Advertising slogan generation system reflecting user preference on the web

摘要: Increased demand for Web advertising has resulted in a corresponding increase the need to develop online personalized advertisements. This paper proposes an advertising-slogan generation system reflecting Web-user preferences. Using social networking service (SNS) site as knowledge base word preferences, and by employing slogan corpus, proposed aims generate slogans that reflect posts on SNS. model selected from corpus containing 24,472 slogans, generates candidates using obtained post These are based following three indexes: natural level given large-scale balanced semantic-relations score preference SNS sites. In particular, extracts data these fan pages estimates each bag-of-words model. enables select timely fashion. The authors conducted subjective experiment examine quality of generated slogans. results show (1) semantic-relation levels effective selecting post, (2) preference-level index contributes selection preferred interesting users.