Structure

Data

By following Twitter data sharing recommendation and policy, check here https://developer.twitter.com/en/developer-terms/more-on-restricted-use-cases, we can only share the tweet ids.

Please check the corresponding tweet by using Twitter API. One simple way is to append the tweet_id after https://twitter.com/i/web/status/, you will be able to see the corresponding tweet.

Our data is under ./public_data.

Take ./public_data/absl.txt as an example: as the following image shows, the first column is the tweet id, and the second column is the label: 'Text' indicates it is not library-related; 'Lib' indicates it is a library-related.

Pre-trained Transformer Models

Install the required packages

$ pip install -r requirements.txt

Go to the ./src folder

Preprocess the data first

$ python preprocessing.py

within-library setting

python BERT_within_library.py --variant BERT --lib_name boto --seed 42

cross-library setting

python BERT_cross.py --variant BERT --seed 42

mixed setting

python BERT_stratified.py --variant BERT --seed 42

For mixed- and cross- settings, the performance will be directly logged inside the log files. For within-library setting, in order to get the performance across five libraries, please run the script ./src/cal_within_result.py.

Note: we stated we used Sigmoid activation, and we implemented it with Softmax. In the binary classification setting, softmax is identical to sigmoid. Reference

Baselines

Strawman

The script is also inside ./src.

python strawman.py

Prasetyo et al.

The script is also inside ./src.

python Prasetyo-approach --variant cross

APIReal

Please refer to ./APIReal/README.md for more details. You can find the original repo here.

After getting the predictions, please run the script inside ./src to calculate precision, recall and F1.

python cal_apireal_result.py

Acknowledgement

We appreciate APIReal authors made their replication package public.

Citation

If you find this repo useful, please consider to cite our work.

@inproceedings{zhang2022benchmarking,
  author = {Zhang, Ting and Chandrasekaran, Divya Prabha and Thung, Ferdian and Lo, David},
  title = {Benchmarking Library Recognition in Tweets},
  year = {2022},
  isbn = {9781450392983},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  url = {https://doi.org/10.1145/3524610.3527916},
  doi = {10.1145/3524610.3527916},
  booktitle = {Proceedings of the 30th IEEE/ACM International Conference on Program Comprehension},
  pages = {343–353},
  numpages = {11},
  keywords = {disambiguation, software libraries, tweets, benchmark study},
  location = {Virtual Event},
  series = {ICPC '22}
}

soarsmu/Library-Recognition-In-Tweets