Sentiment Analysis is the use of Natural Language Processing, analysis of text, linguistics computation and follows a strict process of identifying, extracting, and quantifying the things with respect to any analysis that can be applied to many of the current relatable situations eg:-Twitter Analysis
The major process in sentiment analysis is identifying the text of the sentence, has it may subjective or objective, and this identification is a more complex problem when compared with polarity detection.
Subjective identification may depend on the context of words or sentences and objectivity may depend upon the subjective context, henceforth to overcome these scientists have come up with different methodologies, with respect to automatic learning for major reasons:
- Comprehension because of ambiguity in languages
- Human errors
- Time consumption is more
The 3 main categories of existing approach for Sentiment analysis is:
- Knowledge-based techniques: Classification based on the effect of words
- Statistical approach: Leverage it from Machine learning such as support vector machines, a bag of words etc.
- Hybrid approach: Involves both the approaches of knowledge-based techniques and statistical techniques.
Sentiment analysis can be evaluated using a task-based approach and a separate training model is required for implementation to get the representation that is more accurate for given data set based on sentiment.