Twitter Sentiment Analysis: Survey

Download Full-Text PDF Cite this Publication
Text Only Version


Twitter Sentiment Analysis: Survey

M. Gayathri 1,

M.Phil Research Scholar, PG&Research Department of Computer Science

Sadakathullah Appa College, Tirunelveli, Tamil Nadu, India

S. Shajun Nisha 2

Assistant Professor & Head, PG&Research Department of Computer Science Sadakathullah Appa College, Tirunelveli, Tamil Nadu, India

M. Mohamad Sathik 3

Principle & Research Supervisor, PG&Research Department of Computer Science

Sadakathullah Appa College, Tirunelveli, Tamil Nadu, India

Abstract : Twitter sentiment analysis is a technology that provides the methods to survey about public emotion related to products or events to them.These features are expressed through sentiment words, emojis and so on. Twitter messages are short, and also well suited for our knowledge.These challenges become obstacles in analyzing the accurate meaning of sentiments and detecting the suitable sentiment polarity.And these classficiation to automatically determine whether a tweets sentiment polarity is negative or positive.It is also an automatic process of these tweets under unsupervised learning.

Keywords: Sentiment analysis, Semantic concepts and Polarity detection.


The social media has given users a venue for expressing and sharing their opinions and thought on all kinds of topics and events. Twitter has the users of nearly 700million users and over all 350 million messages for per day. And also it becomes a gold mine for the organizations to monitor their actions over brands, market shares and competitors. The goal of Twitter sentiment classification is also to determine a tweets sentiment is positive or negative and neutral. The word sentiment is the word embeddings as the information. Twitter sentiment has also many advanced techniques because it has become an challenging task due to complexity and variety of the human language. It is the detection of attitudes enduring, affectively colored beliefs, dispositions towards objects or persons

  1. Holder(source) of attitude
  2. Target(aspect0 of attitude
  3. Type of attitude
    • From a set of types
      • Like, love, hate, value, desire, etc.
    • Or(more commonly simple weighted polarity:
      • Positive, negative, neutral. Together with strength
  4. Text containing the attitude
    • Sentence or entire document

The experimental result is also made upon a milestone for twitter users and has also become a benchmark for all their options and their products. And the methods used in this paper have also been shown the accuracy level of both inputs and outputs using the ANN Techniques (Aritifical Neural Network).

The organization of this paper is as follows. Section II discussed related work on this topic. We also proposed the latest algorithm Emotional Classification Method and Data mining Dictionary Classification. The process of the experiment is also shown in section IV.The discusses experiments results also shown in section V.Finally Concludes the paper.


Figure 1: Frame of Twitter sentiment Analysis

In this paper the keyword (Input) is initiated from the tweets and the comments is retrieval from the server it has been preprocessing with the three techniques

like SVM (Support vector machine), ANN (Artifical Nerualnetworks) and Kappa. The technique Kappa is classified for tweet comments and the final output is viewed as the graphical chart reffered to Fig 1.


    There have been many papers written related to this sentiment analysis of blogs and surveys. Until it, has been classified using Deep convolution Neural Network. It was able to achieve only one set of accuracy model of 82.9%using SVM as a unigram model. Researchers have also begun to investigate various ways of automatically collecting training data. The sentiment analysis is also the common type is also called as Polarity detection It also state a statement as Positive, negative and neutral. These sentiment analysis mainly for social media monitoring because it goes future side of the business likes and tweets or retweets and also it provides a qualitative point of view. Because in these trending world the business, share markets depending upon like these social medias, because these types of tweets also explain real-time processing based upon people opinions. The most surveyed paper on twitter sentiment analysis is mainly based upon lexican and supervised method. Lexican based method is also depends upon the accuracy level of the weighted along with their tweets. On othercase, supervised method is also based upon training with the datasets like (Random forest, Navie byes).

    Tweets are usually composed of incomplete expression, a variety of noise and poorly structured sentences because of the frequent presence of acronym, irregular grammar, ill- formed words and non-dictionary terms. Noise and unstructured Twitter data will affect the performance of tweet sentiment classifications.

    Table 1: Statistical values Twitter sentiment Analysis

    Techniques comparedPositiveNegativeTotal

    The feature model is one of most simple and effective representation model for natural language analysis and Twitter sentiment analysis. Some studies have shown state-of-the-art performance for sentiment classification on Twitter data using a unigram model. The number of hashtags, emoticons, negation, POS and the presence of capitalized words are used as features. It checks the access request against the policy specified for every user and yields a decision for the access. The use of multiparty access control mechanism can greatly enhance the flexibility for regulating data sharing in online social networks. Present any mechanism to enforce privacy concerns over data associated with Many userself a user posts a comment in a friends space, he/she can specify which users can view the comment.

  1. Twitter sentiment analysis using Deep convolutional nerualnetworks: Proceedings of the 38th International ACM SIGIR Conference on Research and Development in

    Information Retrieval[


  2. Semantic sentiment analysis on Twitter.[ e/semweb]
  3. Yu, B.: An evaluation of text classification methods for literary study. Literary and Linguistic Computing 23(3), 327 343 (2008)
  4. Sentiment analysis twitters datamining 2015 International conference [ 5753/proceeding]
  5. A survey on sentiment analysis challenges [Doaa MoheyEI- DinMohammed Husseinin year2018 Lexicon-based method Deep sentiment analysis method
  6. Comparsion Research on text pre- processing Methods on Twitter Sentiment Analysis in the year 2017 [Zhao Jianqiang Gui Xiaolin]
  7. Tweet Sentiment Analysis with classierEnsembles [Nadia F.F.da.Silvaa, Eduardo R.Hruschkaa,Estevam R. Hruschka Jr.( 2014)
  8. Tan, C., Lee, L., Tang, J., Jiang, L., Zhou, M., Li, P.: User- level sentiment analysis incorporating social networks. In: Proceedings of the 17th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2011)

Leave a Reply

Your email address will not be published. Required fields are marked *