Social Media Analysis for Accurate Association Relational Engine

社交媒体系统允许用户在网络上协作构建内容。由于其独特的覆盖范围和准确性等特点，已成为一种很有前途的知识抽取语料库。然而，社交媒体分析存在两大技术问题；可扩展性和可信度。在本次研究项目中，为了解决这些问题，我们进行了一系列的研究；1)可扩展关联提取，2)寻找重要特征的机器学习，3)大规模测试集合构建和4)跨不同类型资源的集成分析。

Social media systems allow users to construct contents collaboratively on the Web. Because of the unique characteristics such as coverage and accuracy, it has become one of the promising corpora for knowledge extraction. However, there exist two major technical issues on social media analysis ; scalability and credibility. In this research project, to solve these problems, we have conducted a series of research ; 1) scalable association extraction, 2) machine learning for finding important features, 3) a large scale test collection construction and 4) integrated analysis across different types of resources.