qiujiaxin86/Details-In-Recommendation

Recommender-In-Detail is a package which offers detailed implementations of state-of-the-art techniques and basic methods in recommendation.

Details-In-Recommendation

As its name suggests,Details-In-Recommendation is a package which offers detailed implementations of state-of-the-art techniques and basic methods in recommendation. Most of them could be used in production environments through simple modifications. The structure of the package is as follows:

1. models: contains many sub-package which implements recommendation methods.
2. metrics: some metrics to evaluate recommendation.
3. dataset: dataset to used in examples.
4. examples: examples to test our implementation in models by using dataset in dataset and metrics in metrics.
5. utils: some useful tools, e.g., embedding visualization, and etc.

中文博客-Chinese Blogs.

Implemented Models

MODELS	TAG	PAPERS
DeepWalk	Embedding	DeepWalk: Online Learning of Social Representations
Node2Vec	Embedding	node2vec: Scalable Feature Learning for Networks
FaceBook GBDT+LR	Embedding/CTR	Practical lessons from predicting clicks on Ads at Facebook
Wide&Deep	Deep CTR	Wide & Deep Learning for Recommender Systems
Youtube Candidate Generation	Embedding/Deep CTR	Deep Neural Networks for YouTube Recommendations
Youtube Ranking	Deep CTR	Deep Neural Networks for YouTube Recommendations
DCN	Deep CTR	Deep & Cross Network for Ad Click Predictions
PNN	Deep CTR	Product-based Neural Networks for User Response Prediction
DeepFM	Deep CTR	DeepFM: A Factorization-Machine based Neural Network for CTR Prediction
ESMM	Deep CTR /CVR	Entire Space Multi-Task Model: An Effective Approach for Estimating Post-Click Conversion Rate
DIN	Deep CTR	Deep Interest Network for Click-Through Rate Prediction
XdeepFM	Deep CTR	xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems
NCF	Deep MF	Neural Collaborative Filtering
LFM	MF/CF	funkSVD
wALS	MF/CF	CollaborativeFiltering for Implicit Feedback Datasets
eALS	MF/CF	Fast Matrix Factorization for Online Recommendation with Implicit Feedback
BPR	MF/CF	BPR: Bayesian Personalized Ranking from Implicit Feedback
VBPR	MF/CF	VBPR: Visual Bayesian Personalized Ranking from Implicit Feedback
LSA	MF/topic model
pLSA	topic model	Collaborative topic modeling for recommending scientific articles
LDA	topic model
Item_based	CF	Item-based collaborative filtering recommendation algorithms
Personal Rank	CF	Topic-Sensitive PageRank
Item2Vec	Embedding	Item2Vec:Neural Item Embedding for Collaborative Filtering
AutoEncoder
Tompson Sampling	Bandit
UCB	Bandit
LinUCB	Bandit
Bert	NLP
Causal Embedding	MF/Cause

Metrics

METRICS	PAPERS
ndcg@K

Dataset

https://tianchi.aliyun.com/dataset/

Utils