International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395 -0056
Volume: 04 Issue: 06 | June -2017
p-ISSN: 2395-0072
www.irjet.net
SENTIMENT ANALYSIS ON TWITTER DATA Chandan Arora[1], Dr. Rachna[2] 1Research
Scholar, Department of Computer Science, Global Institutes of Management &Emerging Technologies, Amritsar, India
2Associate
Professor, Department of Computer Science, Global Institutes of Management & Emerging Technologies, Amritsar, India
---------------------------------------------------------------------***--------------------------------------------------------------------website can utilize sentiment analysis to discern if their Abstract - Twitter is one of the most commonly used platforms for sharing opinions, expressing views. Sentiment Analysis on twitter can allow users to understand the opinions expressed in tweets and classifying them in positive or negative categories. The organizations can use sentiment analysis to get an idea of the customer reviews of their products, and subsequently try and improve their services based on the reviews.
products are being liked by the customers or not. The reviews for the products can be generalized into either positive or negative as well as neutral categories[3]. SA can be simply put as “What other people think?”
Keywords: Data Mining, Sentiment Analysis,
Opinion- A conclusion open to dispute
Twitter, Classifiers.
View- A subjective opinion
1. INTRODUCTION
Belief- Deliberate acceptance and intellectual assent
Data Mining refers to extracting knowledge and discovering patterns from large data-sets. Almost all organizations collect and store data, and extract useful information, while discarding unnecessary data. The useful data is then analyzed in order to discover meaningful patterns [1]. Computing large data sets is an integral component of almost all organizations. The goal is to review the data sets and transform them into usable patterns. Data mining is sometimes also referred to as “knowledge discovery from data”, or KDD.Earlier techniques that were used to identify data patterns were Bayes’ theorem and Regression Analysis. As the times have gone by, the size of data sets has increased remarkably. As such, the discoveries in computer sciences like neural networks, clustering, etc. have made it easier to manage these data sets better [2]. The traditional techniques such as database can handle just a limited amount of data. In order to analyze millions of records, data mining has to be used. Specific computer algorithms such as neural networks, decision trees are applied to extract patterns from the given data sets.
Sentiment- opinion representing someone’s feelings[4].
The terms views, belief, sentiment and opinion can be defined as follows:
2.1 SENTIMENT ANALYSIS TECHNIQUES There are basically three techniques to perform Sentiment Analysis. 1. SA using machine learning. 2. SA using lexicon based techniques 3. SA using the above two techniques combined together. 1. Machine learning technique involves both supervised and unsupervised learning. 1.1 Unsupervised Learning is based on just inputs, without any mention of targets. It just relies on clustering. 1.2 Supervised Learning defines pre-specified targets which should be achieved, along with the inputs. Data set are trained to achieve significant outputs when encountered during decision-making.
2. SENTIMENT ANALYSIS
2. Lexicon-Based Approaches: Lexicon based method assigns positive or negative polarity based on the sentiment of each word and then a dictionary is created.
Sentiment Analysis (SA) elucidates users whether information or opinion regarding a certain product is positive, negative or neutral. Sentiment basically refers to any opinion or a feeling expressed by someone. Various organizations use this analysis to understand users’ opinion for their products. For example, a particular e-commerce
© 2017, IRJET
|
Impact Factor value: 5.181
We can use a combining function, for example, sum or average to find out the general sentiment of a document.
|
ISO 9001:2008 Certified Journal
|
Page 31