Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
room_idx
int64
0
175k
room_label
int64
0
3
date
int64
20.3M
20.3M
0
0
20,250,602
1
0
20,250,531
2
0
20,250,528
3
0
20,250,602
4
0
20,250,601
5
1
20,250,521
6
1
20,250,524
7
0
20,250,524
8
3
20,250,520
9
0
20,250,531
10
0
20,250,531
11
0
20,250,531
12
0
20,250,531
13
0
20,250,531
14
0
20,250,531
15
0
20,250,528
16
0
20,250,603
17
0
20,250,531
18
0
20,250,531
19
0
20,250,531
20
0
20,250,531
21
0
20,250,531
22
0
20,250,531
23
0
20,250,531
24
0
20,250,531
25
0
20,250,531
26
3
20,250,531
27
0
20,250,528
28
2
20,250,530
29
0
20,250,531
30
0
20,250,531
31
3
20,250,527
32
0
20,250,528
33
3
20,250,531
34
0
20,250,528
35
0
20,250,521
36
0
20,250,531
37
0
20,250,531
38
0
20,250,531
39
0
20,250,531
40
0
20,250,531
41
1
20,250,522
42
0
20,250,531
43
3
20,250,531
44
0
20,250,603
45
0
20,250,522
46
2
20,250,603
47
0
20,250,530
48
2
20,250,529
49
0
20,250,603
50
0
20,250,530
51
0
20,250,520
52
0
20,250,531
53
0
20,250,531
54
2
20,250,529
55
0
20,250,525
56
0
20,250,531
57
0
20,250,531
58
0
20,250,521
59
0
20,250,522
60
0
20,250,521
61
0
20,250,522
62
0
20,250,522
63
0
20,250,522
64
0
20,250,525
65
0
20,250,530
66
0
20,250,531
67
0
20,250,602
68
0
20,250,525
69
0
20,250,520
70
0
20,250,530
71
0
20,250,602
72
0
20,250,523
73
0
20,250,525
74
0
20,250,522
75
0
20,250,523
76
0
20,250,530
77
0
20,250,522
78
0
20,250,531
79
0
20,250,521
80
0
20,250,522
81
1
20,250,529
82
0
20,250,524
83
0
20,250,523
84
0
20,250,521
85
0
20,250,526
86
0
20,250,521
87
0
20,250,523
88
0
20,250,531
89
2
20,250,603
90
2
20,250,524
91
0
20,250,601
92
0
20,250,521
93
2
20,250,528
94
0
20,250,521
95
0
20,250,525
96
0
20,250,523
97
3
20,250,521
98
0
20,250,523
99
0
20,250,601
End of preview. Expand in Data Studio

Live or Lie — Live Streaming Room Risk Assessment (May/June 2025)

This dataset contains live-streaming room interaction logs for room-level risk assessment under weak supervision. It is the official dataset for the research presented in the papers:

  • Outsmarting the Chameleon: Counterfactual Decoupling for Tactical OOD Shifts in Live Streaming Risk Assessment (Hugging Face Papers)
  • Live or Lie: Action-Aware Capsule Multiple Instance Learning for Risk Assessment in Live Streaming Platforms (arXiv)

Project Page: https://qiaoyran.github.io/LiveStreamingRiskAssessment/ GitHub: https://github.com/bytedance/AC-MIL

Dataset Summary

Each example corresponds to a single live-streaming room and is labeled as risky (> 0) or normal (= 0).

The task is designed for early detection: each room’s action sequence is truncated to the first 30 minutes, and can be structured into user–timeslot capsules for models such as AC-MIL or LPCD.

File Structure

The dataset is organized into two time-indexed subsets (May and June). Large LMDB data files are provided in multiple .part chunks to comply with storage limits.

