Behavior Sequence Transformer for E-commerce Recommendation in Alibaba

Code

• github.com/shenweichen/DeepCTR
  tf★ 8,006
• github.com/alibaba/easyrec
  tf★ 2,296
• github.com/xue-pai/FuxiCTR
  pytorch★ 1,379
• github.com/UlionTse/mlgb
  pytorch★ 1,049
• github.com/massquantity/LibRecommender
  tf★ 474
• github.com/jiwidi/Behavior-Sequence-Transformer-Pytorch
  pytorch★ 176
• github.com/tangxyw/RecAlgorithm
  tf★ 0
• github.com/PaddlePaddle/PaddleRec/tree/master/models/rank/bst
  paddle★ 0
• github.com/nihalsangeeth/behaviour-seq-transformer
  pytorch★ 0

Abstract

Deep learning based methods have been widely used in industrial recommendation systems (RSs). Previous works adopt an Embedding&MLP paradigm: raw features are embedded into low-dimensional vectors, which are then fed on to MLP for final recommendations. However, most of these works just concatenate different features, ignoring the sequential nature of users' behaviors. In this paper, we propose to use the powerful Transformer model to capture the sequential signals underlying users' behavior sequences for recommendation in Alibaba. Experimental results demonstrate the superiority of the proposed model, which is then deployed online at Taobao and obtain significant improvements in online Click-Through-Rate (CTR) comparing to two baselines.

Tasks

Benchmark Results

Reproductions