Association-Based Segmentation for Chinese-Crossed Query Expansion

The continually and high-rate growth of China's economy has attracted more and more international investors. These investors have an urgent need of identifying patterns in Chinese information, which are potentially useful in making competitive decisions. The first step of deeply understanding and analyzing Chinese information is how to effectively search those likely relevant to a user query. However, queries provided by users are often incomplete and inappropriate to the information systems, especially for retrieving Chinese-crossed information. In this paper, we present a segmentation based on actionable Chinese term-association analysis for better understanding user queries so as to automatically generate Chinese-crossed-query expansions. The semantics behind the actionable term-association rules is thus studied. Experiments conducted have shown that our approach is efficient and promising.