This is our official implementation for the paper:

NAIS: Neural Attentive Item Similarity Model for Recommendation Xiangnan He, Zhankui He, Jingkuan Song, Zhenguang Liu, Yu-Gang Jiang, & Tat-Seng Chua IEEE Transactions on Knowledge and Data Engineering (TKDE 2018)

Two collaborative filtering models: NAIS_concat and NAIS_prod. To target the models for implicit feedback and ranking task, we optimize them using log loss with negative sampling.

Also, we implement the baseline: FISM, which is the well-known item-based recommendation model.

Please cite our paper if you use our codes. Thanks!

Corresponding Author: Dr. Xiangnan He (http://www.comp.nus.edu.sg/~xiangnan/)

Run NAIS_prod:

python NAIS.py --dataset pinterest-20 --pretrain 0 --weight_size 16 --embed_size 16 --data_alpha 0 --regs [0,0,1e-6] --alpha 0--beta 0.5 --lr 0.05 --algorithm 0

Run NAIS_concat:

python NAIS.py --dataset pinterest-20 --pretrain 0 --weight_size 16 --embed_size 16 --data_alpha 0 --regs [0,0,1e-6] --alpha 0--beta 0.5 --lr 0.05 --algorithm 1

Run FISM:

python FISM.py --dataset pinterest-20 --pretrain 0 --embed_size 16 --alpha 0 --lr 0.01

For more argument details, you can use python FISM.py -h and python NAIS.py -h to obtain them.

Python 2.7

TensorFlow >= r1.0

Numpy >= 1.12

PS. For your reference, our server environment is Intel Xeon CPU E5-2630 @ 2.20 GHz and 64 GiB memory. We recommend your free memory is more than 16 GiB to reproduce our experiments (and we are still trying to reduce the memory cost...).

We provide two processed datasets: MovieLens 1 Million (ml-1m) and Pinterest (pinterest-20) in Data/

train.rating:

• Train file.
• Each Line is a training instance: userID\t itemID\t rating\t timestamp (if have)

test.rating:

• Test file (positive instances).
• Each Line is a testing instance: userID\t itemID\t rating\t timestamp (if have)

test.negative

• Test file (negative instances).
• Each line corresponds to the line of test.rating, containing 99 negative samples.
• Each line is in the format: (userID,itemID)\t negativeItemID1\t negativeItemID2 ...

Update: February 5, 2018