# Keras-CNN-QA

**Repository Path**: greitzmann/Keras-CNN-QA

## Basic Information

- **Project Name**: Keras-CNN-QA
- **Description**: Keras (re)implementation of paper  "Learning to Rank Short Text Pairs with Convolutional Deep Neural Networks. SIGIR, 2015"
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2020-10-30
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Intro
This code is remplementation of  Learning to Rank Short Text Pairs with Convolutional Deep Neural Networks. SIGIR, 2015 in Keras. 

This code is adapted from repo. https://github.com/aseveryn/deep-qa.

# Depdendencies
- python 2.7+
- numpy
- theano/tensorflow
- keras

# Embeddings 

The pre-initialized word2vec embeddings have to be downloaded from [here](https://drive.google.com/folderview?id=0B-yipfgecoSBfkZlY2FFWEpDR3M4Qkw5U055MWJrenE5MTBFVXlpRnd0QjZaMDQxejh1cWs&usp=sharing).


# Steps to run
To run the model, first run parsing file
>$ python parse.py

Then run 
>$ python ltr_cnn.py

# TO-DO
Currently support for external features (overlapping words from paper) is not supported. 

If anyone is interested, let me know, or you are most welcome to send a PR.