Paper Publish.docx

ABSTRACT Sentiment analysis deals with identifying and classifying opinions or sentiments expressed in source text. Microblogging today has become a very popular communication tool among Internet users. Millions of messages are appearing daily in popular web-sites that provide services for microblogging such as Twitter, Facebook. Authors of those messages write about their life, share opinions on variety of topics and discuss current issues. Because of a free format of messages and an easy accessibility of microblogging platforms, Internet users tend to shift from traditional communication tools to microblogging services. As more and more users post about products and services they use, or express their views, microblogging web-sites become valuable sources of people’s opinions and sentiments. Such data can be efficiently used for marketing or social studies.

INTRODUCTION The age of Internet has changed the way people express their views. It is now done through blog posts, online discussion forums, product review websites etc. When someone wants to buy a product, they will look up its reviews online before taking a decision. The amount of user generated content is too large for a normal user to analyze. So, to automate this, various sentiment analysis techniques are used. Sentiment analysis, or opinion mining, aims at user’s attitude and opinions by investigating, analyzing and extracting subjective texts involving users’ opinions, preferences and sentiment. This is used particularly in data mining field for social media with many applications including product ratings and feedback analysis and customer decision making etc. Presence of emoticons, slang words and misspellings in tweets forced to have a pre-processing step before feature extraction.

There are different feature extraction methods for collecting relevant features from text which can be applied to tweets also. But the feature extraction is to be done in two phases to extract relevant features. In the first phase, twitter specific features are extracted. Then these features are removed from the tweets to create normal text. Again, feature extraction is done to get more features. This is the

idea used in this paper to generate an efficient feature vector for analyzing twitter sentiment. Since no standard dataset is available for twitter posts of electronic devices, we created a dataset by collecting tweets for a certain period. By doing sentiment analysis on a specific domain, it is possible to identify the influence of domain information in choosing a feature vector. Different classifiers are used to do the classification to find out their influence in this domain with this feature vector.

EXISTING SYSTEM A major benefit of social media is that we can see the good and bad things people say about the particular brand or personality. The bigger your company gets difficult it becomes to keep a handle on how everyone feels about your brand. For large companies with thousands of daily mentions on social media, news sites and blogs, it’s extremely difficult to do this manually. To combat this problem, sentimental analysis software is necessary. This software can be used to evaluate the people's sentiment about particular brand or personality.

PROPOSED SYSTEM 1. Data Collection: For performing sentimental analysis we need twitter data consisting of tweets about a particular keyword or query term. For collecting the data and tweets we have used Twitter public API available for general public for free. It is the part of Data Collection. 2. Data Pre-Processing: It is a process to remove the unwanted words from tweets that does not amount to any sentiments.  Emotional Icons- 170 emoticons; identified emotional icons and remove them.

 URLs-does not signify any sentiment; replaced it with a word |URL|  Stop words- words as “a‟, “is”, “the”; does not indicate any sentiment  Usernames and HashTags- @ symbol before the username and # for topic; both replaced with AT_USER.  Repeated Letters- huuuungry, huuuuuuungry, huuuuuuuuuungry into the token “hungry".  Slag Words- Non English words Data 3. Different Ways of Classifications: Binary Classification: It is a two way categorization i.e. positive or negative. 3-Tier: In this, Tweets are categorized as Positive, Negative and Neutral. 5-Tier: Tweets are bucketed in 5 Classes namely: Extremely Positive, Positive, Neutral, Negative and Extremely Neutral.

IMPLEMENTATION CONCLUSION