Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,117 @@
|
|
1 |
-
---
|
2 |
-
license: apache-2.0
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
---
|
4 |
+
|
5 |
+
# LLM4POI
|
6 |
+
|
7 |
+
## Dataset Summary
|
8 |
+
|
9 |
+
A preprocessed version of [LLM4POI](https://github.com/neolifer/LLM4POI), including the FourSquare-NYC, Gowalla-CA, and FourSquare-TKY datasets. Please refer to their repository for more details.
|
10 |
+
|
11 |
+
LLM4POI frames next POI prediction task into a question-answering problem that is fed as prompt into a large language model (LLM). The model is trained to generate the next POI given the current trajectory and the historical trajectory. This repository hosts both the Q&A version and the raw txt and csv versions of the datasets.
|
12 |
+
|
13 |
+
This dataset is used to train GenUP models as described in our paper [GenUP: Generative User Profilers as In-Context Learners for Next POI Recommender Systems](https://arxiv.org/abs/2410.20643).
|
14 |
+
|
15 |
+
### Dataset Sources
|
16 |
+
|
17 |
+
Repository: [neolifer/LLM4POI](https://github.com/neolifer/LLM4POI)
|
18 |
+
|
19 |
+
Paper: [Large Language Models for Next Point-of-Interest Recommendation](https://arxiv.org/abs/2404.17591)
|
20 |
+
|
21 |
+
|
22 |
+
Repository: [w11wo/GenUP](https://github.com/w11wo/GenUP)
|
23 |
+
|
24 |
+
Paper: [GenUP: Generative User Profilers as In-Context Learners for Next POI Recommender Systems](https://arxiv.org/abs/2410.20643)
|
25 |
+
|
26 |
+
## Dataset Structure
|
27 |
+
|
28 |
+
```shell
|
29 |
+
.
|
30 |
+
βββ README.md
|
31 |
+
βββ ca
|
32 |
+
βΒ Β βββ preprocessed
|
33 |
+
βΒ Β βββ test_qa_pairs_kqt.txt
|
34 |
+
βΒ Β βββ train_qa_pairs_kqt.json
|
35 |
+
βΒ Β βββ train_sample.csv
|
36 |
+
βββ nyc
|
37 |
+
βΒ Β βββ preprocessed
|
38 |
+
βΒ Β βββ test_qa_pairs_kqt.txt
|
39 |
+
βΒ Β βββ train_qa_pairs_kqt.json
|
40 |
+
βΒ Β βββ train_sample.csv
|
41 |
+
βββ tky
|
42 |
+
βββ preprocessed
|
43 |
+
βββ test_qa_pairs_kqt.txt
|
44 |
+
βββ train_qa_pairs_kqt.json
|
45 |
+
βββ train_sample.csv
|
46 |
+
```
|
47 |
+
|
48 |
+
### Data Instances
|
49 |
+
|
50 |
+
An example of a line in `test_qa_pairs_kqt.txt`:
|
51 |
+
|
52 |
+
```plaintext
|
53 |
+
<question>: The following data is a trajectory of user 2: At 2010-09-25 01:38:14, user 2 visited POI id 247 which is a Stadium and has Category id 261. At 2010-09-25 02:11:34, ... Given the data, At 2010-09-25 20:34:27, Which POI id will user 2 visit? Note that POI id is an integer in the range from 0 to 9690.<answer>: At 2010-09-25 20:34:27, user 2 will visit POI id 6350.
|
54 |
+
```
|
55 |
+
|
56 |
+
An example of a JSON object in `train_qa_pairs_kqt.json`:
|
57 |
+
|
58 |
+
```json
|
59 |
+
{
|
60 |
+
"question": "The following data is a trajectory of user 2: At 2010-09-25 01:38:14, user 2 visited POI id 247 which is a Stadium and has Category id 261. At 2010-09-25 02:11:34, ... Given the data, At 2010-09-25 20:34:27, Which POI id will user 2 visit? Note that POI id is an integer in the range from 0 to 9690.",
|
61 |
+
"answer": "At 2010-09-25 20:34:27, user 2 will visit POI id 6350."
|
62 |
+
}
|
63 |
+
```
|
64 |
+
|
65 |
+
An example of entries in `train_sample.csv`:
|
66 |
+
|
67 |
+
```csv
|
68 |
+
check_ins_id,UTCTimeOffset,UTCTimeOffsetEpoch,pseudo_session_trajectory_id,UserId,Latitude,Longitude,PoiId,PoiCategoryId,PoiCategoryName
|
69 |
+
126094.0,2010-06-06 18:48:32,1275814112,0,1,37.6163560649,-122.3861503601,445,207,Airport
|
70 |
+
126278.0,2010-06-06 22:11:04,1275826264,0,1,37.7826046833,-122.4076080167,244,1,Coffee Shop
|
71 |
+
126314.0,2010-06-06 22:40:29,1275828029,0,1,37.7831295924,-122.4038743973,9346,121,Conference
|
72 |
+
126622.0,2010-06-07 06:01:04,1275854464,0,1,37.7815086,-122.4050282333,3253,128,Pub
|
73 |
+
```
|
74 |
+
|
75 |
+
### Data Splits
|
76 |
+
|
77 |
+
| | train | test |
|
78 |
+
| --- | ----: | ---: |
|
79 |
+
| NYC | 11022 | 1447 |
|
80 |
+
| CA | 36374 | 2864 |
|
81 |
+
| TKY | 51661 | 7079 |
|
82 |
+
|
83 |
+
## Additional Information
|
84 |
+
|
85 |
+
### Citation
|
86 |
+
|
87 |
+
If you find this repository useful for your research, please consider citing our paper:
|
88 |
+
|
89 |
+
```bibtex
|
90 |
+
@misc{wongso2024genupgenerativeuserprofilers,
|
91 |
+
title={GenUP: Generative User Profilers as In-Context Learners for Next POI Recommender Systems},
|
92 |
+
author={Wilson Wongso and Hao Xue and Flora D. Salim},
|
93 |
+
year={2024},
|
94 |
+
eprint={2410.20643},
|
95 |
+
archivePrefix={arXiv},
|
96 |
+
primaryClass={cs.IR},
|
97 |
+
url={https://arxiv.org/abs/2410.20643},
|
98 |
+
}
|
99 |
+
```
|
100 |
+
|
101 |
+
```bibtex
|
102 |
+
@inproceedings{Li_2024, series={SIGIR 2024},
|
103 |
+
title={Large Language Models for Next Point-of-Interest Recommendation},
|
104 |
+
url={http://dx.doi.org/10.1145/3626772.3657840},
|
105 |
+
DOI={10.1145/3626772.3657840},
|
106 |
+
booktitle={Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval},
|
107 |
+
publisher={ACM},
|
108 |
+
author={Li, Peibo and de Rijke, Maarten and Xue, Hao and Ao, Shuang and Song, Yang and Salim, Flora D.},
|
109 |
+
year={2024},
|
110 |
+
month=jul, pages={1463β1472},
|
111 |
+
collection={SIGIR 2024}
|
112 |
+
}
|
113 |
+
```
|
114 |
+
|
115 |
+
### Contact
|
116 |
+
|
117 |
+
If you have any questions or suggestions, feel free to contact Wilson at `w.wongso(at)unsw(dot)edu(dot)au`.
|