When Reproducibility goes Sideways: The Case of Knowledge-Enhanced Document Embeddings

Repository to contain code and information for:

Knowledge-Enhanced Document Embeddings for IR

Requirements

• ElasticSearch 6.6
• Python 3
  • Numpy
  • TensorFlow >= 1.13
  • Whoosh
  • SQLite3
  • Cvangysel
  • Pytrec_Eval
  • Scikit-Learn
  • Tqdm
  • QuickUMLS
  • Elasticsearch
  • Elasticsearch_dsl
• UMLS 2018AA

Additional Notes

server.py needs to be substitued within QuickUMLS folder as it contains a modified version required to run knowledge-enhanced models.
The folder structure required to run experiments can be seen in folder example. Python files need to be put in root.
Qrels file needs to be in .txt format.
To perform retrofitting run retrofit_doc_vecs.py, whereas to train PV-DM and cDoc2Vec models run gensim_doc2vec.py.
To run BM25, use the Jupyter Notebook file elastic_search.ipynb.
To perform query expansion run qe_combsum.py.