also knows where rashes of credit-card fraud are taking place. 'It keeps getting smarter and smarter all the time. It's almost like artificial intelligence,' [Avivah Litan, vice president and research director at Gartner Inc.] said. The network looks for abnormal patterns, huge purchases, rapid purchases, items not normally bought by the cardholder and purchases too far apart geographically -- and too close chronologically -- for one person to have made them."
>>> Fraud Detection & Prevention, Neural Networks, Banking, Machine Learning, Applications; also see our But is it AI collection

September 21, 2004: Chicago Moving to 'Smart' Surveillance Cameras. By Stephen Kinzer. The New York Times (no fee reg. req'd.). "A highly advanced system of video surveillance that Chicago officials plan to install by 2006 will make people here some of the most closely observed in the world. Mayor Richard M. Daley says it will also make them much safer. ... Police specialists here can already monitor live footage from about 2,000 surveillance cameras around the city, so the addition of 250 cameras under the mayor's new plan is not a great jump. The way these cameras will be used, however, is an extraordinary technological leap. Sophisticated new computer programs will immediately alert the police whenever anyone viewed by any of the cameras placed at buildings and other structures considered terrorist targets wanders aimlessly in circles, lingers outside a public building, pulls a car onto the shoulder of a highway, or leaves a package and walks away from it. Images of those people will be highlighted in color at the city's central monitoring station, allowing dispatchers to send police officers to the scene immediately. ... Many cities have installed large numbers of surveillance cameras along streets and near important buildings, but as the number of these cameras has grown, it has become impossible to monitor all of them. The software that will be central to Chicago's surveillance system is designed to direct specialists to screens that show anything unusual happening. ... 'With the aggressive way these types of surveillance equipment are being marketed and implemented,' Mr. [Edwin C.] Yohnka said, 'it really does raise questions about what kind of society do we ultimately want, and how intrusive we want law enforcement officials to be in all of our lives.' ... 'The value we gain in public safety far outweighs any perception by the community that this is Big Brother who's watching,' Mr. [Ron] Huberman said. 'The feedback we're getting is that people welcome this. It makes them feel safer.'"
>>> Law Enforcement, Vision, Machine Learning, Ethical & Social Implications, Applications; also see the Fall 2002 and Spring 2003 AI in the news columns in AI Magazine

September 16, 2004: Duo-Mining -Combining Data and Text Mining. By Guy Creese. DMReview.com. "As standalone capabilities, the pattern-finding technologies of data mining and text mining have been around for years. However, it is only recently that enterprises have started to use the two in tandem - and have discovered that it is a combination that is worth more than the sum of its parts. First of all, what are data mining and text mining? They are similar in that they both 'mine' large amounts of data, looking for meaningful patterns. However, what they analyze is quite different. ... Collections and recovery departments in banks and credit card companies have used duo-mining to good effect. Using data mining to look at repayment trends, these enterprises have a good idea on who is going to default on a loan, for example. When logs from the collection agents are added to the mix, the understanding gets even better. For example, text mining can understand the difference in intent between, 'I will pay,' 'I won't pay,' 'I paid' and generate a propensity to pay score - which, in turn, can be data mined. To take another example, if a customer says, 'I can't pay because a tree fell on my house;' all of a sudden it is clear that it's not a 'bad' delinquency - but rather a sales opportunity for a home loan."
>>> Data Mining & Discovery, Machine Learning, Natural Language Processing, Banking, Applications

September 15, 2004: Artificial Intelligence lab works to hunt terrorists, cure cancer. By Joe Ferguson. Arizona Daily Wildcat. "The director of the UA's Artificial Intelligence Lab, Hsinchun Chen, said the goal of the AI Lab is to provide academics and professionals with a better way to get information in their high-tech worlds. ... COPLINK takes data from various law enforcement databases, based on existing criminal records, and allows law enforcement to coordinate their information using the software. 'It is like Google for cops,' said Chen. "'But it is much better.' ... Catherine Larson, associate director of the AI Lab, said the lab is also active in the war on terror by helping the Department of Homeland Security identify terrorists. By working closely with available data, the AI Lab identifies patterns with the data to discover the true identities of criminals using aliases, Larson said. ... Chen said the AI Lab also works closely with the University Medical Center's Cancer Center in informatics, a field that analyzes exiting medical data to find patterns that could help researchers at the Cancer Center find new ways to treat cancer."
>>> Law Enforcement, Medicine, Knowledge Management, Data Mining and Knowledge Discovery, Machine Learning, AI Academic Departments (@ Resources for Students)

September 13, 2004: Poly, varsities use software to spot copying from Net - Ngee Ann Poly tries out anti-plagiarism software that dons at 3 varsities are using to spot text students lift from websites. By Lynn Lee. The Straits Times Interactive. " Blithely, the group of undergraduates lifted a chunk of text from the website of a Singapore bank's branch in Thailand and passed it off as their own. ... But their misdeed was exposed swiftly by Turnitin, a software which matches student work against millions of documents on the Internet. ... In the future, artificial intelligence may be used to distinguish one student's essay from others. German academic Joachim Diederich, 46, in Singapore last week to talk about his research on new technologies to fight plagiarism, aims to produce such a software. Professor Diederich, who disclosed his research at the International Conference On Educational Technology, has been conducting his research on the topic since 1999."
>>> Natural Language Processing, Machine Learning, Fraud Detection & Prevention, Applications

September 13, 2004: Pentagon Revives Memory Project. By Noah Shachtman. Wired News. "It's been seven months since the Pentagon pulled the plug on LifeLog, its controversial project to archive almost everything about a person. But now, the Defense Department seems ready to revive large portions of the program under a new name. Using a series of sensors embedded in a GI's gear, the Advanced Soldier Sensor Information System and Technology , or ASSIST, project aims to collect what a soldier sees, says and does in a combat zone -- and then to weave those events into digital memories, so commanders can have a better sense of how the fight unfolded. ... To crunch all the information it receives, ASSIST will have to be smart and able to learn from the experiences its wearers feed it. Building these types of thinking machines has been the goal of Ronald Brachman since he took over Darpa's Information Processing Technology Office in 2002. 'It is the progressive improvement of the knowledge base of the system over time that we believe will best support soldiers on later missions,' Brachman wrote in an e-mail. It will 'allow them to understand what prior patrols saw and heard, and to recognize salient (and potentially life-threatening) changes in the situation when they go out on a mission.'"
>>> Agents, Knowledge Management, Machine Learning, Military, Applications, Ethical & Social Implications; also see related articles in our News Archive such as Helping Machines Think Different (July 2003)

September 11, 2004: CPW studies ways to foil attack on water supplies. By Robert Behre. The Post and Courier & Charleston.net. (no fee reg. req'd.) "John Cook knows the next terrorist strike might not come by air. As assistant manager with Charleston's Commissioners of Public Works, Cook is privy to terrorist talk about undermining the nation's water supply. ... Working with Ed Roehl of the Greenville firm Advanced Data Mining, Cook has drawn up a $400,000 research project that aims to couple everyday sensor technology with advanced number-crunching models to create a reliable, affordable way to monitor water distribution lines. ... The hard part is making sense out of millions of measurements each day. 'No normal person can take all the data in from all these points and analyze it,' Cook says. 'That's where the artificial intelligence is important to make this work.' ... He and Roehl are developing software that can 'learn,' as humans provide it feedback about what caused an abnormal readout, such as a heavy rain."
>>> Public Health & Welfare, Law Enforcement, Machine Learning, Applications

September 9, 2004: Mimicking fraudsters - If your card use has been queried, it's probably because more banks are now using artificial intelligence software to try to detect fraud. By Ken Young. The Guardian. "Credit card fraud losses in the UK fell for the first time in nearly a decade last year, by more than 5% to £402.4m, according to research by the Association of Payment Clearing Services (Apacs). The fall has put a spotlight on the increasing use of neural networks that have the ability to detect fraudulent behaviour by analysing transactions and alerting staff to suspicious activity. As commercial applications of research into artificial intelligence, these systems give the impression of mimicking human abilities for recognising unusual activity. Karina Purang, a financial analyst at Datamonitor in London, says the use of neural networks is growing: 'These systems are very important to banks trying to reduce fraud, and are becoming standard across the card industry to detect unusual spending patterns.' She says Barclays reported that after installing Fair Isaac's Falcon Fraud Manager system in 1997, fraud was reduced by 30% by 2003. The bank attributed this mainly to the new system. ... Nick Sandall, head of retail banking at Deloitte, says that banks also use other technologies. 'The artificial intelligence community is constantly bringing us new solutions. ...'"
>>> Banking, Fraud Detection & Prevention, Neural Networks, Applications, Machine Learning, Industry Statistics

