README.md

---
license: apache-2.0
language:
- en
- multilingual
task_categories:
- token-classification
tags:
- ner
- social-media
- username-extraction
- handle-extraction
- twitter
- bios
- gliner
pretty_name: HandleAtlas Benchmark
size_categories:
- n<1K
configs:
- config_name: default
  data_files:
  - split: test
    path: test.jsonl
---

# HandleAtlas Benchmark

Hand-labeled NER **evaluation set** for extracting social-media handles from
Twitter / X bios. These are the **exact 100 records** (seed = 123) used to
compute the benchmark numbers in the [LumeData/HandleAtlas-166m](https://huggingface.co/LumeData/HandleAtlas-166m)
and [LumeData/HandleAtlas-166m-CPU](https://huggingface.co/LumeData/HandleAtlas-166m-CPU)
model cards.

## Schema

Each record:

```json
{
  "id": 2,
  "text": "🍑 Ig | pea_arunya",
  "entities": [
    {"start": 7, "end": 17, "label": "instagram_username"}
  ]
}
```

- `text` — the raw bio (UTF-8, may contain emojis, multi-language text, line breaks).
- `entities` — list of NER spans. `start` and `end` are **Python character indices**
  (code points, not UTF-16 code units) into `text`. `end` is exclusive — `text[start:end]`
  gives the surface form of the entity.
- Multiple labels on the same span are encoded as separate entries with identical
  `start`/`end` (e.g. one handle written both as "ig + sc" gets two entries).

## Labels

- `instagram_username`
- `snapchat_username`
- `youtube_username`
- `twitch_username`
- `tiktok_username`
- `discord_username`
- `x_username`
- `cashapp_username`
- `onlyfans_username`
- `tumblr_username`
- `github_username`
- `kofi_username`
- `patreon_username`
- `roblox_username`
- `generic_username`

`generic_username` is a fallback for handle-shaped strings without a clear
platform indicator (e.g. `@somebody on twitter` where the user is the writer's
non-Twitter alt).

## Size

| split | records | spans |
|-------|--------:|------:|
| test  | 100 | 137 |

## Reproducing the benchmark numbers

Reference scores on this split:

| model                          | span-only F1 | span+label F1 |
|--------------------------------|-------------:|--------------:|
| HandleAtlas-166m / -CPU        |  0.955       |  0.887        |
| GLiNER-small v2.1 (zero-shot)  |  0.061       |  0.031        |
| OpenAI Privacy Filter (1.5B)   |  0.402       |  n/a (different labels) |

## Usage

```python
from datasets import load_dataset
ds = load_dataset("LumeData/HandleAtlas-benchmark", split="test")
print(ds[0])
# {"id": ..., "text": "🍑 Ig | pea_arunya",
#  "entities": [{"start": 7, "end": 17, "label": "instagram_username"}]}
```

## Provenance

Bios were sampled from a public Twitter/X user-description CSV and labeled
by hand using a single annotator. Each bio was either fully labeled
(at least one platform handle present) or skipped. Skipped records are
**not** included — every record in this dataset has at least one entity.

## Limitations & ethics

- All bios are public Twitter/X profile descriptions but contain handles
  that point to real accounts. Treat the dataset accordingly when training
  downstream models.
- Single-annotator dataset; no inter-annotator agreement was measured.
- Skewed toward the most common platforms (instagram, snapchat). Rare-label
  performance should be validated against your own data before deploying.