.
├── final_May_hard1_masked_encoded.lmdb/
│   ├── data.mdb.00.part
│   ├── data.mdb.01.part
│   ├── data.mdb.02.part
│   ├── data.mdb.03.part
│   └── lock.mdb
├── final_June_hard1_masked_encoded.lmdb/
│   ├── data.mdb.00.part
│   ├── data.mdb.01.part
│   └── lock.mdb
├── May_train.csv
├── May_val.csv
├── May_test.csv
├── June_train.csv
├── June_val.csv
└── June_test.csv

Languages

  • Predominantly Chinese (zh): user behaviors are presented in Chinese, e.g., "主播口播:...". These action descriptions are encoded as action vectors via a Chinese-BERT model.

Data Structure

Each room has a label and a sequence of actions:

  • room_id (string)
  • label (int32, {0,1,2,3}))
  • patch_list (list of tuples):
    • u_idx (string): user identifier within a room
    • t (int32): time index along the room timeline
    • l (int32): capsule index
    • action_id (int32): action type ID
    • action_vec (list<float16> or null): action features encoded from masked action descriptions
    • timestamp (string): action timestamp
    • action_desc (string): textual action descriptions
    • user_id (string): user identifier across rooms

Action Space

The setup includes both viewer interactions (e.g., room entry, comments, likes, gifts, shares, etc.) and streamer activities (e.g., start stream, speech transcripts via voice-to-text, OCR-based visual content monitoring). Text-like fields are discretized as part of platform inspection/sampling.

Data Splits

Split #Rooms Avg. actions Avg. users Avg. time (min)
May train 176,354 709 35 30.0
May val 23,859 704 36 29.6
May test 22,804 740 37 29.7
June train 80,472 700 36 30.0
June val 10,934 767 40 29.1
June test 11,116 725 37 29.1

Quickstart

1. Reconstruct the LMDB files

Before loading the data, you must merge the split parts back into a single data.mdb file for each subset.

# Reconstruct May Dataset
cd final_May_hard1_masked_encoded.lmdb
cat data.mdb.*.part > data.mdb
cd ..

# Reconstruct June Dataset
cd final_June_hard1_masked_encoded.lmdb
cat data.mdb.*.part > data.mdb
cd ..

2. Loading Data

Install the Python package: pip install lmdb

import lmdb
import pickle

# Example: read a specific room
room_id = 0 
lmdb_path = "final_May_hard1_masked_encoded.lmdb"

env = lmdb.open(
    lmdb_path,
    readonly=True,
    lock=False,
    map_size=240 * 1024 * 1024 * 1024,
    readahead=False,
)

with env.begin() as txn:
    value = txn.get(str(room_id).encode())
    if value is not None:
        data = pickle.loads(value)
        patch_list = data["patch_list"] # list of tuples
        room_label = data["label"]

env.close()

Security and Privacy

To ensure the security and privacy of users, all data collected from live rooms has been anonymized and masked, preventing any content from being linked to a specific individual. In addition, action vectors are re-encoded from the masked action descriptions.

Considerations for Using the Data

Intended Use

  • Research on weakly-supervised risk detection / MIL in live streaming.
  • Early-warning room-level moderation signals.
  • Interpretability over localized behavior segments.

Out-of-scope / Misuse

  • Do not use this dataset to identify, profile, or target individuals.
  • Do not treat predictions as definitive enforcement decisions without human review.

Bias and Limitations

  • Sampling bias: negatives are downsampled (1:10); reported metrics and thresholds should account for this.
  • Domain specificity: behavior patterns are platform- and policy-specific; transfer to other platforms may be limited.

License

This dataset is licensed under CC BY 4.0.

Citation

@article{qiao2026live,
  title={Live or Lie: Action-Aware Capsule Multiple Instance Learning for Risk Assessment in Live Streaming Platforms},
  author={Qiao, Yiran and Chen, Jing and Ao, Xiang and Zhong, Qiwei and Liu, Yang and He, Qing},
  journal={arXiv preprint arXiv:2602.03520},
  year={2026}
}
Downloads last month
156

Papers for ByteDance/LiveStreamingRiskControl