Sentiment Analysis Corpus

  Authors

Köksal, Abdullatif ; Özgür, Arzucan

  Item identifier

https://hdl.handle.net/20.500.12913/21

 Referenced by

https://ieeexplore.ieee.org/document/9477814

 Date issued

2021-06-01

 Type

corpus

 Language(s)

 Description

The corpus includes three files which are the train, validation, and test datasets. Each of the datasets is formed of id-value pairs. The id field is the tweet id and the value field is the sentiment (positive, neutral, or negative) of the tweet. Note that tweet ids rather than the tweets themselves are given due to Twitter copyright issues. The tweets can be downloaded via the official Twitter API. Example: {"id": "603539206925852674", "value": "positive"}

 Publisher

Boğaziçi University

 Acknowledgement

TÜBİTAK

Project code: 2211 graduate scholarship

Project name: BİDEB

 Subject(s)

Twitter three-class sentiment analysis

 Collection(s)

Collection

Show full item record