Datasets:
Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code: DatasetGenerationCastError
Exception: DatasetGenerationCastError
Message: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 14 new columns ({'country', 'avg_duration', 'avg_engagement_per_1k', 'median_er', 'total_videos', 'top_hashtag_views', 'avg_velocity', 'platform', 'p95_views', 'avg_share_rate', 'total_views', 'avg_comment_ratio', 'top_hashtag', 'avg_save_rate'}) and 2 missing columns ({'column', 'description'}).
This happened while the csv dataset builder was generating data using
hf://datasets/tarekmasryo/youtube-tiktok-trends-2025/data/country_platform_summary_2025.csv (at revision 2f6c1d9a398f14cdda91124c28bcb4fadd8c6c4f)
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1831, in _prepare_split_single
writer.write_table(table)
File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 714, in write_table
pa_table = table_cast(pa_table, self._schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
return cast_table_to_schema(table, schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
raise CastError(
datasets.table.CastError: Couldn't cast
country: string
platform: string
total_videos: int64
total_views: int64
median_er: double
p95_views: double
avg_duration: double
avg_velocity: double
avg_comment_ratio: double
avg_share_rate: double
avg_save_rate: double
avg_engagement_per_1k: double
top_hashtag: string
top_hashtag_views: int64
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 2005
to
{'column': Value('string'), 'description': Value('string')}
because column names don't match
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1455, in compute_config_parquet_and_info_response
parquet_operations = convert_to_parquet(builder)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1054, in convert_to_parquet
builder.download_and_prepare(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 894, in download_and_prepare
self._download_and_prepare(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 970, in _download_and_prepare
self._prepare_split(split_generator, **prepare_split_kwargs)
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1702, in _prepare_split
for job_id, done, content in self._prepare_split_single(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1833, in _prepare_split_single
raise DatasetGenerationCastError.from_cast_error(
datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 14 new columns ({'country', 'avg_duration', 'avg_engagement_per_1k', 'median_er', 'total_videos', 'top_hashtag_views', 'avg_velocity', 'platform', 'p95_views', 'avg_share_rate', 'total_views', 'avg_comment_ratio', 'top_hashtag', 'avg_save_rate'}) and 2 missing columns ({'column', 'description'}).
This happened while the csv dataset builder was generating data using
hf://datasets/tarekmasryo/youtube-tiktok-trends-2025/data/country_platform_summary_2025.csv (at revision 2f6c1d9a398f14cdda91124c28bcb4fadd8c6c4f)
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
column
string | description
string |
|---|---|
platform
|
Platform (TikTok/YouTube)
|
country
|
Country ISO-2 code
|
region
|
Region macro label (if available)
|
language
|
Primary language inferred from country (fallback to 'en')
|
category
|
Video category (if available)
|
hashtag
|
Primary hashtag aligned with genre
|
title_keywords
|
Short realistic title-like keywords
|
author_handle
|
Creator handle/channel (brand-like, synthetic)
|
sound_type
|
Sound type (if present)
|
music_track
|
Music track (if present)
|
week_of_year
|
ISO week number (1–53)
|
duration_sec
|
Shorts-style duration in seconds (TikTok ~5–75, YouTube ~5–90)
|
views
|
Total views
|
likes
|
Likes count
|
comments
|
Comments count
|
shares
|
Shares count
|
saves
|
Saves count
|
engagement_rate
|
(likes+comments+shares+saves) / views
|
trend_label
|
No description available
|
source_hint
|
No description available
|
notes
|
No description available
|
device_type
|
Android/iOS/Web
|
upload_hour
|
Hour of day video published (0–23)
|
genre
|
Canonical content genre
|
trend_duration_days
|
Days the video remained trending (synthetic)
|
trend_type
|
