A document indexer, consisting of a corpus reader, document processor, tokenizer, a weighted (tf-idf) indexer and a ranked retrieval method

In order to install the required dependencies, run the following command.

pip install -r requirements.txt

To run the indexer execute the following command. If no filepath is provided the indexer will try to open the file 'datasets/metadata_2020-03-27.csv'

cd assignment
python3 indexer.py [filepath]

Two files will be generated: outputs/bmc_weights.csv and outputs/tf_idf_weights.csv. These files will be loaded by the ranking entities

To run the vector space ranking execute the following command. If no input_filepath is provided, then the weights will be loaded from 'outputs/tf_idf_weights.csv'

python3 vector_space_ranking.py [input_filepath]

The results will be generated to 'outputs/vector_space_results.csv'

To run the bm25 ranking execute the following command. If no input_filepath is providedIf no input_filepath is provided, then the weights will be loaded from 'outputs/bmc_weights.csv'

python3 bmc_ranking.py [input_filepath]

The results will be generated to 'outputs/bmc_results.csv'