September 9, 2004: MyVista ready for market. By Charles F. Moreira. The Star Online. "Intelligence Systems Sdn Bhd director See Wan Chee said his company acquired 20 customers for its SmartScan imaging application since it won the regional APICTA 2003 award in the Best in Education category in Bangkok last December.  'Most of them, including Nottingham University's Malaysian campus, use SmartScan to read and mark answers to multiple-choice exam questions, as well as for data collection,' See told In.Tech at ACM2004.  It also indicates students' strengths and weaknesses so that teachers can take remedial action to help students improve in those areas where they are weak.  SmartScan (www.smartscan.com.my) uses artificial intelligence (AI) techniques, including pattern recognition, neural networks and fuzzy logic to analyse answers on an objective test answer sheet."
>>> Pattern Recognition, Neural Networks, Fuzzy Logic, Education, Applications, Machine Learning, Reasoning

September 8, 2004: Artificial Intelligence Creeps into the Commercial Market Despite Initial Hurdles. PhysOrg.com. "When artificial intelligence (AI) was developed to emulate human intelligence, scientists hoped it would be a blockbuster technology. Instead, the inability of end users to deal with its complexity and expensiveness and their lack of understanding of its potential caused these expectations to dwindle. These factors slowed down the adoption rates of AI, but not the efforts of researchers. After a couple of decades, AI, now in the form of applications, is slowly making its way out of laboratories into the mainstream market."
>>> Medicine, Case-Based Reasoning, Scientific Discovery, Bioinformatics, Data Mining & Knowledge Discovery, Customer Relations, Expert Systems, Applications, Reasoning, Machine Learning

September 7, 2004: Security IT tops NSW tech showcase awards. By Fleur Doidge. CRN / available from iTNews Australia. "Two security-focused IT developers creamed the competition this year at the patrons' awards for the NSW Government's export-focused Australian Technology Showcase (ATS). ... Michael Egan, NSW Treasurer and Minister for State Development, said Argus had won for its success in growing export deals. ... Argus' patented iris recognition system had netted $600,000 in export sales in two years -- a considerable achievement for a new, innovative technology, he said. ... Second place went to another IT surveillance system developer, iOmniscient, based in Sydney's Chatswood. ... iOmniscient had patented a surveillance system using artificial intelligence to detect unmoving, suspicious objects -- such as bags and boxes that could contain bombs -- in crowded areas such as airports and train stations, [Warren Dick] said."
>>> Law Enforcement, Biometrics (@ Image Understanding), Pattern Recognition, Vision, Machine Learning, Applications

September 4, 2004: Robots invade the table football pitch. By Duncan Graham-Rowe. New Scientist Magazine (appears on page 18 with the title: Play table football against a robot). "Fans of table football, or foosball, will no longer have to hang around at the pub waiting for a friend to turn up before they can play. A robotic foosball table will be able to give them just as good a game. ... To allow the control system to track the ball, the base of the table is made of translucent glass, tinted green. A camera underneath photographs the ball 50 times per second, and sends this data to a built-in computer that maps the ball's position. Intelligent software then works out the effect of one of the figures kicking the ball. ... [Bernhard Nebel's University of Freiburg] team is now working on being able to stop the ball and pass it -- a capability that will be essential if the robot is ever going to beat good players."
>>> Sports, Entertainment, Vision, Planning, Machine Learning, Robots, Reasoning, Applications

September 4, 2004: Brain research? Pay it no mind. Mystery of consciousness still outwitting scientists. By Philip Marchand. The Toronto Star. "Scientists who have been trying to understand the brain have recently tried to measure neural activity of Republicans and Democrats to see if political affiliations had anything to do with brain chemistry. The results were inconclusive. ... What really caught my eye about a New York Times Magazine article on the topic was the following statement: 'One of the most celebrated insights of the past 20 years of neuroscience is the discovery -- largely associated with the work of Antonio Damasio -- that the brain's emotional systems are critical to logical decision-making. People who suffer from damaged or impaired emotional systems can score well on logic tests but often display markedly irrational behaviour in everyday life.' I'm sure Damasio has done good work, rooting around the neocortex. But what does it say for neuroscience that one of its 'most celebrated insights' is something we've known for three or four millennia? ... The bravest of the neuroscientists are trying to tackle the toughest nut of all, the mystery of consciousness. ... A professor named Howard Gardner, for example, whose 1985 book The Mind's New Science helped to popularize the field of cognitive science, told Horgan that questions such as consciousness and free will were 'particularly resistant' to the scientific habit of trying to break down a subject into its most elemental parts, like neural pathways in the brain. ... The human brain is so complex it simply defies the same kind of analysis that scientists devote to subatomic particles or human immune systems. 'Like neuroscientists, researchers in evolutionary psychology and artificial intelligence are both bumping up against the Humpty Dumpty dilemma,' [John] Horgan writes. 'They can break the mind into pieces, but they have no idea how to put it back together again.'"
>>> Emotion, Creativity, Cognitive Science, Reasoning, Philosophy, Neural Networks, Machine Learning

September 3, 2004: Software firm finds gold in diverse data mining. By Mary Ann Azevedo. Houston Business Journal / also available from MSNBC (9/5/04). "Customers of PolyVista range from technology giant Hewlett-Packard Co. and global energy conglomerate British Petroleum to telecom heavyweight Verizon and the University of Texas M.D. Anderson Cancer Center. The company founded by Shabhaz Anwar in 1998 helps users discover anomalies, identify trends and pinpoint relationships within huge databases over a short period of time. The data is mined by letting unstructured text interact with structured mathematical information, according to Anwar, who uses decks of cards to describe the software system. 'Imagine having six decks of cards all mixed up in a big pile,' Anwar explains. 'It's a tedious and time-consuming process to separate them manually into six separate packs. PolyVista takes the pile and automatically stacks the cards in order, in six separate stacks -- all in one step.' The process yields nuggets of information that humans might miss in the overwhelming abundance of data, he says, and the software can be customized to meet the needs of diverse users. ... The company's success can be attributed to the fact that it focuses on data mining, which is increasingly becoming more a part of data access for most companies, according to [William] McKnight."
>>> Data Mining and Knowledge Discovery, Business, Applications, Machine Learning

September 1, 2004: Turn Search Into Find. By Nathaniel Palmer. Transform Magazine. "Web-based customer self-service is gaining rapid adoption as one of the most promising opportunities for customer-facing firms in all industries to decrease customer transaction costs while maintaining or improving service quality. ... Information retrieval, or search, software is built upon two fundamental components: an indexing engine that maps and categorizes content, and a retrieval engine that deploys algorithms to find and return indexed content. ... A taxonomy refers to structures built to organize information -- a collection of relevant topics and subtopics arranged in a hierarchical structure. Humans use taxonomies to make sense of formerly unstructured information. ... Taxonomy and classification within customer self-service solutions is enabled by software that creates hierarchical structures and defines characteristics throughout the branches of the structures. Once these structures are in place, classification is accomplished by parsing collections of content and assigning individual documents to appropriate categories within the taxonomy structure. This can be done manually (with the aid of software) or automatically based on specific algorithms (see 'Behind the Jargon: Five Approaches to Classification')."

  • sidebar: Behind the Jargon - Five Approaches to Classification. Transform Magazine. "In the statistical analysis approach, subsets of documents are identified manually and presented to the software as 'exemplary' to a given topic or node of the taxonomy. The provided sample content is analyzed and from this the taxonomy is further refined and the rules of classification established. These rules are then used to automate the analysis of new documents and their classification into the taxonomy. This approach is also referred to as 'machine learning.' The Bayesian probability approach attempts a concept-based analysis by learning the probabilities of words being related in a given category. ... Neural networks create a matrix of computational nodes. These nodes track and compare topic similarity. A neural network utilizes artificial intelligence to build an interconnected system of processing elements, each with a limited number of inputs and outputs. Rather than being programmed, these systems learn to recognize patterns. ... Support vector machine algorithms are derived from statistical learning theory.... Semantic analysis and clustering supports both taxonomy creation and content categorization."

