/Best-Match-Models

This is Information Retrieval HW2

Primary LanguagePython

Introduction

  • This is the Information Retrieval HW2
  • Using TF-IDF to compute the relation between given querys and documents

Usage

code.py [-h] [-B B] [-K1 K1] [-K3 K3] [-use_q_tf {T,F}]

optional arguments:
  -h, --help       show this help message and exit
  -B B             (default: 0.75)
  -K1 K1           (default: 3.5)
  -K3 K3           (default: 1000)
  -use_q_tf {T,F}  whether to use query's tf; will deactivate K3 (default: F)

Approach

  • IDF different from original formula
    • Square IDF make it more important
  • not use query's TF term that in formula (having 's term), i.e.