https://github.com/xiangwang1223/tree_enhanced_embedding_model

TEM: Tree-enhanced Embedding Model for Explainable Recommendation, WWW2018
https://github.com/xiangwang1223/tree_enhanced_embedding_model

attention-mechanism decision-trees explainable-recommendations www2018 xgboost

Last synced: about 11 hours ago
JSON representation

TEM: Tree-enhanced Embedding Model for Explainable Recommendation, WWW2018

• Host: GitHub
• URL: https://github.com/xiangwang1223/tree_enhanced_embedding_model
• Owner: xiangwang1223
• Created: 2018-02-12T11:16:15.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2019-05-02T02:59:48.000Z (over 5 years ago)
• Last Synced: 2024-08-03T17:11:07.926Z (3 months ago)
• Topics: attention-mechanism, decision-trees, explainable-recommendations, www2018, xgboost
• Homepage:
• Size: 32.3 MB
• Stars: 73
• Watchers: 3
• Forks: 15
• Open Issues: 4
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-drug-discovery - [Python Reference

README

        
# Tree-enhanced Embedding Model

This is our project for the paper:

>Xiang Wang, Xiangnan He, Fuli Feng, Liqiang Nie and Tat-Seng Chua (2018). [TEM: Tree-enhanced Embedding Model for Explainable Recommendation](https://dl.acm.org/citation.cfm?id=3178876.3186066). In WWW'18, Lyon, France, April 23–27, 2018.

Author: Dr. Xiang Wang (xiangwang at u.nus.edu)

## Introduction

Tree-enhanced Embedding Mode (TEM) is a new recommendation framework, which combines the strong representation ability of embeddingbased and interpretability of tree-based models. At its core is an easy-to-interpret decision-tree and attention network, making the recommendation process fully transparent and explainable.

## Citation 

If you want to use our codes and datasets in your research, please cite:

```

@inproceedings{TEM2018,

  author    = {Xiang Wang and

               Xiangnan He and

               Fuli Feng and

               Liqiang Nie and

               Tat{-}Seng Chua},

  title     = {{TEM:} Tree-enhanced Embedding Model for Explainable Recommendation},

  booktitle = {{WWW}},

  pages     = {1543--1552},

  year      = {2018},

}

```

## Codes

We are finding license suitable to release this software. Currently codes are under request and will be released later.

## Dataset

We provide two rich-attribute datasets: London-Attractions (LON-A) and New-York-City-Restaurant (NYC-R) datasets, which have user profiles and item attributes, and are collected from [TripAdvisor](https://www.tripadvisor.com.sg/).

* `London_Attractions_Complete_Review.csv`

  * All positive instances.

  * Each line is a review, where the fields of user profiles and item attributes start with 'u' and 'i', respectively.

* `New_York_City_Restaurant_Complete_Review.csv`

  * All positive instances.

  * Each line is a review, where the fields of user profiles and item attributes start with 'u' and 'i', respectively.