1. Remove all urls, hashtags, and targets.
2. Correct the spellings.
3. Replace all the emoticons with their sentiment.
4. Remove all punctuations ,symbols, numbers.
5. Remove Stop Words.
6. Expand Acronyms.
7. Remove Non-English Tweets.

University of Michigan - Ann Arbor

Social networking platform, including Twitter, Facebook, Google+, allow people to view hot topics and newest events and express their attitudes. Given the quantity of the data, the dataset is extremely important for verifying the opinion mining and sentiment analysis algorithms' performance.

Sentiment Analysis of Twitter Data

https://arxiv.org/ftp/arxiv/papers/1601/1601.06971.pdf

Sentiment Analysis of Twitter Data: A Survey of
Techniques


1. HASH
2. EMOT
3. ISIEVE
4. Columbia univ.dataset
5. Patient Dataset
6. Sample
7. Stanford Dataset
8. Stanford
9. Spam Dataset
10. Soe Dataset

Publicly Available Datasets for Twitter Sentiment Analysis

1. Word and their frequencies.
2. Parts of Speech Tags.
3. Opinion Words and Phrases.
4. Position of Terms.
5. Negation.
6. Syntax.

Learn Before

Related

Related