File size: 1,569 Bytes
aab72cc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
---
license: other
language:
  - en
tags:
  - privacy
  - pii
  - token-classification
  - named-entity-recognition
  - synthetic-data
  - data-masking
task_categories:
  - token-classification
pretty_name: Maskara Extensive PII
size_categories:
  - 10K<n<100K
---

# Maskara Extensive PII

This dataset is the larger training corpus used for the real Maskara
Transformers token-classification model.

It contains 75,000 examples:

- `train.jsonl`: 67,500 examples
- `validation.jsonl`: 3,750 examples
- `test.jsonl`: 3,750 examples

## Sources

- Maskara synthetic v2 generator
- Public examples normalized from `ai4privacy/pii-masking-300k`

The dataset includes chat prompts, email drafts, support tickets, delivery
messages, identity prompts, finance prompts, code/log snippets, JSON snippets,
calendar messages, map/travel prompts, and hard negatives.

## Schema

```json
{
  "id": "maskara-extensive-0000001",
  "text": "Email Maya Rao at maya.rao@example.org.",
  "entities": [
    {"start": 6, "end": 14, "label": "PERSON_NAME", "text": "Maya Rao"},
    {"start": 18, "end": 38, "label": "EMAIL", "text": "maya.rao@example.org"}
  ],
  "domain": "email_draft",
  "locale": "en-US",
  "difficulty": "mixed",
  "source": "maskara_synthetic_v2"
}
```

## Labels

- `ADDRESS`
- `API_KEY`
- `CREDIT_CARD`
- `DATE_OF_BIRTH`
- `DRIVER_LICENSE`
- `EMAIL`
- `IP_ADDRESS`
- `LOCATION`
- `PASSWORD`
- `PERSON_NAME`
- `PHONE`
- `SSN`
- `USERNAME`

## Notes

This is not real user vault data. Maskara should never train on real local
vault mappings or private user prompts.