>>> Information Retrieval, Customer Service, Machine Learning, Probability, Bayes (@ Namesakes), Neural Networks, Natural Language Processing, Representation, Reasoning, Applications

August 30, 2004: In Search Of Better Video Search. IBM, Microsoft, and academic researchers are trying to invent ways to find specific images in video footage. By Aaron Ricadela. InformationWeek. "At a conference in Cambridge, England, last week, an IBM researcher gave the first public demonstration of a computer system called Marvel that uses statistical techniques to learn about relationships between colors, shapes, patterns, sounds, and other clues from video footage that can help identify its content. IBM's prototype then labels the footage so users can go back and find individual shots. That could be a boon not only to TV news producers but intelligence analysts watching surveillance video and even PC users editing home movies. Today's state of the art relies on searching for keywords embedded in video files, says IBM Research senior manager John Smith, who heads the project. ... Smith's team also is working with Columbia University's digital video multimedia lab on a project to search news footage from U.S. and foreign broadcasters for related topics, combining computer vision and image understanding with machine learning approaches that analyze each station's signature approach to a story."
>>> Information Retrieval, Image Understanding, Machine Learning, Vision, Law Enforcement, Applications

August 30, 2004: An apple for the computer - Machines are so sophisticated they can be used to grade essays. But in some ways, artificial intelligence still lacks common sense. By Faye Flam. Philadelphia Inquirer. "First, computers learned to beat people at chess, then they started answering 411 calls. Now, computers endowed with artificial intelligence are going where only teachers ventured before: They're grading essays. At least three companies are marketing computerized essay graders, and thousands of schools across the country are using them as teaching tools and to score standardized tests. ... Jill Burstein, [E-rater's] lead scientist and a computational linguist, said the computer is 'trained' by feeding it thousands of essays that have already been scored and then asking the system to look for patterns that distinguish the good from the bad. ... [E]ssay-scoring programs will work for students who make a good-faith effort, said Harry Barfoot, vice president for marketing and sales at Vantage Learning. 'It can't score poetry and creative writing,' he said, but that was never promised. ... [Henry] Lieberman and other artificial intelligence researchers say computers could become dramatically smarter and more humanlike in the future. The brain is just a physical machine, albeit a complicated one we don't yet understand, they argue. 'People have this illusion that what we do is magic and it will never be automated,' said University of Pennsylvania computer science professor Lyle Ungar. When he first started studying artificial intelligence, he said, no one thought a computer could play chess well enough to beat the masters. Today, computers can beat everyone at chess, he said, and we're no longer impressed."
>>> Education, Intelligent Tutoring Systems, Pattern Recognition, Commonsense, The AI Effect, Machine Learning, Applications

August 26, 2004: From factoids to facts. At last, a way of getting answers from the web. The Economist. "Ask MSR is still a prototype, although Microsoft is trying to improve it and it may be launched commercially under the name AnswerBot. Dr [Eric] Brill, meanwhile, has moved to a more difficult task. One of his most recent papers, written jointly with Radu Soricut of the University of Southern California, is entitled 'Beyond the Factoid'. It describes his efforts to build a system capable of providing 50-word answers to questions such as "What are the rules for qualifying for the Academy Awards?" This is harder than finding a single-word answer, but Dr Brill thinks it should be possible using something called a 'noisy channel' model. Such models are already employed in spell-checking and speech-recognition systems. They work by modelling the transformation between what a user means (in spell-checking, the word he intended to type) and what he does (the garbled word actually typed). ... Rather than relying on a traditional 'artificial intelligence' approach of parsing sentences and trying to work out what a question actually means, this quick-and-dirty method draws instead on the collective, ever-growing intelligence of the web itself."
>>> Information Retrieval, Web-Searching Agents, Natural Language Processing, Machine Learning, Applications

August 25, 2004: Card fraud prevention 'pays off'. BBC News. "Market analyst Datamonitor said credit card fraud fell 5% to £402.4m last year, from £424.6m in 2002. ... 'The efforts spent by the various players in preventing card fraud are finally paying off,' report author Karina Purang said. She added that the introduction of new technology - such as neural network systems which flag up transactions that do not match a cardholder's usual spending behaviour - had helped to curtail card fraud."
>>> Fraud Detection & Prevention, Banking, Neural Networks, Machine Learning, Applications

August 23, 2004: WebGen keeps rooms cool and electric bills down at UM. The Miami Herald & Herald.com. "[W]hen you're the University of Miami , it's next to impossible to monitor what's happening in each classroom, office and lab on a 260-acre campus that includes two colleges and seven schools. Enter WebGen, a Cambridge, Mass., company that has developed a software-based system that allows businesses and organizations to control their energy use and costs. ... At UM, each floor in a building is divided into zones. ... The WebGen systems check in every two minutes. ... The system also factors in the weather and how the room is being used. At UM, WebGen is tied into the school's scheduling system so it knows when classes are in session and rooms in use. ... The four partners who wrote the business plan for WebGen came from disparate backgrounds, but they brought the expertise needed to make the company work. ... Dirk Mahling provided the technology through his work with artificial intelligence and neural networks...."
>>> Neural Networks, Smart Houses, Applications, Machine Learning

August 20, 2004: Diverse Sciences Propel Bioinformatics. By Jessica D. Tenenbaum. eWeek. " At conferences in computational biology, speakers generally start with questions: 'How many people in the room are biologists? Computer scientists? Other?' It can be hard to predict what kinds of experts will show up in the audience. This year's Computational Systems Bioinformatics Conference, the third of its kind, was no exception. The CSB 2002 Web site described the conference's goal as bringing together 'biology and computer science' experts. This year, the conference organizers hope to 'promote a systems biology approach that links biology, computer science, mathematics, chemistry, physics, medicine and engineering.' That's five new disciplines in two years. Even so, we've left out statistics. ... One is struck both by how far the field has come in a relatively short period of time, and also by how far it has yet to go. In the past 10 years, the numbers of sequences stored in public databases such as GenBank, SwissProt and even the Protein Data Bank all have increased exponentially. ... The conference agenda itself highlighted how interdisciplinary this field is. ... Other presentations included methods from high-throughput microscopy, text processing, data mining, artificial intelligence and more. Fusions of fields are not just expected but required. Stephen Wong of Harvard University explained how to use robotic automation and digital microscopy to screen thousands of cells simultaneously for, among other tasks, high-throughput drug screening."
>>> Bioinformatics, Data Mining, Applications, Machine Learning, Robots

August 19, 2004: Future Route releases AI-based fraud detection product. finextra news. "UK-based Future Route is releasing a new card fraud detection system, iHex, based on artificial intelligence technology developed at Oxford University's computing laboratories for bio-informatics. The product has been designed for use by financial services firms, government agencies and corporations. IHex detects fraud using Inductive Logic Programming (ILP) techniques - an artificially intelligent method of identifying fraud patterns and anomalies. The vendor says unlike many other pattern detection products, the system automatically generates and continuously enhances underlying rules."
>>> Fraud Detection & Prevention, Banking, Pattern Recognition, Machine Learning, Expert Systems, Applications

August 18, 2004: Popular stock market invt theories. By Richard J. Maturi. Sify.com. "There's a myriad of broad based investment theories within which numerous investment strategies can be implemented. Here we will look at the rationale behind these theories and how they work. ... Jerry Felson offers an alternative to the efficient market theory in his book, Cybernetic Approach to Stock Market Analysis (Exposition Press, 1975) in order to bypass its perceived limitations and deficiencies. ... Using cybernetics concepts (the science and control of communication, and mathematical analysis of the flow of information) and artificial intelligence (advanced cybernetics) techniques, Felson proposes developing judgmental decision-making processes by weighing evidence and formalizing investment analysis. In plain language, the cybernetics approach automates the investment decision-making process through the use of pattern recognition, learning system theory, and other methods, removing the imperfect human factor and theoretically improving investment returns"
>>> Finance & Investing, Pattern Recognition, Machine Learning, Applications

August 17, 2004: Funding for UCD-based Lightwave. Radio Telefís Éireann (RTÉ), the Irish Public Service Broadcasting Organisation. "Lightwave is close to developing its first product, called ICE (Intelligent Control of Energy), which uses artificial intelligence techniques to anticipate how a building will react to new conditions such as the outside temperature or the number of people occupying the building."

  • Also see: Lightwave secures funding. Ireland On-Line (August 17, 2004). "Lightwave Technologies, an innovative environmental technology company located at NovaUCD, has secured seed capital funding of ¤300,000 from a group of private investors. Lightwave Technologies uses artificial intelligence techniques to make efficient decisions for controlling energy usage in commercial buildings with the objective of saving up to 30% of energy costs for clients. ... This system is using new advances in computer science in the area of neural networks."
>>> Neural Networks, Smart Houses, Applications, Machine Learning

August 17, 2004: The 'Nose' Knows A Sweet Smell Of Success. SpaceDaily. "What about detecting chemical leaks in enclosed spaces, like the International Space Station or Space Shuttle? NASA built 'E-Nose' to come to the rescue. The Agency's Jet Propulsion Laboratory in California and the California Institute of Technology jointly developed a method for a machine to 'smell.' ... E-Nose technology has the ability to send a signal to an environmental control system where a central computer decides how to handle the problem, without human interaction. The device also can be 'trained' in one session to detect many specific contaminants. ... Commercial companies were quick to see E-Nose's potential. In March 1997, JPL licensed the technology to Cyrano Sciences, of Pasadena, Calif. The company renamed the device 'Cyranose 320' and put it to work in the food industry, testing for spoilage. The technology is also being tested to detect toxic materials, water pollutants and chemical leaks."
>>> Artificial Noses, Hazards & Disasters, Applications, Resources for Educators, Machine Learning

August 12, 2004: When machines breed - Evolvable hardware -- gadgets that design themselves -- can get the job done, even if humans have no idea how they do it. By Sam Williams. Salon.com (subscribe or watch a brief ad to get a free day pass). "Paul Layzell is a specialist in the budding field of evolvable hardware. Simply put, he helps machines design themselves, using principles borrowed directly from biological evolution. ... Using evolutionary processes to optimize machine performance is nothing new. Since the 1960s, artificial intelligence researchers have exploited the dynamics of Darwinian evolution to solve software problems in fields as diverse as financial investment, manufacturing and biochemistry. What is new, however, is the application of evolutionary processes in the hardware realm."
>>> Genetic Algorithms, Engineering, Machine Learning, Finance & Investing, Manufacturing, Bioinformatics, Applications

