Personalized Internet Advertisement Recommendation Service Based on Keyword Similarity
Wei-Hao Hwang, Yeong-Sheng Chen, Tsang-Ming Jiang · 2015
Internet advertising, or E-Marketing, is an important marketing tool in today's on-line world. This study focus on developing a framework for personalized Internet advertisement recommendation service. At first, all the advertisements are categorized by referring to the commercial categories provided by Yahoo. Then, keywords are extracted from each advertisement, and the most correlated keywords to each category are identified through Term Frequency and Inverted Domain Frequency (TF-IDF) analysis. Thus, the ontology of the advertisement world is built. Normalized Google Distance (NGD) values between keywords are computed to derive the characteristic vector of each advertisement. Also, through Logistic Regression, the user profile, which describes a user's preferences, is established based on the user's responses to some certain advertisements. Finally, for a new advertisement, a recommendation value is computed by using the characteristic vector of this advertisement and the user profile. The value is thus used to determine whether this advertisement should be recommended to the user or not. A prototype website for verifying the proposed schemes was developed. Experiment results showed that better marketing effectiveness can be achieved with the proposed personalized advertisement recommendation.