2

I am doing a research study on some of the most recent Chinese Sentiment Analysis tools. My purpose is to evaluate the analysis results of each tool on a set of sentences – I am looking for sentence-level analysis instead of document-level analysis. The set of sentences would be randomly selected from social networking sites, i.e. the sentences will most likely be colloquial. In particular, I am looking for tools (at least 4) with the following specifications:

  • The analysis scale must have at least two values, i.e. positive or negative. (I believe this is true for all sentiment analysis tools...just in case)
  • The tool could preferably obtain a sentiment score for any Chinese sentences, i.e. the dataset it uses is large enough to cover most of the commonly used words.
  • The tool is preferably open-source.
  • The tool is preferably an SDK or API.

Thanks in advance!

2

I think that what you are looking for is the combination of python and the natural language toolkit, (nltk).

The book Bird, Steven, Edward Loper and Ewan Klein (2009), Natural Language Processing with Python. O’Reilly Media Inc. states in its opening chapter:

Technologies based on NLP are becoming increasingly widespread. For example, phones and handheld computers support predictive text and handwriting recognition; web search engines give access to information locked up in unstructured text; machine translation allows us to retrieve texts written in Chinese and read them in Spanish; text analysis enables us to detect sentiment in tweets and blogs. By providing more natural human-machine interfaces, and more sophisticated access to stored information, language processing has come to play a central role in the multilingual information society.

and a quick check on the sentiment analysis finds both positive and negative results.

Please note that the latest version of NLTK has been made for python3 and that the book is in the process of being updated for that version.

To cover your requirements one by one:

  • The analysis scale must have at least two values, i.e. positive or negative. (I believe this is true for all sentiment analysis tools...just in case) - Yes
  • The tool could preferably obtain a sentiment score for any Chinese sentences, i.e. the dataset it uses is large enough to cover most of the commonly used words. - Yes but you may have to locate a good corpus for this a good starting point is here.
  • The tool is preferably open-source. - Yes completely
  • The tool is preferably an SDK or API. - Yes

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.