August 3, 2004: Neural network mimics human thought process. HP interested in technology invented at U of I. By Julie Howard. The Idaho Statesman. "Creation of a machine that thinks like a person is a step closer to reality with a discovery from a team of researchers at the University of Idaho in Moscow. The new technology, with a patent pending, 'opens the door' for computers or robots to do rapid computations in a much more complex way than scientists had thought possible, said inventor Richard Wells. ... The technology is called a neural network, and it operates much as the human brain does, Wells said. Instead of a microprocessor, which performs computation after computation, a neural network can do several computations simultaneously. In addition, a 'neurocomputer' uses what's called 'neuro-fuzzy logic,' meaning it deals with uncertainty, the missing function of traditional programmed integrated circuitry. ... Having a neural network formed on a computer chip could revolutionize the way computers function, said Gene Merrell, acting director of the Idaho Research Foundation Inc., based at the university."
>>> Neural Networks, Fuzzy Logic, Machine Learning, Systems

August 2004: A Machine With a Mind of Its Own - Ross King wanted a research assistant who would work 24/7 without sleep or food. So he built one. By Oliver Morton. Wired Magazine (Issue 12.08). "For a machine that's changing the world, the device on the lab bench in front of me doesn't look very impressive - it just goes back and forth, back and forth, back and forth. ... [Ross] King's humble robot is based on a Biomek 2000, a low-rent fluid-handling device that goes for only $37,900. But it can do something its more nimble cousins can't. Its components - the tireless robot arm, an incubator in which cells cultured on the platter either wither or thrive, and a plate reader that examines the little depressions to see whether anything is growing there - are linked up to a much more exceptional brain. The artificial intelligence routines in that brain can look at the results of an experiment, draw a conclusion about what the results might mean, and then set off to test that conclusion. The 'robot scientist' (King has resisted the temptation of a jazzy acronym) may look like a mere labor-saving gizmo, shuttling back and forth ad nauseam, but it's much more than that. Biology is full of tools with which to make discoveries. Here's a tool that can make discoveries on its own. ... Studying AI at the Turing Institute in Glasgow, [King] set about using machine-learning techniques to predict the shapes of proteins, one of the fundamental challenges of bioinformatics. King, though, found a twist. With his friend Colin Angus, whom he'd met at Aberdeen, he developed software that translated protein structures into musical chord sequences.... Stephen Muggleton argues that the life sciences are peculiarly well suited to machine learning. 'There's an inherent structure in biological problems that lends itself to computational approaches,' he says. In other words, biology reveals the machinelike substructure of the living world; it's not surprising that machines are showing an aptitude for it."
>>> Bioinformatics, Machine Learning, Robots, Scientific Discovery, Applications, Music

July / August 2004: Spotting Cancer Sooner - Blood tests that detect cancer in its early stages would save countless lives. The first could arrive within a year. By Ken Garber. Technology Review. "The individual fates of the 1.3 million Americans diagnosed with cancer this year will be largely decided by one simple factor: at what stage was the disease spotted? ... The problem, of course, is that cancers, which begin with just a few deviant cells, are by their very nature hard to diagnose early. In the last few years, though, a new method has emerged that promises to deliver simple blood tests that identify the telltale molecular profiles of various cancers easily and accurately. ... Like [George] Wright, [Emanuel] Petricoin and [Lance] Liotta used a Ciphergen system to generate protein profiles from blood samples. Their early attempts to find cancer patterns failed, though, because they were simply trying to juggle too much information. Then, in June 1999, a solution appeared. Petricoin and his friend Peter Levine, a Maryland lawyer with a background in data analysis, were chatting about the problem over brunch; Levine suggested using pattern recognition algorithms to make sense of the massive amount of data. Levine, who had considered using such algorithms to analyze stock market trends and commodities trading, sketched out the cancer idea on a napkin. 'In about five minutes, we both realized this would be a really fascinating approach,' Petricoin recalls. So they tested it, together with Ben Hitt, a software engineer who borrowed the necessary algorithms from artificial-intelligence theory. In fact, cancer patterns did emerge, and in 2000 Levine and Hitt founded Correlogic Systems to develop blood tests for cancers. In early 2002, the researchers published results in the British medical journal Lancet , showing they could use a specific protein pattern to spot ovarian cancer."
>>> Medicine, Pattern Recognition, Finance & Investing, Machine Learning, Applications

July 29, 2004: Organic PC goal of UK project. By Harry Yeates. Electronics Weekly. "In the future, alongside the box of lifeless silicon you call your PC, you might find a little tub of living tissue. For particular specialist tasks involving complex, non-linear problems your inorganic circuits would find daunting, you would turn to the box of organics. That's the ultimate aim of a new £1.2m, four year research project involving the universities of the West of England (UWE), Leeds and Sussex. 'For fifty years AI has been trying to build systems that have got complicated behaviour, with some success,' said Dr Larry Bull from UWE, who will lead the project. 'But given this complex behaviour seems to be easy in the natural world, networks of neurons and chemical systems, why don't we try to build AI systems out of that stuff, rather than try to write clever programmes?'"
>>> Systems, Neural Networks, Machine Learning

July 28, 2004: Amplified Intelligence - The AI Problem. Interview with Ken Ford. Astrobiology Magazine. "Astrobiology Magazine (AM): The IMHC [Interdisciplinary Study of Human & Machine Cognition] research agenda broadly seems to cover robotics, cognition and simulations. Are there parts of machine intelligence that your research institute doesn't cover today, but that you see as growth areas? Ken Ford (KF): Don't forget that second letter is 'H'. Although a lot of our research could be categorized as AI, and five of our researchers are AAAI (American Association for Artificial Intelligence) Fellows, IHMC is not a traditional machine intelligence laboratory. The focus and theme of our research is what has become known as human-centered computing which, in a nutshell, is about fitting technology to people instead of fitting people to technology. The human is part of the system, and it is the performance of the whole system, including the human, that we are interested in. This requires that machines should be designed to fit us physically, cognitively, and perhaps even socially. We think of AI as meaning 'Amplified Intelligence.' The interesting thing is that many traditional AI technologies in fact are being used in just this way. We like to refer to it as building cognitive prostheses, computational systems that leverage and extend human intellectual capacities, just as eyeglasses are a kind of ocular prosthesis. Building cognitive prostheses is fundamentally different from AI's traditional Turing Test ambitions -- it doesn't set out to imitate human abilities, but to extend them. ... AM: In your opinion, how well do the machine intelligence problems (like navigation, data-mining, or simulations with agents) map to the basic computer science [CS] problem of efficient 'search'? KF: Wow, efficient search is a 'basic computer science problem'? Not long ago, search was being suggested as a defining characteristic of AI to distinguish it from 'mainstream' CS. But to return to the question: search is certainly a central technique in AI, but the search spaces arising in AI are often impossibly huge, and a more interesting aspect is not so much how to search them efficiently as how to re-cast problems so that the search space itself is reduced in size. Searching is what you do when you can't think of anything smarter."
>>> Interfaces, Search, Robots, Space Exploration, Data Mining, Household Appliances, Interviews, AI Overview, Applications, Reasoning, Machine Learning, Turing Test

July 16, 2004: I.T. May Help Clean a Polluted Sea, Say Researchers. By Mike Martin. NewsFactor Network. "If an article in this week's journal Science is on target, air pollution fouls not only our skies but our oceans as well. ... But software and information technology may play an equally important role, claim the authors of a study published in a recent special issue of the journal Management of Environmental Quality, which is devoted to 'information technologies in environmental engineering.' 'Rapid environmental changes call for continuous surveillance and online decision-making -- two areas where I.T. can be valuable,' say study authors Ioannis Athanasiadis and Pericles Mitkas. Both are computer science researchers at the Informatics and Telematics Institute Center for Research and Technology in Thessaloniki, Greece. In their study, entitled 'An Agent-Based Intelligent Environmental Monitoring System,' the researchers 'present a multi-agent system for monitoring and assessing air-quality attributes, which uses data coming from a meteorological station.' Their system, the study explains, uses a 'community of software agents to monitor and validate measurements coming from several sensors to assess air-quality.' Software agents are computer systems to which an operator can delegate tasks. Like the robots in the new movie 'I, Robot,' software agents are more autonomous, proactive and adaptive than the everyday software we normally use. ... Using agents to monitor the environment is a branch of 'enviromatics -- the research initiative examining the application of information technology in environmental research, monitoring, assessment, management and policy,' Athanasiadis explains. ... 'In O3RTAA, several software agents operate in a distributed-agent society in order to monitor both meteorological and air pollutants, to evaluate air quality and, ultimately, to trigger alarms' about environmental damage, Mitkas explains, adding that the system uses machine-learning algorithms and data-mining methodologies for 'extracting knowledge.'"
>>> Multi-Agent Systems, Data Mining, Natural Resource Management & The Environment, Agents, Machine Learning, Applications

July 16, 2004: Movie tests Asimov's moral code for robots. By Will Knight. New Scientist News. "The possibility of developing truly intelligent machines, and their potential to be friend or foe to humanity, gets the Hollywood treatment in a new blockbuster film I, Robot, which opens in the US on Friday. At the heart of the movie are Isaac Asimov's 'Three Laws of Robotics', invented as a simple, but immutable moral code for robots. ... [R]obotics and artificial intelligence experts admit they are a long way from having to worry about such rules yet. 'The difficulty is building something that would understand them,' says Alan Bundy, at Edinburgh University's Artificial Intelligence Institute in the UK. 'That is well beyond the state of the art at the moment.' Bundy notes that simple safety measures are already a crucial part of the design of industrial robots, which have in rare cases caused the death of people. ... 'Asimov's laws are about as relevant to robotics as leeches are to modern medicine,' says Steve Grand, who founded the UK company Cyberlife Research and is working on developing artificial intelligence through learning. 'They stem from an innocent bygone age, when people seriously thought that intelligence was something that could be 'programmed in' as a series of logical propositions.'"
>>> Robots, Science Fiction, Ethical & Social Implications, Machine Learning, Manufacturing

July 14, 2004: Computer brains. e4engineering.com. "A team of computer scientists and mathematicians at Palo Alto, CA-based Artificial Development are developing software to simulate the human brain's cortex and peripheral systems. As a first step along the way, the company recently disclosed that it has completed the development a realistic representation of the workflow of a functioning human cortex. Dubbed the CCortex-based Autonomous Cognitive Model ('ACM'), the software may have immediate applications for data mining, network security, search engine technologies and natural language processing."
>>> Neural Networks & Connectionist Systems, Machine Learning, Natural Language Processing, Cognitive Science, Data Mining, Information Retrieval, Networks, Applications

July 14, 2004: Attack of the killer vacuum cleaners. By Charles Arthur. The Belfast Telegraph Digital. "Things are about to happen with robots, because the element they need to make them truly useful - the software, which needs to be able to adapt to a wide range of situations - is getting cheaper all the time. Future Horizons, a semiconductor analyst based in Kent, forecasts that by 2010 there will be 55.5 million robots, in a world market worth £30bn - up from £2.4bn last year. 'The electronics industry is on the cusp of a robotics wave, a period in which applications are aimed at labour-saving and extending human skills,' it reports. Of those, it says that 39 million will be domestic robots, and 10.5 million 'domestic intelligent service' robots. That is because there's a growing need for robots to help the elderly and handicapped. ... But the real explosion in robotics is coming among the 'immobots' - or, more simply, just 'bots'. These are bits of software that are incorporated into larger objects, and that remove a lot of the strain of having to decide what to do next. We're getting glimpses of how good these could be at present: the tiny number of Britons with a TiVo personal video recorder have something that decides, based on the programmes they choose to record, what other programmes they might like to see, and records those, too. ... The reason why we can't yet declare 'The Year of the Robot', however, is that researchers are still fundamentally split about how robots should behave and learn. One group favours the 'top-down' approach, in which all the behaviour of the robot is mapped out, and its software is written to fill out that behaviour. The Roomba vacuum cleaner is a classic example.... The alternative is something assembled from smaller, self-contained units, which creates a gestalt of behaviour based on that. Thus the system that controls the legs learns to 'walk' independently.... Sony's Aibo draws on a form of this....
>>> Robots, Science Fiction, Agents, Systems, Assisitive Technologies, Household Appliances, Industry Statistics, Applications, Reasoning, Machine Learning

July 7, 2004: Software aids future tennis stars. BBC News. " As Britons bemoan another year without a Wimbledon hero, there could be some hope in a computer model being worked on at Kingston University in London. ... It will create a computer-generated competitor which rival players can pit themselves against. The system will analyse video footage of champions and allow other players to explore tactics to beat them. ... The research will focus initially on tennis but will move on to look at more complex sports such as football and basketball. 'As well as helping specialised sports training, the technology we are developing could have benefits in fields such as realistic computer gaming, virtual reality and surveillance,' said Dr Ahmed Shihab of the School of Computing and Information Systems at Kingston University."
>>> Sports, Pattern Recognition, Vision, Machine Learning, Video Games, Law Enforcement, Information Retrieval, Applications

July 6, 2004: Programmer seems to have a technology that does everything. By Rachel Melcer. St. Louis Post-Dispatch / STLtoday.com. "Steven Thaler, founder of Imagination Engines Inc. in Maryland Heights, says he has a unique challenge: figuring out what to do with a technology that does everything. He and his supporters say his creation, a computer program called the Creativity Machine, has huge economic potential. It could be the first successful form of artificial intelligence, a machine that learns and thinks by simulating the human brain's activity. ... Imagination Engines also is experimenting with spinoff companies that license the core technology and adapt it for specific uses. ... The first spinoff, Synaptrix Financial Prediction LLC, was created last year as a partner for Stann Financial. It aims to analyze a real-time flow of information on trades in the financial markets to predict the best time to buy or sell a particular stock. The project showed early promise, reaching a 60 percent to 65 percent accuracy rate, but it stalled over problems with the information feed and the need to refine its programming, [John] Stann said. ... Synaptrix Parts Inspection LLC, another of his spinoffs, combines an ordinary video camera with the Creativity Machine's neural network and custom software to perform quality-control checks in manufacturing. The system is 'shown' a variety of objects that it can learn to instantly identify for sorting or to use as an ideal to spot defects and variations. ... On the government side, Imagination Engines is part of a consortium developing an airport-security system for the Department of Homeland Security. The group recently got an 18-month, $800,000 grant to design and test a series of smart sensors at an airport in Butte, Mont. The system would be able to identify vehicles on airport property, monitor them, spot and warn of suspicious activity, Thaler said."
>>> Neural Networks, Image Understanding, Finance & Investing, Manufacturing, Law Enforcement, Machine Learning, Vision, Applications

July 6, 2004: Evolution could speed net downloads. By Will Knight. New Scientist News. "Transferring popular data across the internet repeatedly can be inefficient and costly, so networking companies have developed ways of temporarily storing, or 'caching', data at different locations to reduce costs and increase download speeds. But figuring out where to store data and for how long is a complex problem. One solution might be to have caches 'talk' to each other repeatedly, but this is inefficient as it takes up a lot of bandwidth. To tackle the challenge, Pablo Funes of US company Icosystem and Jürgen Branke and Frederik Theil of the University of Karlsruhe in Germany used 'genetic algorithms', which mimic Darwinian evolution, to develop strategies for internet servers to use when caching data. Using a simulation they were able to improve download speeds over existing caching schemes. ... Funes told New Scientist the scheme could eventually be used to allow caches to automatically 'evolve' their configuration."
>>> Genetic Algorithms, Telecommunications, Networks, Machine Learning, Applications

July 4, 2004: His quest - Do Disney in a day. By Larry Bleiberg. The Dallas Morning News / available from Mickey News. "Rich Vosburgh worked out hard, spending four months with a personal trainer. He scrutinized maps and a detailed timetable. He even deployed a secret weapon: artificial-intelligence research to chart a course through death-defying drops, torrents of water and fiery heat.And when this Texas adventurer clambered out of a floating log a year ago, he had reached his holy grail: visiting - in a single day - each of the 41 operating rides, attractions and shows at the Everest of theme parks, Walt Disney World's Magic Kingdom. His time: a record 10 hours, 40 minutes. ... At heart, the challenge is an enduring and perplexing quandary: What's the most efficient way to route someone to multiple places, taking into account constantly changing conditions? Logistics and timing Mathematicians call it the Time Dependent Traveling Salesman Problem. The answer could help fighter-jet pilots chart bombing targets or freight companies schedule package deliveries."
>>> Traveling Salesperson Problem, Planning and Scheduling, Search, Genetic Algorithms, Machine Learning, Reasoning, Transportation, Applications, Games & Puzzles

July 2004 [issue date]: Homeland Security as Catalyst - Innovative software firms are answering the call from U.S. government agencies for advanced analytics to help combat terrorism and criminal activity. What's the potential of this software for strategic business applications? By Jesus Mena. Intelligent Enterprise Magazine. "Ever heard of NORA? Or how about these guys: InferAgent, CopLink, NameHunter, Bladeworks, and Sentinel? These ominous-sounding fellows are products from tiny software firms that are developing some of the most advanced analytic technologies today for homeland security. Some provide solutions for the conversion of garbled text into knowledge discovery. Others tend to the unearthing of associations of individuals to actions, locations, and events from hundreds of thousands of internal and external records. Still others offer innovative methods for detecting fraud, categorizing foreign names, and virtual, remote analysis of data or text from any database in the world for agencies such as the U.S. Department of Homeland Security's Terrorist Threat Integration Center (TTIC). Given the growing diversity and globalization of business enterprises, is it possible that these innovative technologies, finding clear purpose for homeland security, could also be of interest to private business enterprises? In this article, I will describe some of these new technologies and how they may be applied to your company today and tomorrow. Who Are These Guys? Innovative products I mentioned at the beginning are commercial off-the-shelf (COTS) software -- a term favored by military and government agencies -- originating from such companies as Attensity, InferX, Infoglide, Knowledge Computing Corp. (KCC), Language Analysis Systems (LAS), Searchspace, System Research & Development (SRD), and others. Almost all have developed applications based on artificial intelligence technologies to meet demand from first military and intelligence communities, and now from the emerging homeland security market."
>>> Law Enforcement, Military, Information Extraction, Business, Knowledge Management, Applications, Machine Learning, Natural Language Processing, Agents

