A Preliminary Evaluation of the Impact of Syntactic Structure in Semantic Textual Similarity and Semantic Relatedness Tasks

The well related tasks of evaluating the Se- mantic Textual Similarity and Semantic Relat- edness have been under a special attention in NLP community. Many different approaches have been proposed, implemented and evalu- ated at different levels, such as lexical similar- ity, word/string/POS tags overlapping, seman- tic modeling (LSA, LDA), etc. However, at the level of syntactic structure, it is not clear how significant it contributes to the overall ac- curacy. In this paper, we make a preliminary evaluation of the impact of the syntactic struc- ture in the tasks by running and analyzing the results from several experiments regarding to how syntactic structure contributes to solving these tasks.