Neural-Symbolic Recommendation Model based on Logical Query

This repository includes the implementation for Neural-Symbolic Recommendation Model based on Logical Query

Example to run the codes

We provide the trained model file, and you can run the test using the following command.

> cd \src  
> sh testml100k.sh  

Due to file size constraints, we have uploaded the model files for the other two datasets to the following link, which you can use to download the files and place them in the src directory for testing

model file of 5MoviesTV:https://drive.google.com/file/d/1R5u5G_LoK4rW6qawzEUzW7GSuhR1u_u1/view?usp=share_link
model file of 5KindleStore:https://drive.google.com/file/d/1xrJTAjFimy76hVX8-YyiZpGXkGrB6OdL/view?usp=share_link

> cd \src  
> sh test5movie.sh  
> sh test5kindle.sh  

You can also train your own model and get results by modifying the parameters of train and load, we will add more details on how to use them when the paper is accepted

Base line experiments

For experiments related to baseline, please refer to：https://github.com/rutgerswiselab/NLR

##Result Since some reader may want to see more result, we provide that here:

ml100k	ndcg5	ndcg10	ndcg20	hit1	hit5	hit10	hit20
BPRMF	0.3051	0.3599	0.4085	0.1517	0.4566	0.6265	0.8184
SVDPP	0.3168	0.3778	0.4203	0.1565	0.4732	0.6624	0.8307
DMF	0.3023	0.3643	0.4084	0.1479	0.4489	0.6488	0.8124
NCF	0.3041	0.3595	0.4066	0.1504	0.4512	0.6338	0.8081
STAMP	0.3353	0.3907	0.4303	0.1747	0.4893	0.6602	0.8173
GRU4Rec	0.3536	0.4094	0.4421	0.1816	0.5129	0.6838	0.8296
NARM	0.3511	0.4084	0.4435	0.1870	0.5065	0.6833	0.8210
NLR	0.3466	0.4059	0.4323	0.1824	0.4995	0.6672	0.8135
NCR	0.3504	0.4064	0.4389	0.1886	0.5133	0.6948	0.8291
NPQA	0.3792	0.4316	0.4684	0.1983	0.5504	0.7128	0.8575
5Movies	ndcg5	ndcg10	ndcg20	hit1	hit5	hit10	hit20
BPRMF	0.4037	0.4460	0.4764	0.2515	0.5421	0.6728	0.7926
SVDPP	0.4157	0.4569	0.4869	0.2706	0.5477	0.6770	0.7971
DMF	0.4059	0.4445	0.4786	0.2503	0.5469	0.6775	0.8019
NCF	0.3943	0.4398	0.4734	0.2461	0.5242	0.6645	0.7971
STAMP	0.4047	0.4458	0.4764	0.2576	0.5385	0.6655	0.7863
GRU4Rec	0.4209	0.4616	0.4914	0.2716	0.5560	0.6818	0.7991
NARM	0.4210	0.4612	0.4909	0.2744	0.5533	0.6797	0.7987
NLR	0.4220	0.4626	0.4921	0.2728	0.5567	0.6822	0.7984
NCR	0.4374	0.4767	0.5058	0.2856	0.5728	0.6964	0.8109
NPQA	0.4654	0.5033	0.5307	0.3153	0.5997	0.7168	0.8244

5Kindle	ndcg5	ndcg10	ndcg20	hit1	hit5	hit10	hit20
BPRMF	0.4626	0.5024	0.5286	0.2973	0.6120	0.7343	0.8372
SVDPP	0.4815	0.5205	0.5470	0.3209	0.6217	0.7422	0.8442
DMF	0.4489	0.4975	0.5237	0.2972	0.5972	0.7194	0.8138
NCF	0.4568	0.5059	0.5287	0.2968	0.6038	0.7259	0.8253
STAMP	0.4517	0.4906	0.5189	0.2964	0.5923	0.7121	0.8234
GRU4Rec	0.4782	0.5138	0.5381	0.3229	0.6159	0.7253	0.8211
NARM	0.4619	0.4982	0.5239	0.3087	0.5984	0.7101	0.8112
NLR	0.4539	0.4934	0.5214	0.2968	0.6066	0.7225	0.8335
NCR	0.4708	0.5139	0.5424	0.3025	0.6218	0.7426	0.8452
NPQA	0.5538	0.5864	0.6069	0.3969	0.6908	0.7909	0.8718