Spring 2004: What We Don't Know Can Hurt Us. By Heather Mac Donald. City Journal (Vol. 14, No. 2). "Immediately after 9/11, politicians and pundits slammed the Bush administration for failing to 'connect the dots' foreshadowing the attack. What a difference a little amnesia makes. For two years now, left- and right-wing advocates have shot down nearly every proposal to use intelligence more effectively -- to connect the dots -- as an assault on 'privacy.' Though their facts are often wrong and their arguments specious, they have come to dominate the national security debate virtually without challenge. The consequence has been devastating: just when the country should be unleashing its technological ingenuity to defend against future attacks, scientists stand irresolute, cowed into inaction. 'No one in the research and development community is putting together tools to make us safer,' says Lee Zeichner of Zeichner Risk Analytics, a risk consultancy firm, 'because they're afraid' of getting caught up in a privacy scandal. The chilling effect has been even stronger in government. 'Many perfectly legal things that could be done with data aren't being done, because people don't want to lose their jobs,' says a computer security entrepreneur who, like many interviewed for this article, was too fearful of the advocates to let his name appear. ... The goal of TIA [the Total Information Awareness project] was this: to prevent another attack on American soil by uncovering the electronic footprints terrorists leave as they plan and rehearse their assaults. ... TIA would have been the most advanced application yet of a young technology called 'data mining,' which attempts to make sense of the explosion of data in government, scientific, and commercial databases. Through complex algorithms, the technique can extract patterns or anomalies in data collections that a human analyst could not possibly discern. ... Without question, TIA represented a radical leap ahead in both data-mining technology and intelligence analysis, not surprising for a visionary group like DARPA, which created the Internet. ... As with any public or private power, TIA's capabilities could have been abused -- which is why DARPA planned to build safeguards throughout the system. But it differed from existing law enforcement and intelligence techniques only in degree, not kind. Though the scale of data it would have made immediately available to government was unprecedented, the type of evidence was identical to what government had had legal access to for decades. ... Information technology can help government in its constitutional responsibilities to protect the nation; indeed the congressional jo int inquiry into September 11 found that 'a reluctance to develop and implement new technical capabilities aggressively' was a cause of the pre-9/11 intelligence failures. The report added: 'While technology remains one of this nation's greatest advantages, it has not been fully and most effectively applied in support of U.S. counterterrorism efforts.' The privocrats will rightly tell you that eternal vigilance is the price of liberty; trouble is, they are aiming their vigilance at the wrong target." [Other projects discussed in this article: Human Identity at a Distance ; LifeLog; CAPPS II, Computer Assisted Passenger Prescreening System; MATRIX, Multistate Anti-Terrorism Information Exchange; and FIDNet.]
>>> Ethical & Social Implications, Data Mining, Law Enforcement, Military, Applications, Agents, Vision, Machine Learning

June 29, 2004: Panel Seeks Protections From Data Mining. By Brian Bergstein. Associated Press / available from The Herald News. "Even as the government increasingly relies on of data mining - scouring databases in search of clues about terrorism and everyday waste and fraud - there aren't clear rules about the practice. Privacy activists say it's like the wild West, dangerously unregulated. ... The data mining frontier could finally be seeing some civilizing influences take shape, particularly in the recommendations of a panel headed by former Federal Communications Commission chief Newton Minow that are getting particular praise. The panel's report, released in early June, acknowledged the importance of data mining in fighting terrorism. But it also said broad searches through reams of records and commercial files, on citizens who have done nothing to warrant individual suspicion, threaten fundamental protections in the Bill of Rights. To strike a balance, the group, known as the Technology and Privacy Advisory Committee (TAPAC), called for technological changes that would 'anonymize' data so investigators could hunt for suspicious activities and associations without immediately knowing whom they were probing."
>>> Ethical & Social Implications, Data Mining, Law Enforcement, Machine Learning, Applications

June 24, 2004: 2020 Vision has CCTV intelligent cameras deal in focus. The Journal / available from ic Newcastle. "Security specialist 2020 Vision Systems has secured an exclusive deal to provide artificial intelligence systems for CCTV cameras. The technology developed by Australian company, iOmniscient, allows security cameras to 'learn' to recognise anomalies in an area while ignoring routine movements. ... Using the technology, a camera can be 'taught' to recognise when valuable objects - such as paintings in a gallery - are moved, while ignoring people walking."
>>> Law Enforcement, Machine Learning, Applications, Vision

June 24, 2004: SkyNet Autonomy - Smart Satellite to Monitor Flood Gates. By Ed Stiles. University of Arizona report / available from Astrobiology Magazine. "There's nothing worse than a satellite that can't make decisions. Rather than organizing data, it simply spews out everything it collects, swamping scientists with huge amounts of information. It's like getting a newspaper with no headlines or section pages in which all the stories are strung together end-to-end. Researchers at the University of Arizona (UA), Arizona State University (ASU) and the Jet Propulsion Laboratory (JPL) are working to solve this problem by developing machine-learning and pattern-recognition software. This smart software can be used on all kinds of spacecraft, including orbiters, landers and rovers. Scientists currently are developing this kind of software for NASA's EO-1 satellite. The smart software allows the satellite to organize data so it sends back the most timely news first, while holding back less-timely data for later transmission. Although the project, called the Autonomous Sciencecraft Experiment (ASE), is still in the test and development stage, software created by UA hydrologists has already detected flooding on Australia's Diamantina River. ... The flood-detection software compares images from the satellite's cameras with images stored in its computer memory. If the rivers are not flooding and images come close to matching, the satellite remains silent. But if the satellite's computer finds significant differences, it takes more photos and notifies scientists."
>>> Space Exploration, Earth Science, Natural Resource Management, Pattern Recognition, Machine Learning, Applications

June 24, 2004: Informed decisions - CHEO team tests artificial intelligence in neo-natal unit. By Andrew Mayeda. Ottawa Citizen (subscription required). "When a baby is born prematurely, parents must often make a heartbreaking decision of whether to continue care or to simply let go. While that decision will never be easy, a pair of Ottawa researchers have developed artificial-intelligent tools that could at least make it more informed. The result is a software system [Parents Assisting Decision Support] that lets parents know their child's chances of survival, and allows them to weigh the pros and cons of treatment options while consulting their doctor or nurse. ... PADS, as it is called for short, is the brainchild of Dr. Robin Walker and Monique Frize, who have worked together for more than a decade."
>>> Case-Based Reasoning, Neural Networks, Medicine, Applications, Reasoning, Machine Learning

June 23, 2004: The Futurist - The Intelligent Internet. The Promise of Smart Computers and E-Commerce. By William E. Halal. Government Computer News Daily News (GCN). "Information and communication technologies are rapidly converging to create machines that understand us, do what we tell them to, and even anticipate our needs. We tend to think of intelligent systems as a distant possibility, but two relentless supertrends are moving this scenario toward near-term reality. Scientific advances are making it possible for people to talk to smart computers, while more enterprises are exploiting the commercial potential of the Internet. ... [F]orecasts conducted under the TechCast Project at George Washington University indicate that 20 commercial aspects of Internet use should reach 30% 'take-off' adoption levels during the second half of this decade to rejuvenate the economy. Meanwhile, the project's technology scanning finds that advances in speech recognition, artificial intelligence, powerful computers, virtual environments, and flat wall monitors are producing a 'conversational' human-machine interface. These powerful trends will drive the next generation of information technology into the mainstream by about 2010. ... The following are a few of the advances in speech recognition, artificial intelligence, powerful chips, virtual environments, and flat-screen wall monitors that are likely to produce this intelligent interface. ... IBM has a Super Human Speech Recognition Program to greatly improve accuracy, and in the next decade Microsoft's program is expected to reduce the error rate of speech recognition, matching human capabilities. ... MIT is planning to demonstrate their Project Oxygen, which features a voice-machine interface. ... Amtrak, Wells Fargo, Land's End, and many other organizations are replacing keypad-menu call centers with speech-recognition systems because they improve customer service and recover investment in a year or two. ... General Motors OnStar driver assistance system relies primarily on voice commands, with live staff for backup; the number of subscribers has grown from 200,000 to 2 million and is expected to increase by 1 million per year. The Lexus DVD Navigation System responds to over 100 commands and guides the driver with voice and visual directions. ... BCC Corporation estimates total AI sales to grow from $12 billion in 2002 to $21 billion in 2007. ... This scenario is not without uncertainties. Cynicism persists over unrealized promises of AI, and the Intelligent Internet will present its own problems. If you think today's dumb computers are frustrating, wait until you find yourself shouting at a virtual robot that repeatedly fails to grasp what you badly want it to do. ... The main obstacle is a lack of vision among industry leaders, customers, and the public as scars of the dot-com bust block creative thought."
>>> AI Overview, Applications, Natural Language Processing, Speech, Interfaces, Systems, Machine Learning, Customer Relations & E-Commerce, Information Retrieval, Networks, Industry Statistics

