/DBRD

Primary LanguageJupyter Notebook

Duplicate Bug Report Detection (DBRD) using Siamese Neural Networks with Spatio-Temporal Locality

This repository contains the code and data used in our research on improving Duplicate Bug Report Detection (DBRD) using Siamese Neural Networks with Spatio-Temporal Locality.

Data

We utilized the dataset provided by Lazar et al., which can be found here. It includes bug reports from three open-source projects: OpenOffice, Eclipse, and NetBeans. The processed versions of the datasets for training are stored in .rar format. Simply download and extract them to use.

  • deal_datasetexample.ipynb: An example notebook showcasing the process of handling the datasets.
  • component.ipynb: A notebook demonstrating the analysis of certain attributes.

Models

  • DBRD.ipynb: This notebook presents the network model structure used in our research.
  • embedding.ipynb and tokenize.ipynb: These notebooks show how to obtain corresponding embedding vectors through BERT pre-processing.
  • normalization.ipynb: A notebook illustrating the process of handling penalty terms.

Baseline Models

  • bert-mlp.py and dc-cnn.py: These files contain the model structures for the baseline models.

Data Preparation and Testing

  • dataprepara-dccnn.py: A script showing how to prepare the dataset for input into the DC-CNN model.
  • test.ipynb: A notebook for testing the baseline models with the prepared dataset.