Short (≤7), Medium (8–21), Evergreen (≥22)
|
engagement_velocity
|
views / trend_duration_days
|
dislikes
|
Dislikes (synthetic, platform-aware)
|
comment_ratio
|
comments / views
|
share_rate
|
shares / views
|
save_rate
|
saves / views
|
like_dislike_ratio
|
likes / (dislikes+1)
|
publish_dayofweek
|
Day of week of publish_date
|
publish_period
|
Part of day bucket (Morning/Afternoon/Evening/Night)
|
event_season
|
Seasonal/contextual event (Ramadan, SummerBreak, BackToSchool, HolidaySeason, None)
|
tags
|
YouTube-like comma-separated tags aligned with genre
|
sample_comments
|
One short synthetic multilingual comment
|
creator_avg_views
|
Avg views per video for the creator (across dataset rows)
|
creator_tier
|
Creator tier based on avg views: Micro / Mid / Macro / Star
|
season
|
Climatological season (Winter/Spring/Summer/Fall)
|
publish_date_approx
|
ISO date reconstructed/approximated within 2025 (clipped to 2025-09-12)
|
year_month
|
Publish year-month for time-series aggregation
|
title
|
Short realistic video title (synthetic)
|
title_length
|
Character count of title
|
has_emoji
|
Whether title contains emoji (1/0)
|
avg_watch_time_sec
|
Estimated average watch time (seconds)
|
completion_rate
|
avg_watch_time_sec / duration_sec
|
device_brand
|
If mobile: device brand (iPhone, Samsung, Huawei, Xiaomi, Oppo, Vivo, Pixel, Other); Web → Desktop
|
traffic_source
|
Coarse discovery source (TikTok: ForYou/Following/Search/External; YouTube: Home/Suggested/Search/External)
|
is_weekend
|
Publish on Fri/Sat/Sun = 1
|
row_id
|
Deterministic MD5 over [platform, country, author_handle, title, publish_date_approx, duration_sec] (primary key)
|
engagement_total
|
likes + comments + shares + saves
|
like_rate
|
likes / views
|
dislike_rate
|
dislikes / views
|
engagement_per_1k
|
Total engagements per 1,000 views
|
engagement_like_rate
|
Likes divided by Views; NaN when Views <= 0
|
engagement_comment_rate
|
Comments divided by Views; NaN when Views <= 0
|
engagement_share_rate
|
Shares divided by Views; NaN when Views <= 0
|
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
null | null |
End of preview.
🎬 YouTube Shorts & TikTok Trends 2025
Author: Tarek Masryo · Kaggle
License: CC0 (Public Domain) — Free for open research & education
📌 Dataset Summary
A comprehensive snapshot dataset capturing short-form video activity on YouTube Shorts & TikTok in 2025 (Jan–Aug).
- 🌍 Coverage: 100+ countries, 2 major platforms
- 📂 Package: raw video-level file, ML-ready version, monthly summaries, country rollups, top creators, top hashtags, and a data dictionary
- ⚡ Features: standardized schemas, deduplicated IDs, engagement metrics (views, likes, comments, shares, saves, completion rate)
💡 For machine learning tasks, use the ML-ready file (youtube_shorts_tiktok_trends_2025_ml.csv).
⚠️ trend_label is a snapshot approximation (not full time-series). It’s a challenging ML target (≈25–35% baseline accuracy).
🗂 Dataset Structure
Main Files
youtube_shorts_tiktok_trends_2025.csv→ 48,079 rows × 58 columns (raw video-level data: platform, country, region, language, category, hashtags, author_handle, sound/music metadata, full engagement metrics).youtube_shorts_tiktok_trends_2025_ml.csv→ 50,000 rows × 32 columns (ML-ready version: cleaned & feature-engineered for faster modeling).
Companion Files
monthly_trends_2025.csv→ 480 rows × 8 cols (video counts, views, avg engagement, velocity)country_platform_summary_2025.csv→ 60 rows × 14 cols (totals, medians, percentiles)top_creators_impact_2025.csv→ 1,000 rows × 20 cols (creator stats, cumulative engagement, avg rates)top_hashtags_2025.csv→ 82 rows × 18 cols (hashtag usage, reach, ratios, velocity)DATA_DICTIONARY.csv→ 58 rows (column names, descriptions, types)
🚀 How to Use
from datasets import load_dataset
import pandas as pd
# Load dataset from HuggingFace
ds = load_dataset("TarekMasryo/YouTube-Shorts-TikTok-Trends-2025")
df = ds["train"].to_pandas()
print(df.head())
# Example: work with companion file
hashtags = pd.read_csv("top_hashtags_2025.csv")
print(hashtags.sort_values("views", ascending=False).head(10))
- Downloads last month
- 319