June 21, 2004: The Future of Business Intelligence & Predictions For BI's Future By Mitch Betts. Computerworld. [These articles are part of their special Business Intell igence report.] "We asked some industry leaders for their boldest predictions about the future of business intelligence tools, and here's our collection of the most interesting ideas. ... BI meets AI. In the near future, business leaders will manage by exception, and automated systems will handle significant loads of routine tasks. Today, automated systems in banking match incoming customer requests and inquiries with basic cross-sell and upsell oriented advertising. Over the next five years, these systems will become increasingly complex by considering customer financial status and wealth, transactional history, and even family and business relationships, to produce complex man/machine interactions that resemble artificial intelligence. The viability of artificial intelligence to solve real-world problems is being made possible by the convergence of hardware capabilities (faster processors, memory expansion and higher bandwidth) and sophisticated software (neural networks, probability models and rules analysis). -- Mike Covert, chief operating officer, Infinis Inc., Columbus, Ohio ... Automatic insurance decisions. By 2009, 50% of all insurance underwriting decisions will be automated using data mining technology. -- Richard Vlasimsky, chief technology officer, Valen Technologies."
>>> Business, Banking, Customer Relations, Image Understanding, Data Mining, Expert Systems, Neural Networks, Probability, Machine Learning, Reasoning, Vision, Systems, Applications

June 21, 2004: Text mining tools take on unstructured data - Companies are increasingly using text mining tools to harness the information in their unstructured data. By Drew Robb. Computerworld. [This article is part of their special Business Intell igence report.] "Unstructured data, most of it in the form of text files, typically accounts for 85% of an organization's knowledge stores, but it's not always easy to find, access, analyze or use. ... But a new generation of text mining tools allows companies to extract key elements from large unstructured data sets, discover relationships and summarize the information. Many organizations are deploying or considering such software to deal with their mountains of text, despite the need for specialized skills to make implementations work. ... Text mining tools take a variety of approaches. ClearResearch uses a proprietary pattern-matching methodology to search for information, categorize it and graphically show its relationship to other data. 'The software can see, discover and extract concepts, not just words," says Shabrang. "It gives us a pictorial representation of the text in the documents in an easy-to-understand chart.'"

June 18, 2004: Breeding Race Cars to Win. By Michelle Delio. Wired News. "A technology that allows robots to rebuild themselves and computer programs to evolve and become better on their own is now being used to breed super-fast Formula One race cars. ... The breeding was done solely with computer-generated simulations using genetic algorithms -- programs that combine Mother Nature's laws and computer science to mimic the natural process of evolution. Using this sort of programmed procreation, the Digital Biology Interest Group [at University College London] has made self-healing battlefield surveillance robots -- gadgets that look like robotic snakes that can figure out how to wiggle home even when severely damaged, unlike less-evolved robots that typically just give up when one of their critical components goes out of commission."
>>> Genetic Algorithms, Engineering, Machine Learning, Robots, Applications

June 16, 2004: Research: From lab to market. By Michael Kanellos. CNET News. "Data mining, the ability to find unexpected patterns in accumulated data, was born during a lunch break. At a customer conference in the early 1990s, an executive at British department store chain Marks & Spencer was explaining his database woes to Rakesh Agrawal, an information retrieval specialist at IBM. The store was collecting all sorts of data but didn't know what to do with it. So Agrawal and his team began devising algorithms for asking open-ended queries, eventually authoring a 1993 paper that would become required reading in data-mining science. The report has been cited in more than 650 other studies, making it one of the most widely cited papers of its kind. ... Agrawal, the data-mining pioneer, is today working on a system that will scramble customer data in a way that will allow companies to study buying trends or other patterns while preserving strict privacy. ... In its Beijing labs, researchers are tinkering with handwriting recognition systems for Asian languages and a digital home in which appliances--lights, alarm systems, dishwashers, computers--can be operated through voice commands."
>>> Data Mining, History, Machine Learning, Natural Language Processing, Networks, Information Retrieval, Marketing, Smart Houses, Applications, Ethical & Social Implications

June 15, 2004: NASA Evolutionary Software Automatically Designs Antenna. Press release available from SpaceRef. "NASA artificial intelligence (AI) software - working on a network of personal computers - has designed a satellite antenna scheduled to orbit Earth in 2005. The antenna, able to fit into a one-inch space (2.5 by 2.5 centimeters), can receive commands and send data to Earth from the Space Technology 5 (ST5) satellites. ... NASA scientists have spent two years developing the evolutionary AI software that designed the antenna. 'The AI software examined millions of potential antenna designs before settling on a final one,' said project lead Jason Lohn, a scientist at NASA's Ames Research Center, located in California's Silicon Valley. 'Through a process patterned after Darwin's 'survival of the fittest,' the strongest designs survive and the less capable do not.' The software started with random antenna designs and through the evolutionary process, refined them. The computer system took about 10 hours to complete the initial antenna design process. ... 'Not only can the software work fast, but it can adapt existing designs quickly to meet changing mission requirements,' he said. ... Scientists also can use the evolutionary AI software to invent and create new structures, computer chips and even machines, according to Lohn. ... 'The software also may invent designs that no human designer would ever think of,' Lohn asserted."
>>> Genetic Algorithms, Engineering, Space Exploration, Telecommunications, Machine Learning, Applications

June 14, 2004 [issue date]: Innovators / Artificial Intelligence: Forging the Future - Rise of the Machines - These visionaries are making robots that can perform music, rescue disaster victims and even explore other planets on their own. By Dan Cray, Carolina A. Miranda, Wilson Rothman, Toko Sekiguchi. Time Magazine. "The Bionic Engineer - Driving School On Mars. Television critics will tell you that The Bionic Woman was just another cheesy '70s sci-fi series, but for Ayanna Howard it was a springboard to a career. When she was 12 years old, she became so captivated by the show's cyborg premise that she started reading books that reaffirmed the concept of integrating machines with humans. A thousand reruns and an electrical-engineering Ph.D. later, she's creating robots that think like humans for NASA's Jet Propulsion Laboratory. ... Three years ago, hoping to encourage others to follow in her footsteps, Howard launched a math-and-science mentoring program for at-risk junior high school girls. ... Howard hopes the program will help steer more young women into robotics, a field she says that within a decade will produce robots that mimic human thought processes. ... The Swarm Keeper - Metal Insects On Wheels. When James McLurkin was a high school junior on Long Island, N.Y., he built his first robot: a toy car that he rigged with a keypad, an LED display and a squirt gun. ... Now a graduate student in computer science at M.I.T., the young scientist is on the forefront of developing 'swarmbots'--packs of dozens of small robots that communicate with one another and work in harmony to complete an assignment. They have no centralized command system and can cover vast terrain; if one is destroyed, others fill in. ... Rescuer By Remote - Need Help? Send In The Robot. Within 24 hours of the 9/11 attacks on the World Trade Center, Robin Murphy was on the scene with a team of robots to help sort through the debris. It was the first real-world test of the Center for Robot-Assisted Search and Rescue in Tampa, Fla., the only unit of its kind on the planet. ... The Mimic Maker - The Android Who Learned To Dance. Mitsuo Kawato is fascinated with the brain -- so he helped build one. The biophysics engineer and computer researcher led a team at the Advanced Telecommunications Research Institute International in Kyoto, Japan, that spent five years constructing a humanoid equipped with artificial intelligence. Completed in 2001, the 6-ft. 2-in., 175-lb. robot was named Dynamic Brain, or DB for short. Says Kawato: 'We built an artificial brain hoping that it'll help us understand the real one.' ... So far, the robot has acquired about 30 skills, including juggling, air hockey, yo-yoing, folk dancing and playing the drum."
>>> AI Overview, Space Exploration, Neural Networks, Reasoning, Robots, Multi-Agent Systems, Artificial Life, Military, Hazards & Disasters, Applications, Machine Learning, Cognitive Science, Careers in AI (@ Resources for Students)

