Skip to content

Latest commit

 

History

History
16 lines (13 loc) · 1.18 KB

README.md

File metadata and controls

16 lines (13 loc) · 1.18 KB

TAROT

This the repository for our ASE'2022 paper "Using Consensual Biterms from Text Structures of Requirements and Code to Improve IR-Based Traceability Recovery". In this paper, TAROT is proposed to improve IR-based requirement traceability recovery through using consensual biterms extracted from requirements and code, which can also be combined with other enhancing strategies to promote their performance.

The framework of TAROT

Running TAROT

  • Please run main() in Start.java.
  • You can choose different evaluated system (9 projects) and IR models (e.g., VSM, LSI, and JSD).

Extract Consensual Biterms

  • Step1: Run experiment.preprocess.parsecode.ParseCodeAST to extract identifier names from source code, and experiment.preprocess.biterm.ExtractReqBiterm to extract initial biterms from code.
  • Step2: Run methodextractIssueBiterm() and extractUcBiterm() in experiment.preprocess.biterm.ExtractReqBiterm to extract initial biterms from issue and uc respectively.
  • Step3: Run experiment.preprocess.biterm.ConsensualBiterm to get consensual biterms both shared by requirements and source code.