readme.txt

-------------------------------------------------------------
buildindex.py : Build inverted index
Description:
Take corpus (eg. dev.docs), then output a index file in format
where each line follows

word doc_id tf

example:

informaiton MED-3140 10

meaning word information has tf 10 in document 3140.

To run buildindex.py
`python3 buildindex.py <inputfile> <indexfile>`
---------------------------------------------------------------
vsm.py : compute similarity between query and doc.
Description:
Take a index file and query file (e.g. dev.title.queries) and parameter file
and output a retrieval result file.

parameter file should be in following format

parameter_name        options

weight_func        tf, tf_log, tf_norm

documents_returns      `<int>`

.....                   .....

retrival result file should be in following format
query-id document-id score

-----------------------------------------------------------------
>>> This, we can probably use galago eval function instead<<<
eval.py : evaluate performance of the model
Description:
Takes a retrieval result file and judgement file (e.g. qrel), and output a evaluation file

evaluation file format

query-id ap

....     ...

....     ...

map ...