The Influence Of Sentiment Analysis

Sentiment analysis is the classification of text into pre-defined sentiment classes e.g. positive or negative i.e. polarity of the text. Sentiment analysis research has attracted a large number of researchers around the globe. Sentiment analysis attempts to determine whether a given text is subjective or objective and further, whether a subjective text contains positive or negative opinion. A lot of research has been done for detecting sentiment from the text. Still, there is a huge scope of improvement of these existing sentiment analysis models. The performance of the existing methods can be further improved by including more semantic information. Sentiment analysis techniques are categorized into machine learning (ML) and semantic orientation

Now a days when everybody is so addicted to social media and everyone is an opinionated person, people frequently and openly share their views on any happenings, events, political situation, product, entertainment industry etc. It would be quite likable and perplexing to work on this particular aspect. Sentiment analysis can be done at several level of granularity. Figure 2 shows some of the application areas of sentiment analysis.

Sentiment analysis and many other text processing based research areas highly require deep natural language processing algorithms to solve various problems in the respective fields. Extracting and understanding of sentiments from a set of documents heavily depends on NLP. For example, very basic NLP tasks are required for pre-processing of the text, such as, sentence boundary detection, word-tokenization, stemming, etc. More in-depth language processing is required to actually extract the sentiments from a document based on lexicons and other important

Dealing with noisy data is challenging e.g. follow the link and special prizes.

 Negation when attached to positive sentence gives negative meaning e.g. not very attractive.

 Longer distance dependency e.g. the sentence “The computer installed last month in our firm placed on the fifth floor crashed” has long distance dependency. Last few words do not exactly define the object that crashed. Actual target noun is at the start of the sentence whereas the verb acting on it is in the last part of the sentence.

 Sometimes negation intensify concerns e.g. not only good but amazing.

 Languages differ from one geographical region to another.

 Deliberate spelling mistakes, use of abbreviations/short forms (I m f9 feeling gr8) and emoticons, use of roman language, and non-standard standardized words (e.g. un-friend, tweeting)

 Mixed opinion like “Well as usual Keanu Reeves is nothing special, but surprisingly, the very talented Laurence Fish Bourne is not so good either, I was surprised.”