Bidirectional Long Short-Term Memory for Entailment Identification in Requirement Specifications Using Information from Use Case Diagrams

Dony Bahtera Firmawan, Daniel Siahaan

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

Text entailment is a field of natural language processing research that concerned with understanding the meanings or semantics of sentences or text fragments. An entailment statement is a relationship between sentences where the truth of one sentence necessarily implies the truth of another. In requirements engineering (RE), identifying entailments between requirements is crucial for modeling the dependencies. Current entailment methods are not suitable for this purpose. This research proposes a new architecture for classifying entailments between requirements in the software specification document. The proposed architecture builds a training model using use case diagram and its description. The training model is later used for classifying entailments between requirements statements. Based on the case study, the training model can identify entailments between requirement statements.

Original languageEnglish
Title of host publication2021 International Seminar on Machine Learning, Optimization, and Data Science, ISMODE 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages331-336
Number of pages6
ISBN (Electronic)9781665405447
DOIs
Publication statusPublished - 2022
Event2021 International Seminar on Machine Learning, Optimization, and Data Science, ISMODE 2021 - Jakarta, Indonesia
Duration: 29 Jan 2022 → …

Publication series

Name2021 International Seminar on Machine Learning, Optimization, and Data Science, ISMODE 2021

Conference

Conference2021 International Seminar on Machine Learning, Optimization, and Data Science, ISMODE 2021
Country/TerritoryIndonesia
CityJakarta
Period29/01/22 → …

Keywords

  • BiLSTM
  • Software specification document
  • Text entailment
  • Use case diagram

Fingerprint

Dive into the research topics of 'Bidirectional Long Short-Term Memory for Entailment Identification in Requirement Specifications Using Information from Use Case Diagrams'. Together they form a unique fingerprint.

Cite this