XiaoBB commited on
Commit
aca7cb5
·
verified ·
1 Parent(s): f76a99a

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +121 -29
README.md CHANGED
@@ -1,31 +1,123 @@
1
  ---
2
- dataset_info:
3
- features:
4
- - name: seq_idx
5
- dtype: int64
6
- - name: seq_len
7
- dtype: int64
8
- - name: description
9
- dtype: string
10
- - name: metadata
11
- dtype: string
12
- - name: time_since_start
13
- list: float64
14
- - name: time_since_last_event
15
- list: float64
16
- - name: type_event
17
- list: string
18
- - name: type_text
19
- list: string
20
- splits:
21
- - name: test
22
- num_bytes: 10102134
23
- num_examples: 219
24
- download_size: 3489422
25
- dataset_size: 10102134
26
- configs:
27
- - config_name: default
28
- data_files:
29
- - split: test
30
- path: data/test-*
31
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ task_categories:
6
+ - time-series-forecasting
7
+ tags:
8
+ - temporal-point-process
9
+ - event-sequences
10
+ - github
11
+ - software-engineering
12
+ - developer-workflows
13
+ - marked-temporal-point-process
14
+ size_categories:
15
+ - n<1K
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
16
  ---
17
+
18
+ # GitHub User Event Streams
19
+
20
+ Curated event sequences from individual GitHub developers, designed for temporal point process (TPP) and marked temporal point process (MTPP) modeling. Each sequence captures a single developer's activity across multiple repositories over a one-week window.
21
+
22
+ ## Dataset Description
23
+
24
+ - **Source:** [GH Archive](https://www.gharchive.org/) raw JSON dumps
25
+ - **Time window:** June 1 - 7, 2024
26
+ - **Grouping:** Events grouped by user (actor)
27
+ - **Sequences:** ~230 (varies with filtering)
28
+ - **Sequence length:** 80-100 events per sequence
29
+ - **Language:** English only
30
+
31
+ ## Schema
32
+
33
+ Each record is a dictionary with 8 fields:
34
+
35
+ | Field | Type | Description |
36
+ |---|---|---|
37
+ | `seq_idx` | int | Sequence index |
38
+ | `seq_len` | int | Number of events in the sequence |
39
+ | `description` | str | User name, active repos (with descriptions), and event window |
40
+ | `metadata` | str (JSON) | `actor_login`, `num_repos`, `top_repos` |
41
+ | `time_since_start` | list[float] | Hours since the first event in the sequence |
42
+ | `time_since_last_event` | list[float] | Hours since the previous event |
43
+ | `type_event` | list[str] | Event type labels (see below) |
44
+ | `type_text` | list[str] | Natural language description of each event |
45
+
46
+ ## Event Types (8 categories)
47
+
48
+ | Label | Description |
49
+ |---|---|
50
+ | `issue_opened` | New issue filed |
51
+ | `issue_closed` | Issue resolved/closed |
52
+ | `pr_opened` | Pull request opened |
53
+ | `pr_merged` | Pull request merged |
54
+ | `push` | Code pushed (with commit messages) |
55
+ | `release` | Release published |
56
+ | `pr_reviewed` | PR approved or changes requested |
57
+ | `comment` | Comment on issue, PR, or review |
58
+
59
+ ## Curation Filters
60
+
61
+ Sequences are selected to represent moderately active developers working across a focused set of repositories:
62
+
63
+ | Filter | Value | Description |
64
+ |---|---|---|
65
+ | `min-raw-events` | 50 | Min kept-type events in archive (pre-classification) |
66
+ | `max-raw-events` | 300 | Max kept-type events in archive (pre-classification) |
67
+ | `min-events` | 80 | Min classified events per sequence |
68
+ | `max-events` | 100 | Max classified events (skip, not truncate) |
69
+ | `min-unique-types` | 5 | At least 5 distinct event types |
70
+ | `min-repos` | 2 | Active in at least 2 repos |
71
+ | `max-repos` | 5 | Active in at most 5 repos |
72
+ | `min-avg-text-len` | 150 | Min average text length per event |
73
+ | `max-type-fraction` | 0.5 | No single event type >50% of events |
74
+ | `require-english` | true | All event texts must be in English |
75
+
76
+ Additional processing:
77
+ - Bot users are excluded (detected by name patterns like `[bot]`, `dependabot`, etc.)
78
+ - Consecutive duplicate events (same type + text) are deduplicated
79
+
80
+ ## Key Difference from Repository Events
81
+
82
+ | | User Events | Repo Events |
83
+ |---|---|---|
84
+ | Grouping | By developer | By repository |
85
+ | Cross-repo | Yes (2-5 repos) | No (single repo) |
86
+ | Text perspective | Mentions repository name | Mentions actor name |
87
+ | Use case | Developer workflow modeling | Project activity modeling |
88
+
89
+ ## Example
90
+
91
+ ```json
92
+ {
93
+ "seq_idx": 0,
94
+ "seq_len": 90,
95
+ "description": "GitHub user alice: active in org/repo-a (A web framework for Rust), org/repo-b Event window: June 02 - 07, 2024.",
96
+ "metadata": "{\"actor_login\": \"alice\", \"num_repos\": 2, \"top_repos\": [\"org/repo-a\", \"org/repo-b\"]}",
97
+ "time_since_start": [0.0, 0.004, ...],
98
+ "time_since_last_event": [0.0, 0.004, ...],
99
+ "type_event": ["pr_reviewed", "pr_merged", "push", ...],
100
+ "type_text": ["PR approved in org/repo-a: Add new feature...", ...]
101
+ }
102
+ ```
103
+
104
+ ## Intended Use
105
+
106
+ - Training and evaluating temporal point process models
107
+ - Studying individual developer workflow patterns across repos
108
+ - Benchmarking next-event prediction and event forecasting
109
+ - Modeling marked temporal point processes with rich text marks
110
+
111
+ ## Citation
112
+
113
+ If you use this dataset, please cite:
114
+
115
+ ```bibtex
116
+ @dataset{github_user_events_2024,
117
+ title={GitHub User Event Streams},
118
+ author={XiaoBB},
119
+ year={2024},
120
+ url={https://huggingface.co/datasets/XiaoBB/github_user_events},
121
+ note={Curated from GH Archive, June 2024}
122
+ }
123
+ ```