A Synthesis of Stochastic Petri Net (SPN) Graphs for Natural Language Understanding (NLU) Event/Action Association

被引：1

作者：

Psarologou, Adamantia ^{[1
]}

Bourbakis, Nikolaos ^{[1
]}

Esposito, Anna ^{[2
]}

机构：

[1] Wright State Univ, Coll Engn, CART, Dayton, OH 45435 USA

[2] Univ Naples 2, Dept Psychol, I-81100 Caserta, Italy

来源：

2015 IEEE 27TH INTERNATIONAL CONFERENCE ON TOOLS WITH ARTIFICIAL INTELLIGENCE (ICTAI 2015) | 2015年

关键词：

Natural Language Understanding (NLU); Sotchastic Petri Nets (SPNs); Coloured Petri Nets (CPIV); combination of graphs; Anaphora Resolution (AR); kernel extraction;

D O I：

10.1109/ICTAI.2015.43

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This paper focuses on the combination of Stochastic Petri Net (SPN) graphs for event association in the context of Natural Languages Understanding (NLU). Our general goal is to develop a new NLU methodology. In this paper we present some of its components which are: the use of Anaphora Resolution (AR), the extraction of kernel(s) based on the structure of their parse trees, and the synthesis of SPN graphs. Each extracted kernel is represented with SPN graphs. Here we define when, why and how to synthesize the produced SPN graphs for event/action association, while preserving their timing and information flow. Finally, we provide examples of combined and uncombined SPN graphs of different NL texts created using our proposed methodology.

引用

页码：218 / 225

页数：8