RandolphVI/Deep-Relevance-Ranking

Deep Relevance Ranking in QA & IR

Deep Relevance Ranking

This repository contains my implementations of some models (e.g., DRMM, PACRR, ABEL-DRMM, etc.) for deep relevance ranking in QA & IR.

Requirements

• Python 3.6
• Tensorflow 1.1 +
• Numpy
• Gensim

Introduction

Data

See data format in data folder which including the data sample files.

You need to download the dataset(s) you intend to use (BioASQ and/or TREC ROBUST2004).

cd data
sh get_bioasq_data.sh
sh get_robust04_data.sh

Data Format

This repository can be used in other QA & IR datasets in two ways:

1. Modify your datasets into the same format of the sample.
2. Modify the data preprocess code in data_helpers.py.

Anyway, it should depend on what your data and task are.

Network Structure

DRMM

References:

• A deep relevance matching model for ad-hoc retrieval

---

PACRR

References:

• PACRR: A position-aware neural IR model for relevance matching

---

ABEL-DRMM

POSIT-DRMM

References:

• Deep relevance ranking using enhanced document-query interactions

---

About Me

黄威，Randolph

SCU SE Bachelor; USTC CS Ph.D.

Email: chinawolfman@hotmail.com

My Blog: randolph.pro

LinkedIn: randolph's linkedin