A graph-based approach to mining multilingual word associations from wikipedia

Zheng Ye, Xiangji Huang, Hongfei Lin

Abstract: In this paper, we propose a graph-based approach to constructing a multilingual association dictionary from Wikipedia, in which we exploit two kinds of links in Wikipedia articles to associate multilingual words and concepts together in a graph. The mined association dictionary is applied in cross language information retrieval (CLIR) to verify its quality. We evaluate our approach on four CLIR data sets and the experimental results show that it is possible to mine a good multilingual association dictionary from Wikipedia articles.