hw5-JohayerChowdury

hw5-JohayerChowdury created by GitHub Classroom MSCI-541-HW5 Johayer Rahman Chowdury

Handed In: December 06, 2022

To run IndexEngine, open the terminal and run 'python IndexEngine.py' with the following arguments:

• absolute/relative path to the latimes.gz file (restricted TREC data)
• absolute/relative path to the directory to store the documents and their metadata
• (optional argument) 0 for no stemming, 1 for stemming

For example:

• python IndexEngine.py ./latimes.gz ./latimes-index

To run GetDoc, open the terminal and run 'python GetDoc.py' with the following arguments:

absolute/relative path to the directory created after running IndexEngine One of the following words: docno, id The docno or id of the document based on the word chosen in the previous argument. For example:

python GetDoc.py ./latimes-index docno LA010389-0090 python GetDoc.py ./latimes-index id 24

To run BM25, open the terminal and run 'python BM25.py' with the following operators and arguments:

• --index-path, 'Path to the directory location of your index'
• --queries-path, 'Path to the queries file'
• --output-path, 'Path to store output in a new file'
• --toStem, type=int, help='Optional argument: 0 for basic tokenization, 1 for stemming in tokenization (default is 0)'

To run GetDocBM25stem, open the terminal and run 'python GetDOCBM25stem.py' with the following operators and arguments:

1. a relative path to the BM25 stem results.
2. a relative path to the BM25 stem index.
3. a relative path to the dictionary of queries.

After running GetDocBM25stem.py, three files will be created. 3 pickled files; 1 that stores the documents' text in a list, 1 that stores the documents' docnos in a list, and 1 that stores the queries associated for that document in a list.

To run monoBERT_ranking_results.py, use the three files created after running GetDocBM25stem.py in conjunction with the ExtractQueries program that creates a pickled file that stores the dictionary of queries and topic numbers.

To run Topic Evaluation Score Calculator (TESC), open the terminal and run 'python TESC.py' with the following operators and arguments:

• --qrel Path_to_qrel
• --document-directory Path_to_directory_containing_documents
• --results Path_to_file_containing_results
• (optional) --max-results Max_number_of_results_to_iterate_for_each_query

For example:

• python TESC.py --qrel ../hw3-files/qrels/LA-only.trec8-401.450.minus416-423-437-444-447.txt --document-directory ./latimes-index --results ../hw3-files/results-files/student1.results
• python TESC.py --qrel ../hw3-files/qrels/LA-only.trec8-401.450.minus416-423-437-444-447.txt --document-directory ./latimes-index --results ../hw3-files/results-files/student1.results --max-results 500