フォロワーの興味抽出における系列パターン評価指標利用の検討

抄録

Extracting users' interests from text data is one of the most important issues of people, who want to target adequate customer in many business fields. On the other hand, twitter follow-followed relationship is useful for gathering customers, who are interested in each account. However, the relationships do not reflect detailed interest of users, and they still need to know about the detailed interest by using some keyword extraction methods of tweets of the followers. In this paper, I apply some sequential pattern evaluation indices for mining key-phrases from the text data, which gathered from the followers of three famous e-commerce accounts.