https://github.com/ucasir/NPRF

NPRF: A Neural Pseudo Relevance Feedback Framework for Ad-hoc Information Retrieval
https://github.com/ucasir/NPRF

information-retrieval neural-network pseudo-relevance-feedback

Last synced: 13 days ago
JSON representation

NPRF: A Neural Pseudo Relevance Feedback Framework for Ad-hoc Information Retrieval

• Host: GitHub
• URL: https://github.com/ucasir/NPRF
• Owner: ucasir
• License: apache-2.0
• Created: 2018-08-16T06:49:45.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2018-11-26T10:44:13.000Z (almost 6 years ago)
• Last Synced: 2024-07-31T23:45:10.191Z (3 months ago)
• Topics: information-retrieval, neural-network, pseudo-relevance-feedback
• Language: Python
• Size: 2.15 MB
• Stars: 33
• Watchers: 4
• Forks: 10
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-few-shot-meta-learning - code - official (TF+Keras)

README

        
# NPRF

NPRF: A Neural Pseudo Relevance Feedback Framework for Ad-hoc Information Retrieval [[pdf]](http://aclweb.org/anthology/D18-1478)

 





If you use the code, please cite the following paper: 

```

@inproceedings{li2018nprf,

  title={NPRF: A Neural Pseudo Relevance Feedback Framework for Ad-hoc Information Retrieval},

  author={Li, Canjia and Sun, Yingfei and He, Ben and Wang, Le and Hui, Kai and Yates, Andrew and Sun, Le and Xu, Jungang},

  booktitle={Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing},

  year={2018}

}

```

## Requirement

* Tensorflow

* Keras

* gensim

* numpy

## Getting started

### Training data preparation

To capture the top-k terms from top-n documents, one needs to extract term document frequency from index. Afterwards, you are required to generate the similarity matrix upon the query and document given the pre-trained word embedding (e.g. word2vec). Related functions can be found in preprocess/prepare_d2d.py.

### Training meta data preparation

We introduce two classes for the ease of training. The class **Relevance** incorporates the relevance information from the baseline and qrels file. The class **Result** simplify the write/read operation on standard TREC result file. Other information like query idf is dumped as a pickle file.

### Model training

Configure the MODEL_config.py file, then run 

```

python MODEL.py --fold fold_number temp_file_path

```

You need to run 5-fold cross valiation, which can be automatically done by running the *runfold.sh* script. The temp file is a temporary file to write the result of the validation set in TREC format. A training log sample on the first fold of TREC 1-3 dataset is provided for reference, see [sample_log](https://github.com/ucasir/NPRF/blob/master/sample_log).

### Evaluation

After training, the evaluation result of each fold is retained in the result path as you specify in the MODEL_config.py file. One can simply run `cat *res >> merge_file` to merge results from all folds. Thereafter, run the [trec_eval](https://trec.nist.gov/trec_eval/) script to evaluate your model.

## Reference

Some snippets of the code follow the implementation of [K-NRM](https://github.com/AdeDZY/K-NRM), [MatchZoo](https://github.com/faneshion/MatchZoo).