8000 GitHub - YijunSu/IJCNN2020_FGRec: Official code for FGRec: A Fine-Grained Point-of-Interest Recommendation Framework by Capturing Intrinsic Influences (IJCNN 2020)
[go: up one dir, main page]

Skip to content

Official code for FGRec: A Fine-Grained Point-of-Interest Recommendation Framework by Capturing Intrinsic Influences (IJCNN 2020)

Notifications You must be signed in to change notification settings

YijunSu/IJCNN2020_FGRec

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FGRec

FGRec: A Fine-Grained Point-of-Interest Recommendation Framework by Capturing Intrinsic Influences (IJCNN 2020)

Performance

Details for Precision:

| Dataset    | Precision@5 | Precision@10 | Precision@20   | Precision@50 |
| ---------- | ------------| -------------| ---------------| -------------|
| Foursquare | 0.04712     | 0.03772      | 0.02885        | 0.0204       |
| Yelp       | 0.030783    | 0.0258491    | 0.0214022      | 0.015348     |
Details for Recall:

| Dataset    | Recall@5    | Recall@10    | Recall@20      | Recall@50    |
| ---------- | ------------| -------------| ---------------| -------------|
| Foursquare | 0.02867     | 0.04446      | 0.06681        | 0.115        |
| Yelp       | 0.031421    | 0.05193      | 0.085184       | 0.1479221    |
Details for MAP:

| Dataset    | MAP@5        | MAP@10       | MAP@20         | MAP@50      |
| ---------- | ------------ | -------------| ---------------| ------------|
| Foursquare | 0.01733      | 0.01987      | 0.02219        | 0.02491     |
| Yelp       | 0.0202166    | 0.021        | 0.0231568      | 0.026059    |
Details for NDCG:

| Dataset    | NDCG@5       | NDCG@10      | NDCG@20        | NDCG@50      |
| ---------- | ------------ | -------------| ---------------| -------------|
| Foursquare | 0.04894      | 0.04192      | 0.03448        | 0.02585      |
| Yelp       | 0.03161      | 0.0279923    | 0.02425        | 0.0186004    |
  • The performance of our framework on Foursquare.

The performance of our framework on Foursquare

  • The performance of our framework on Yelp.

The performance of our framework on Yelp

Requirements

  • python==3.7

Datasets

We use two real-world LBSN datasets from Foursquare and Yep.

Statistics:

| Dataset    | Number of users | Number of POIs | Number of categories | Number of check-ins    | Number of social links  | User-POI matrix density|
| ---------- | --------------- | -------------- | ---------------------| ---------------------- |-------------------------|----------------------- |
| Foursquare | 2,551           | 13,474         | 10                   | 124,933                | 32,512                  |0.291%                  |
| Yelp       | 30,887          | 30,887         | 624                  | 860,888                | 860,888                 |0.14%                   |

How to run FGRec model

1.python recommendation_Foursquare.py
2.recommendation_Yelp.py

Citation

Please cite our paper if you use the code or datasets:

@inproceedings{su2020fgrec,
  title={FGRec: A Fine-Grained Point-of-Interest Recommendation Framework by Capturing Intrinsic Influences},
  author={Yijun Su, Jia-Dong Zhang, Xiang Li,  Daren Zha, Ji Xiang, Wei Tang and Neng Gao},
  booktitle={IEEE International Joint Conference on Neural Networks, {IJCNN} 2020}, 
  pages={1-9},
  doi={10.1109/IJCNN48605.2020.9207571},
  year={2020}
}

Contact

If you have any questions, please contact us by suyijun.ucas@gmail.com, we will be happy to assist.

Last Update Date: October 1, 2021

About

Official code for FGRec: A Fine-Grained Point-of-Interest Recommendation Framework by Capturing Intrinsic Influences (IJCNN 2020)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

0