Hidden Markov Modeling for Semantic Analysis—On the Combination of Different Decoding Strategies

摘要

Different strategies to enhance the semantic decoding accuracy of a stochastic parser are discussed and comparatively evaluated on a corpus containing dialogues between two persons scheduling a meeting. Using a stochastic parsing method the human effort can be limited to the task of data labeling, which is much simpler than the design, maintenance and extension of grammar rules, especially for non-experts. Since a stochastic method automatically learns the semantic formalism through an analysis of these data, it is comparatively flexible and robust and can easily be ported to different applications, domains and human languages. The performance of the parser was improved by subsequently adding valuable and removing redundant semantic information, as well as by combining several decoding methods either sequentially or in parallel.