dahara1/soccerscope-worldcup2026

SoccerScope: FIFA World Cup 2026 YouTube Trends Dataset

Release / リリース: v1.0.0

Daily snapshots of FIFA World Cup 2026 football videos trending on YouTube across countries, plus the official 104-match tournament results. The three schemas are exposed as separate Hugging Face configurations so each can be viewed and loaded independently.

• データ提供元 / Dataset provider: TubeSaku
• 調査・分析：Webbigdata配信データ研究所（TubeSaku運営）
• Research and analysis by Webbigdata Live streaming Data Research Institute, operator of TubeSaku.
• Live analysis pages: https://webbigdata-jp.github.io/soccerscope/
• Search / report agent: https://soccer.tubesaku.com
• Hugging Face dataset: https://huggingface.co/datasets/dahara1/soccerscope-worldcup2026
• Version-specific archive / バージョン固定版: Zenodo DOI 10.5281/zenodo.21466996
• YouTube-analysis coverage: 2026-06-11 to 2026-07-21 (41 days)
• Official match-results coverage: 2026-06-11 to 2026-07-19 (104 matches)
• License: CC BY 4.0 for the aggregated statistics, derived analysis, and this compilation. Match scores and outcomes are factual records compiled from FIFA's official competition summary. The underlying YouTube videos, titles, thumbnails, and referenced comments remain the property of YouTube and their respective creators and are not relicensed by this dataset. Raw comment text is not included.

Dataset configurations

Use the configuration selector in the Dataset Viewer, or load a configuration explicitly. To reproduce the fixed v1.0.0 release, specify the Hugging Face revision tag.

from datasets import load_dataset

repo_id = "dahara1/soccerscope-worldcup2026"
revision = "v1.0.0"

daily = load_dataset(repo_id, "daily_snapshots", revision=revision)
videos = load_dataset(repo_id, "videos", revision=revision)
matches = load_dataset(repo_id, "matches", revision=revision)

daily_snapshots — daily_snapshots.jsonl (41 rows)

One row per collection day.

field	description
date	Collection date, YYYY-MM-DD
videos_analyzed	Number of videos analyzed that day
countries	Number of countries whose YouTube search results were scanned
views_total / likes_total / comments_total	Summed statistics across analyzed videos
comments_analyzed	Number of comments used for sentiment analysis
sentiment_positive_avg / sentiment_negative_avg / sentiment_neutral_avg	Average sentiment ratio (%) across videos
top_teams	List of {team, mentions, lean, positive, neutral, negative}
top_positive_themes / top_negative_themes	List of {theme_en, theme_ja, mentions}

videos — videos.jsonl (783 rows)

One row per video that appeared in a day's top-by-views and/or top-by-reach ranking, merged into one row when it appeared in both.

field	description
date	Collection date, YYYY-MM-DD
video_id	YouTube video ID
title / url	Video title and YouTube URL
views	View count at collection time
reach	Number of countries whose search results surfaced the video
countries	Country codes where the video appeared
narration_en / narration_ja	Short Gemini-generated summary
rank_by_views / rank_by_reach	Daily ranking, or null if absent

matches — matches.jsonl (104 rows)

All 104 tournament matches from FIFA's official Competition Summary. Dates and times are local to the listed host venue.

field	description
match_number	Official FIFA match number, 1–104
date / local_time / venue	Venue-local date, time, and host city/area
stage / group	Tournament stage and group letter where applicable
home_team / away_team	Teams in FIFA's listed order
home_score / away_score	Final match score before any shoot-out
home_penalty_score / away_penalty_score	Shoot-out score, otherwise null
winner / outcome	Winning team and home/away/draw label
decided_by	regulation, extra_time, penalties, or draw
official_result	FIFA's compact result notation
source_url / source_version	Provenance for the match row

Match source: FIFA World Cup 2026 Competition Summary, Version 34, published 2026-07-19 22:04 UTC.

Example use cases

• Align match outcomes with daily changes in national-team mentions and positive/negative fan sentiment.
• Compare which teams and themes different language regions discussed around the same match day.
• Study whether wins, losses, extra-time matches, or shoot-outs are associated with changes in video reach and comment reception.

How the YouTube-derived data was produced

The dataset is provided by TubeSaku. Raw collection uses the YouTube Data API for search, metadata, and comments, followed by Gemini-based comment sentiment analysis. The published dataset is generated from the daily stats.json files written by build_stats_page.py alongside the public HTML pages at https://webbigdata-jp.github.io/soccerscope/.

調査・分析：Webbigdata配信データ研究所（TubeSaku運営）

Research and analysis by Webbigdata Live streaming Data Research Institute, operator of TubeSaku.

Summary stats at build time

• Release: v1.0.0
• Daily snapshot rows: 41
• Ranked video rows: 783
• Official match rows: 104
• Total views across analyzed videos, summed across days: 15,445,999,226
• Total comments used for sentiment analysis, summed across days: 767,361

Citation / 引用

Use the following short attribution consistently.

日本語

出典：TubeSaku「SoccerScope: FIFA World Cup 2026 YouTube Trends Dataset」

English

Source: TubeSaku, “SoccerScope: FIFA World Cup 2026 YouTube Trends Dataset.”

BibTeX

@dataset{soccerscope_worldcup2026_v1_0_0,
  author    = {{TubeSaku}},
  title     = {{SoccerScope: FIFA World Cup 2026 YouTube Trends Dataset}},
  year      = {2026},
  version   = {1.0.0},
  publisher = {{Zenodo}},
  doi       = {10.5281/zenodo.21466996},
  url       = {{https://doi.org/10.5281/zenodo.21466996}},
  note      = {{Research and analysis by Webbigdata Live streaming Data Research Institute, operator of TubeSaku.}}
}

About

TubeSaku is the dataset provider and provides the public analysis services linked above.

TubeSaku is operated by GIPU Inc.