---
dataset_info:
  features:
  - name: source_id
    dtype: string
  - name: response_row
    dtype: int64
  - name: text
    dtype: string
  - name: label
    dtype: int64
  - name: perspective
    dtype: string
  - name: label_name
    dtype: string
  - name: parent_id
    dtype: string
  - name: augmentation
    dtype: string
  - name: is_augmented
    dtype: bool
  splits:
  - name: train
    num_bytes: 402051
    num_examples: 1352
  - name: validation
    num_bytes: 3654
    num_examples: 14
  - name: test
    num_bytes: 4148
    num_examples: 16
  download_size: 102448
  dataset_size: 409853
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
language:
- en
task_categories:
- text-classification
pretty_name: text-test
tags:
- text
- education
- 24-679
- data-augmentation
---

# 24-679 (Fall 2026): Cat vs. Fish Perspective Classification

OmarK211/text-test

Each retained sample contributes a cat-perspective text and a fish-perspective text.

## Source and task

The preparation notebook reads `test-text.csv`, retains the two specified perspective columns,
and reshapes paired answers into separate records. Labels are denoted as **0 = cat
perspective; 1 = fish perspective**.

Preparation source: [24-679 Text Data-OK notebook](https://colab.research.google.com/drive/1tTLsHs7hjhWqVgF4_CgbGAIfC6FCoIrx).
Course: 24-679, Fall 2026, Carnegie Mellon University. Repository maintainer: the account shown above.

## Fields

| Field | Meaning and modeling role |
|---|---|
| `text` | Input text sample, with line breaks retained. |
| `label` | Classification target: 0 = cat perspective; 1 = fish perspective. |
| `perspective`, `label_name` | Readable target descriptions; exclude both from model inputs. |
| `response_row` | Paired entries; grouping/provenance only. |
| `source_id`, `parent_id` | Unique example key and its original text key; never input features. |
| `augmentation`, `is_augmented` | Method identifier and synthetic-row flag; provenance only. |

## Splits and original-source counts

These counts are computed from the packaged splits for this run.

| Split | Original rows | Synthetic rows | Total rows |
|---|---:|---:|---:|
| train | 70 | 1282 | 1352 |
| validation | 14 | 0 | 14 |
| test | 16 | 0 | 16 |

Synthetic rows in training only.

| Split | Label | Rows |
|---|---:|---:|
| train | 0 | 676 |
| train | 1 | 676 |
| validation | 0 | 7 |
| validation | 1 | 7 |
| test | 0 | 8 |
| test | 1 | 8 |

Requested holdout fraction: 30%; test receives 50% of that holdout.
Small-sample rounding changes the realized proportions. The first split uses seed 20314, and the
holdout split uses seed 20315.

Split the original input keys first, then carry both texts into the same partition. Every
generated variant stays in training with its parent. The split is grouped by response row; it does
not establish that different responses came from different people. Reuse these partitions for all
model comparisons. Do not randomly re-split the augmented rows.

## Augmentation and preprocessing

Character rate in this run: **30%** of eligible words, capped at **3** words per line. The corpus contains **50** distinct lines for each perspective.

Each method independently starts from an original training line; transformations are not stacked.

- **Internal character swaps:** swap adjacent, distinct interior letters in selected words.
- **Internal character deletions:** remove one interior letter from each selected word.
- **Within-line word swap:** exchange two distinct words in one line, preserving that line's word inventory.
- **WordNet synonym replacement:** use NLTK's English POS tagger and the first matching WordNet sense to replace one eligible base-form noun, verb, adjective, or adverb with a single-word alternative. Proper nouns, inflected forms, pronouns, negation, and explicit human/AI/machine terms are protected or excluded.

The character methods apply the configured rate to eligible words of at least four letters, round
up, and respect the per-line cap; first and last letters remain intact. A line with no eligible
change is skipped. The WordNet first-sense rule is a heuristic, not contextual sense disambiguation.

| Training method | Stored rows |
|---|---:|
| `internal_character_deletions` | 350 |
| `internal_character_swaps` | 350 |
| `none` | 70 |
| `within_line_word_swap` | 347 |
| `wordnet_synonym_replacement` | 235 |

## Intended use and limitations

Use for teaching/learning data preparation, augmentation audits, and perspective classification.vDo not use these labels to
identify whether a cat is secretly a fish.

## Privacy and licensing

The prepared schema has no dedicated name, email, or timestamp field, but free text may contain
personal details or recognizable writing. Pseudonymous IDs and grouping by response row do not
establish anonymity or participant consent. Review content and applicable course permissions before
reuse. No license was specified in the repository metadata when this card was first added; this card
does not assign one.

## Load and compare

```python
from datasets import load_dataset
ds = load_dataset("OmarK211/text-test")
# Train with ds["train"], choose settings with ds["validation"], then score ds["test"].
```

Use an account with access if repository visibility changes. For reproducible comparisons, record the
dataset commit and model/environment versions. Regenerate this card with the preparation notebook after
changing the data; its counts are calculated from the actual packaged splits. The YAML schema and split
configuration are preserved from the upload.