Wrapper induction of news information for feeding to social networking service on smartphone

In this paper, we propose NewsFeedAndroid, a novel system that interconnects a social networking service and online newspaper sites in order to extracts news articles from the online news sites and to perform feeding of news articles to social network service (SNS) users. In NewsFeedAndroid, news information agents extract news article information from the news and portal sites using Minimum Description Length (MDL) wrapper induction algorithm. The news document collecting module regularly gathers news list information from news list page in the news sites and portals. In the collected documents, the document preprocessing module removes tags that are unnecessary for news information extraction. Lexical analyzer converts the rest text information and tags to a sequence of tokens, and news information is obtained by matching token patterns to the sequence. Those extracted news information from the various sites are integrated in the system and supplied to the end users through the social networking service on a smartphone. NewsFeedAndroid demonstrates a novel usage of integrating social networking services and online newspaper sites.