Equivalence detection using parse-tree normalization for math search

In recent years, efforts have begun to put math contents on the Web. As for other types of Web information, search capabilities should be provided to enable users to find what they need because without the ability to search the data for specific items, the data is useless. Conventional (i.e. textbased or even multimedia-based) search engines fall short of providing math-search capabilities. Preliminary efforts to create math-search systems have started, and many of the issues and the challenges for building such systems have been identified. One of the more difficult challenges is the detection of mathematical equivalence between expression in users’ queries and expressions in math contents. The purpose of this research is to develop techniques and algorithm for equivalence-detection based math search. In particular, this research aims to explore some proposed normalization rules, then to develop a general way that can be utilized to transform both the repository contents and users’ input expressions into a unified normalized form.