Automatically Mapping Natural Language Requirements to Domain-Specific Process Models

Authors

  • Uthayasanker Thayasivam University of Georgia
  • Kunal Verma Accenture
  • Alex Kass Accenture
  • Reymonrod Vasquez Accenture

DOI:

https://doi.org/10.1609/aaai.v25i2.18863

Abstract

For large scale enterprise implementations, a key problem, that has not been tackled much, is the ability to automatically map users’ requirements to reference process models. We present a tool called Process Model Requirements Gap Analyzer (ProcGap), which uses a combination of natural language processing, information retrieval and semantic reasoning to automatically match and map textual requirements to industry-specific process models. We present the results of mapping requirements from an industry project to an existing process model. We compare our approach to two previously implemented approaches and show that our approach outperforms them. In a case study, we also found that a user group with ProcGap had better performance than a user group that performed the same task manually.

Downloads

Published

2011-08-11

How to Cite

Thayasivam, U., Verma, K., Kass, A., & Vasquez, R. (2011). Automatically Mapping Natural Language Requirements to Domain-Specific Process Models. Proceedings of the AAAI Conference on Artificial Intelligence, 25(2), 1695-1700. https://doi.org/10.1609/aaai.v25i2.18863