Title: Algorithmic Improvements in Natural Language Parsing within Dialogue Systems: Priority Patterns and Wildcards
Abstract:Nowadays, human-machine interaction is changing dramatically toward spoken dialogue. One of the subtasks performed by a dialogue system (DS) is the natural language (NL) parsing in order to extract me...Nowadays, human-machine interaction is changing dramatically toward spoken dialogue. One of the subtasks performed by a dialogue system (DS) is the natural language (NL) parsing in order to extract meaning from spoken user input. The most regular tool used for NL parsing is a context-free (CF) grammar. Jay Earley’s parser is a well-known algorithm for parsing in an arbitrary CF grammar. In this paper, an improved variant of Earley’s parser is proposed in order to solve two essential problems that belong to NL parsing and arise from dialogue management. One of them is the use of partially-specified patterns, i.e., the use of wildcards in the right-handsides of the rewriting rules. The second one is the use of priority patterns, i.e., the assignment of priorities to rewriting rules. These problems mostly are not handled in the state-of-the-art DSs. The proposed parser has been implemented in a DS core application called the CAML Core, that is used to implement DSs in several domains like conversational and multi-modal interfaces for help desk applications, and for chat bots.Read More
Publication Year: 2004
Publication Date: 2004-01-01
Language: en
Type: article
Access and Citation
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot