What is Sentiment Analysis ?
Sentiment analysis and opinion mining technology is an important tool for businesses, and governments to judge and evaluate the reaction of people for their services.
Here’s more formal definition of sentiment analysis : “Sentiment analysis, also called opinion mining, is the field of study that analyzes people’s opinions, sentiments, evaluations, appraisals, attitudes, and emotions towards entities such as products, services, organizations, individuals, issues, events, topics, and their attributes.”
What are Applications of Sentiment Analysis ?
Some governments have already started using this technology in health, tourism and other domains. This technology is also being used to encourage various vendors to improve their performance.
This technology is also being used by businesses to get a detailed, and targeted customer feedback for improving their services. Currently, this technology is being used to identify opinion holders, sentiment and target of opinion/sentiment.
Sentiment Analysis Using Opinion Tagger
The opinion tagger is a rule and dictionary based tagger. It detects positive and negative polarity words (such as ‘nice’ and ‘awful’), as well as intensifiers or weakeners (such as ‘very’ and ‘hardly’) and polarity shifters (such as ‘not’).
In addition, the module includes some simple rules that detect the holders and targets of the opinions related to the positive and negative polarity words.
A sentiment analysis task is accomplished in two steps using following modules:
- Polarity Tagger: Detects and tags polarity words(nice, awful, beautiful, amazing etc) and sentiment modifiers(very, not)
- Opinion Miner: Detects and tags opinions in text, extracting the three opinion entities defined: expression, holder and target.
Sentiment Analysis Example
Consider this simple example:
“My wife said that her room in that hotel was really dirty”
The polarity tagger module should detect and extract these elements:
- dirty → a word with a negative polarity
- really → a polarity intensifier
The second module, the opinion miner, detects opinions and their elements (expression, holder and target) in the input text.
This module takes as input a text tagged with the polarity tagger and provides the following output:-
Step 1) Expression detection → “really dirty”
Step 2) Target detection → “her room”
Step 3) Holder detection → “My wife”