Extracting and searching news articles in web portal news pages

摘要

Recently, a large amount of news articles is being created online, and news articles are important resources for understanding social phenomena and trends. Accordingly, a web portal service provides a 'portal news page' that classifies news articles published from various news sources into sections and provides each news article with a certain structure. Therefore, by analysing portal news pages, it is possible to automatically extract information about news articles. In this paper, we introduce a prototype that extracts and searches key information of news articles for analysis. Specifically, we describe: 1) a crawler that collects, analyses and parses news articles; 2) an Elasticsearch server that indexes and searches news information; and 3) a front-end application that provides a search user interface. These systems are expected to provide the foundation for news analytics and forecasting services.