June 13, 2004: A Computer That Has an Eye for Van Gogh. By Douglas Heingartner. The New York Times (no fee reg. req'd.). "Who can say for sure that a great artwork is the real deal? ... Now a team of researchers at the University of Maastricht, here in the Netherlands, are taking a stab at rationalizing connoisseurship, a word that in its art-historical context refers to the formal process of determining who created a work of art. They have developed a computer system that quickly examines hundreds of paintings for telltale patterns. The results, they say, can lend credence to existing attributions or help dismiss them. Members of the team make modest claims for their system. 'The computer will come up with data that show some patterns, but we cannot decide whether these patterns are meaningful or not,' said Dr. Eric Postma, the leader of the project, known as Authentic, which is currently analyzing all paintings attributed to Vincent van Gogh. 'For that purpose we need experts. We can provide them with numbers, and they can interpret the numbers. And this interaction is where the real value of the project is.' ... Dr. Postma compares this pattern-seeking technique to chess. ... This is not the first time artificial intelligence has been used in authentication. In Germany in 1998, a team at the University of Bremen's Center for Computing Technologies trained their computer to identify the drawings of Delacroix, which it managed to do with 87 percent accuracy. ... In a more recent project at the Catholic University of Rio de Janeiro, a computer distinguished between 23 paintings made by the popular Brazilian painter Candido Portinari and five by his contemporary Enrico Bianco."
>>> Art, Pattern Recognition, Chess, Machine Learning, Law Enforcement, Applications

June 10, 2004: A golden vein - Computing: Analysis of customer information, better known as "data mining", is finally delivering on its promises-and expanding into some promising new areas. The Economist Technology Quarterly. "In the old days, knowing your customers was part and parcel of running a business, a natural consequence of living and working in a community. But for today's big firms, it is much more difficult: a big retailer such as Wal-Mart has no chance of knowing every single one of its customers. So the idea of gathering huge amounts of information and analysing it to pick out trends indicative of customers' wants and needs -- data mining -- has long been trumpeted as a way to return to the intimacy of a small-town general store. But for many years, data mining's claims were greatly exaggerated. ... In recent years, however, improvements in both hardware and software, and the rise of the world wide web, have enabled data mining to start delivering on its promises. Richard Neale of Business Objects, a software company based in San Jose, California, tells the story of a British supermarket that was about to discontinue a line of expensive French cheeses which were not selling well. But data mining showed that the few people who were buying the cheeses were among the supermarket's most profitable customers -- so it was worth keeping the cheeses to retain their custom. As data mining has matured, examples like this are plentiful. ... The traditional British pub seems like an unlikely place to find the latest in data mining. But some pub chains now change the prices of different drinks from day to day, using software that assesses the impact that 'happy hour' offers have on sales. ... Privacy advocates have long been wary of data mining, demonising supermarket loyalty cards, for example, as 'spies in your shopping'. Like any technology, of course, it can be misused. ... Forrester predicts that sales of BI [business intelligence] software, currently around $2 billion a year, will grow by 8.5% a year over the next three years. If new tricks like predictive analytics and unstructured-data analysis catch on, that could prove to be a conservative figure."
>>> Data Mining, Marketing, Business, Fraud Detection & Prevention, Law Enforcement, Machine Learning, Applications, Ethical & Social Implications, Industry Statistics

June 10, 2004: Fuzzy logic and neural nets: still viable after all these years? Though no longer headliners, fuzzy logic and neural networks are options in tackling challenging applications. By Graham Prophet. EDN Magazine. "[B]oth still have their place in your engineering tool kit. The two techniques are essentially unrelated, except that they both provide control methodologies to handle highly nonlinear or poorly specified problems, they both came to some prominence at about the same time, and they both faded from view in much the same way. Both neural networks and fuzzy logic aspire to allow electronic systems, built with familiar circuit techniques or employing conventional computing technologies, to attack certain problems in a way that mimics human responses and abilities. ... One of the intimidating aspects of fuzzy logic is the name itself, which has connotations of imprecision. On the contrary, however, fuzzy logic is capable of precise responses. It allows systems built around Boolean logic, handling binary values, to work with imprecisely defined values that you might express verbally as 'more,' 'less,' 'high,' 'low,' and so on. ... Neural networks, unlike fuzzy logic, seek to reproduce the versatility of the human brain in recognizing the end-to-end, input-to-output behavior of a system without understanding all the processes taking place within it. Taking as a fundamental model the interconnections of nervous systems within the brain -- neurons and synapses -- neural networks have the attributes of memory and learning. ... What happens to the expertise built up in neural and fuzzy techniques from their first flush of popularity? If you set about tracking down some of the pioneering companies from as much as a decade ago, you'd find that, although many no longer exist, some have transformed themselves into software-design and consultancy operations. These businesses are applying the same neural and fuzzy techniques but mainly in software simulation running on conventional computers, in areas such as financial modeling, financial services, and data mining."
>>> Fuzzy Logic, Neural Networks, Applications, Banking, Finance & Investing, Data Mining, Reasoning, Machine Learning

June 10, 2004: Brain learns like a robot - Scan shows how we form opinions. By Tanguy Chouard. Nature Science Update. "Researchers may have pinpointed the brain regions that help us work out good from bad. And their results suggest that humans and robots are more alike than we may care to admit, as both use similar strategies to make value judgements. ... The team also plotted brain activity on a graph to give a mathematical description of processes that underlie the formation of value judgements. The patterns they saw resembled those made by robots as they learn from experience. 'The results were astounding,' says study co-author Peter Dayan. 'There was an almost perfect match between the brain signals and the numerical functions used in machine learning,' he says. This suggests that our brains are following the laws of artificial intelligence."
>>> Cognitive Science, Machine Learning, Robots

June 8, 2004: Man who cracked computer engima. Opinion by Andrew Hodges. Edinburg Evening News / available from Scotsman.com News. "[Alan] Turing was fascinated by the concept of creating a mathematical machine to represent thought processes, and it was the 'Turing Machine' which became the foundation of the modern theories of computer science. He also envisaged a 'Universal Turing Machine' - one machine for all possible tasks - which embodied the essential principle of the computer. Turing's originality lay in seeing the relevance of mathematical logic to a problem originally seen as one of physics. He made a bridge between thought and action, which crossed conventional boundaries. All this was when he was just 24. Then he left Cambridge for a spell at Princeton and right away saw a link from 'useless' logic to practical purposes. ... In 1944, following the invasion of Normandy that Allied control of the Atlantic allowed, Alan Turing was almost uniquely in possession of three key ideas - his own 1936 concept of the universal machine, the potential speed and reliability of electronic technology and the inefficiency in designing different machines for different logical processes. Combined, these ideas provided the principle, the practical means and the motivation for the modern computer. ... From October 1947, the National Physical Laboratory allowed, or perhaps preferred, that he should spend the academic year at Cambridge. Out of this came a pioneering paper on what would now be called neural nets. ... Though marginalised in practice, he published his theoretical ideas on artificial intelligence in 1950 in a paper which is now one of the most quoted in science. His 'Turing Test' for intelligent machinery now has a long and entertaining history."
>>> History, Alan Turing (@ Namesakes), Neural Networks, Turing Test, Machine Learning

June 7, 2004: Cognitive Personal Assistant. AI-based systems could handle routine administrative tasks. Future Watch by Thomas Hoffman. Computerworld. "Researchers at Carnegie Mellon University are developing a computer-based administrative assistant that draws upon artificial intelligence (AI) techniques to perform routine tasks such as scheduling meetings for busy managers and filtering and prioritizing their e-mail. ... The project, called Radar (short for Reflective Agent with Distributed Adaptive Reasoning), is being funded by the Defense Advanced Research Projects Agency under a program called PAL, or Personalized Assistant that Learns. ... Using AI, Radar will draw on statistical and symbolic learning. Say a manager demonstrates a tendency to deny e-mail requests to hold meetings on Fridays over the course of a few months. Radar will pick up on this pattern and send a message to the manager asking whether the manager prefers to avoid meetings on Fridays."
>>> Agents, Natural Language Processing, Machine Learning, Probability, Applications

June 4, 2004: Programs of the Mind. Review by Gary Marcus. Science Magazine (subscription required). "Eric Baum's What Is Thought? [MIT Press, Cambridge, MA, 2004], consciously patterned after [Erwin] Schrödinger's book [What Is Life?], represents a computer scientist's look at the mind. Baum is an unrepentant physicalist. He announces from the outset that he believes that the mind can be understood as a computer program. Much as Schrödinger aimed to ground the understanding of life in well-understood principles of physics, Baum aims to ground the understanding of thought in well-understood principles of computation. In a book that is admirable as much for its candor as its ambition, Baum lays out much of what is special about the mind by taking readers on a guided tour of the successes and failures in the two fields closest to his own research: artificial intelligence and neural networks. ... Advocates of what the philosopher John Haugeland famously characterized as GOFAI (good old-fashioned artificial intelligence) create hand-crafted intricate models that are often powerful yet too brittle to be used in the real world. ... At the opposite extreme are researchers working within the field of neural networks, most of whom eschew built-in structure almost entirely and rely instead on statistical techniques that extract regularities from the world on the basis of massive experience."
>>> AI Overview, Cognitive Science, Philosophy, Neural Networks, Machine Learning

FOR MORE ARTICLES, SEE THE MACHINE LEARNING NEWS ARCHIVE