1. What are the contributions in "Semantic processing using the hidden vector state model" ?
This paper discusses semantic processing using the Hidden Vector State ( HVS ) model.. Furthermore, when measured by its ability to extract attribute-value pairs from natural language queries in the travel domain, the HVS model outperforms a conventional finite-state semantic tagger by 4.. 6 % in F -measure for Communicator, suggesting that the benefit of the HVS model ’ s ability to encode context increases as the task becomes more complex.
read more
2. What are the future works mentioned in the paper "Semantic processing using the hidden vector state model" ?
Another important topic for future work is to port the HVS model to a real dialogue application and explore the possibilities for on-line learning.. For example, if the mapping from an input utterance to semantic interpretation was initially wrong but was then repaired during the subsequent dialogue, it may be possible to use the corrected semantics to update the HVS model online.. Improved performance may be gained by using the decoder to score and rank hypotheses encoded in a word lattice.. Also, since the semantic decoder provides additional knowledge about dialogue acts, such knowledge may be passed back to the speech recognizer in order to constrain the recognition, for example, by conditioning the recognizer ’ s N-gram language model.
read more
