Naohiro Matsumura, Hikaru Yamamoto, Daisuke Tomozawa
In this paper, we attempt to specify the influencer and the influential terms in consumer package goods by using the influence diffusion model (IDM). IDM calculates the spread of a term in the blogosphere recursively, and evaluates the influence of terms, blog entries, and bloggers. The model was applied to the official data from a blog hosting service. The results of the model were compared with the term frequency (TF). Then, the influencers of the IDM were examined and compared with the results of other network indices such as number of links and trackbacks. The results indicate that IDM performs better in finding blogs posted by bloggers with product usage experience or with an interest in the product. A propagation network which visualizes the consumer-oriented key marketing elements is also presented. An empirical analysis suggests the validity of IDM for marketing decision marking.
Subjects: 1.10 Information Retrieval; 10. Knowledge Acquisition
Submitted: Feb 15, 2008