Turkish Tweet Datasets for Top-N Emoji Recommendation

A collection of 42 datasets for training and testing emoji recommendation models based on Turkish tweets. The datasets were created by author Yiltan Bitirim and are associated with a research article on the topic. The data was last updated on 2026-04-25 and is available as a 117.2 MB RAR archive.

Use Cases

• Training emoji recommendation models based on Turkish tweet text.
• Evaluating the performance of Top-N recommendation algorithms.
• Analyzing the relationship between Turkish language constructs and emoji usage.

Strengths

• Comprises 42 distinct datasets for training and testing.
• Data is specifically focused on Turkish-language tweets, a defined domain.
• Dataset size is 117.2 MB, indicating a medium-scale collection.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Yiltan Bitirim

Collection Method

Likely collected from the social media platform X (formerly Twitter).

Freshness

Last updated 2026-04-25 05:49:56; freshness should be verified.