Enhancing Software Failure Analysis through Semantic Similarity and Knowledge Graph Modeling

This repository related to the paper

Will update soon

Data Available

Download data from: Google Drive

Directories

• preprocessing: Contains data from the FAIL database and relationships identified between the data.
• model: Contains model training files.
• train: Contains data used to train the model.
• result: Contains the results of the models.
• evaluate: Contains evaluation results for each model.

Files

• Interactive_Product.ipynb: Jupyter notebook containing an interface to retrieve relevant information from the FAIL database.
• sentence_embeddings.npy: Numpy file containing pre-computed embeddings of all entries in the FAIL database.

Contact

Boyang Zhang (bryanzh123 [at] gmail.com)