Hello,

this repository contains an implementation of the DRMM model.

Original paper: A Deep Relevance Matching Model for Ad-hoc Retrieval (Guo et al.) https://dl.acm.org/citation.cfm?id=2983769

Some code is taken from MatchZoo (referenced in the respective files).

Table of Contents

• Preprocessing: Contains all preprocessing code (starting from raw TREC data) to create all input files for the neural ranking model (this includes generating the histogram, idf informations). See the peprocessing Readme for more.

• Neural Ranking: Contains the neural ranking model

• Results: Contains a description of experiments & their raw and trec_eval evaluated results.

Dependencies

• TREC 8 corpus data (not preprocessed: fbis,fr94,ft,latimes)
• Python 3: Latest Keras and Tensorflow