Hybrid natural language processing in a customer-care environment

From FSL
Jump to: navigation, search

Hybrid natural language processing in a customer-care environment
David Reitter, Stefan Covaci, Florin Oltean, Catalin Bacanu and Traian Florin Serbanuta
TaCoS'01. 2001.
Abstract. CyMON is an industrial platform for one-to-one customer care applications. It compromises facilities to implement user-friendly websites with an emotional, learning interface. The CyMON Natural Language Processing Engine can understand written input and react accordingly. It is based on hybrid linguistic and statistical algorithms to analyze natural language input. Each sentence is assigned an appropriate contextualized meaning representation. This paper describes the results of an evaluation of basic pattern matching techniques and argues in favor of their augmentation with more sophisticated models taken from linguistics and statistics. The natural language processing techniques applied in CyMON are described in detail. An overview of related work is provided.
PDF, citeseer, TaCoS, BIB

Personal tools