Datasets:

Twitter
/

TwitterFaveGraph

user_index int64 1 13.1M	tweet_index int64 1 6.76M	time_chunk int64 1 192
2,664,563	4,656,426	134
9,109,889	3,354,941	32
10,473,449	4,994,572	31
4,288,842	5,859,769	133
6,032,357	6,472,618	10
2,034,269	4,368,917	109
8,830,130	2,783,046	158
3,130,320	4,495,471	127
4,910,071	2,566,715	175
7,580,155	3,233,726	151
4,355,276	6,368,528	178
8,667,380	4,952,085	84
12,941,558	4,803,221	33
1,341,629	6,720,280	79
3,473,283	1,844,222	180
1,839,383	650,480	62
3,955,051	1,653,189	152
11,628,634	2,871,166	7
4,542,224	5,654,400	32
10,742,976	5,813,199	62
1,848,639	3,585,192	102
16,580	404,387	128
3,287,633	5,254,735	176
6,884,229	957,687	60
4,717,784	6,066,840	182
8,088,419	6,217,010	154
6,161,302	5,694,436	134
5,830,754	4,152,514	8
8,020,419	787,108	175
12,985,202	5,633,427	62
655,546	2,126,712	109
4,342,424	5,502,958	153
8,918,424	1,875,414	80
8,603,479	4,316,984	61
3,173,458	4,200,119	130
514,570	679,046	151
482,049	1,720,401	126
8,796,519	5,731,579	179
7,422,893	4,338,144	104
10,944,500	269,659	126
7,399,359	942,486	128
4,238,331	6,380,956	154
6,140,467	5,557,615	126
13,039,379	4,624,073	86
1,263,595	6,574,338	135
6,159,531	1,878,551	131
3,057,060	4,950,019	175
88,729	6,686,896	130
3,693,767	103,425	59
6,766,427	5,014,210	8
2,361,171	4,350,252	56
1,727,962	5,064,409	153
3,292,335	1,935,364	13
446,486	4,848,429	102
1,763,453	6,119,819	13
1,542,019	1,218,314	175
12,797,703	1,211,405	178
2,870,242	5,753,434	110
214,471	1,838,293	134
2,647,355	580,176	134
6,871,883	6,357,155	158
161,942	4,018,286	61
663,909	3,269,354	106
11,949,629	2,206,061	153
7,730,801	1,999,714	85
5,311,774	5,539,567	56
8,890,353	4,762,346	60
9,352,412	418,003	37
131,743	3,557,524	38
18,536	5,930,742	105
10,441,845	3,043,515	36
6,990,047	771,352	130
12,068,390	2,558,621	128
6,169,384	2,529,048	127
2,487,379	1,566,610	181
4,412,175	2,217,450	59
1,630,083	3,619,648	31
12,292,267	1,112,918	102
3,278,632	507,274	151
7,394,017	5,540,012	157
69,896	3,317,325	37
9,987,289	632,902	61
3,809,680	117,645	55
1,288,130	3,689,673	60
10,055,664	4,423,103	85
7,276,720	5,664,099	103
638,912	5,615,928	103
9,199,978	3,589,933	55
9,643,661	4,069,822	8
13,077,203	3,304,622	158
5,719,436	3,158,686	56
720,826	6,134,862	103
9,000,088	3,333,881	109
12,525,523	5,860,095	38
5,062,775	4,227,670	152
4,420,188	2,282,132	135
7,909,713	1,112,595	181
1,105,133	4,758,551	80
7,429,342	3,546,313	132
8,630,789	596,256	179

End of preview. Expand in Dataset Viewer.

MiCRO: Multi-interest Candidate Retrieval Online

This repo contains the TwitterFaveGraph dataset from our paper MiCRO: Multi-interest Candidate Retrieval Online.
[PDF] [HuggingFace Datasets]

This work is licensed under a Creative Commons Attribution 4.0 International License.

TwitterFaveGraph

TwitterFaveGraph is a bipartite directed graph of user nodes to Tweet nodes where an edge represents a "fave" engagement. Each edge is binned into predetermined time chunks which are assigned as ordinals. These ordinals are contiguous and respect time ordering. In total TwitterFaveGraph has 6.7M user nodes, 13M Tweet nodes, and 283M edges. The maximum degree for users is 100 and the minimum degree for users is 1. The maximum degree for Tweets is 280k and the minimum degree for Tweets is 5.

The data format is displayed below.

user_index	tweet_index	time_chunk
1	2	1
2	1	1
3	3	2

Citation

If you use TwitterFaveGraph in your work, please cite the following:

@article{portman2022micro,
  title={MiCRO: Multi-interest Candidate Retrieval Online},
  author={Portman, Frank and Ragain, Stephen and El-Kishky, Ahmed},
  journal={arXiv preprint arXiv:2210.16271},
  year={2022}
}

Downloads last month: 49

Size of downloaded dataset files:

2.37 GB

Size of the auto-converted Parquet files:

3.95 GB

Number of rows:

283,837,117