Create README.md

# YouTube Alignment: Recommendation Title Embeddings

Dataset accompanying the paper **"Evaluating feedback mechanisms for aligning YouTube recommendations with user interests"** (Cho, Hale, Zhao, Shadbolt & Przybylski; University of Oxford).

## Dataset Summary

This dataset contains YouTube homepage recommendation titles and their semantic embeddings, collected from 12 simulated adolescent puzzle-gamer accounts across four experimental personas. It supports a controlled output-based audit of how YouTube's explicit feedback controls ('like' and 'not-interested') shift recommendation content relative to implicit engagement alone.

## Experimental Design

A 2×2 between-subjects design crossed two explicit feedback conditions:

| Persona | Like watched videos | Not-interested on non-puzzle gaming |
|---|---|---|
| **Watch** (control) | No | No |
| **Like** | Yes | No |
| **Not-Interested** | No | Yes |
| **Combined** | Yes | Yes |

Three independent accounts were created per persona (12 accounts total). Accounts were modelled on a 13-year-old UK adolescent puzzle-gamer archetype (DoB: 12 March 2012; gender: Rather not say). All accounts were trained in synchrony over 7 days on a randomised playlist of 70 puzzle-gaming videos (10 videos/day). Homepage recommendations were scraped hourly during a 5-day maintenance phase (days 3–7). The 2-day warm-up period was excluded from analysis.

Device and software were fixed across accounts (MacBook Pro 13", macOS Ventura 13.0.1, Chrome 134.0.6998.89; single UK geolocation) to minimise confounds.

## Data Collection

Recommendation titles were collected using an automated scraper adapted from [TheirTube](https://www.tomokihara.com/en/theirtube.html) (Kihara 2020). Standard video recommendations were retained; Shorts and advertisements were excluded. Titles were cleaned by removing emojis, punctuation, and excess whitespace, and only English-language titles were kept.

**Total recommendations collected:**

| Persona | N |
|---|---|
| Watch (control) | 1,517 |
| Like | 1,359 |
| Not-Interested | 1,200 |
| Combined | 1,220 |
| **Total** | **5,296** |

## Embeddings

Titles were embedded using [Sentence-BERT](https://arxiv.org/abs/1908.10084) `all-MiniLM-L6-v2`). Embeddings were reduced from 768 to 100 dimensions via PCA prior to clustering and t-SNE visualisation.

## Content Categories

Titles were assigned to one of three content categories using a combination of HDBSCAN clustering (min cluster size = 50) and GPT-4 zero-shot classification (`gpt-4-0613`):

- **Puzzle Gaming** – niche target interest
- **Generic Gaming** – adjacent gaming content
- **Mainstream** – non-gaming content

## File Description

`allDfEmbed.pkl` Pandas DataFrame (pickle) containing recommendation titles, persona/account labels, SBERT embeddings.

### Loading the data

```python
import pandas as pd
df = pd.read_pickle("allDfEmbed.pkl")
```

Ethics

This study received institutional ethics approval from the University of Oxford (CUREC Ref: 1562454). Data collection used only newly created accounts with age-appropriate content, limiting the risk of affecting real users' recommendations. The study followed Oxford CUREC Best Practice Guidance 06 for internet-mediated research.

Citation

If you use this dataset, please cite the accompanying paper:

@article {cho2026youtube,
title={Evaluating feedback mechanisms for aligning YouTube recommendations with user interests},
author={Cho, Desiree and Hale, Scott A. and Zhao, Jun and Shadbolt, Nigel and Przybylski, Andrew K.},
year={2026},
note={Manuscript under review}
}

Authors

Desiree Cho, Scott A. Hale, Jun Zhao, Nigel Shadbolt, Andrew K. Przybylski
Oxford Internet Institute / Department of Computer Science / Institute for Ethics in AI, University of Oxford
Correspondence: desiree.cho@cs.ox.ac.uk

README.md

@@ -0,0 +1,15 @@
+---
+license: cc-by-4.0
+language:
+- en
+task_categories:
+- text-classification
+tags:
+- recommender-systems
+- youtube
+- algorithmic-auditing
+- social-media
+- nlp
+- embeddings
+pretty_name: YouTube Alignment – Recommendation Title Embeddings
+---