These articles can help you in this regard,
1. Sentiment analysis algorithms and applications: A survey.
2. Opinion Mining and Sentiment Analysis
3. Sentiment Analysis and Opinion Mining
Synthesis Lectures on Human Language Technologies
As far as the implementation is concerned, most of these machine learning algorithms are implemented in various languages specially in Java and python and tools such as Weka, R etc. can be used to implement such algorithms for research work.
If you are looking for open source: StanfordNLP does a decent job. If proprietary works, Semantria api's does really well.
Typically, there are two ways to crack sentiment analysis problem supervised and unsupervised. Supervised gives relatively better accuracy however training data for the same is difficult to find in most of the cases. Whereas for unsupervised learning, it is complicated. Please read Mining and summarizing customer reviews its a seminal paper for sentiment analysis.
Please read the papers, blogs of Prof. Bing Liu. Amazing work in this domain. You can also get amazon reviews data dump from Dr.Bind Liu's blog.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence