Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Manju Bala
I. P. College for Women,
Delhi University, Delhi
manjugpm@gmail.com
Abstract: In past few years, the data available on internet has multiplied at an alarming rate. Tweets, reviews, blogs and comments on social
media have been a huge factor which has resulted in such a huge amount of increase in the available data. Because of this datasets being highly
unstructured and of high dimensionality, sentiment classification becomes a very tiresome task. Sentiment Analysis is used to estimate the user
opinion on various issues. It consequently mines states of mind and perspectives of clients on particular issues. It‟s a multistep preparation where
choosing and extracting elements is an indispensable stride that controls execution of sentiment classifier. In this paper we have used three
supervised techniques namely SVM, Decision Tree and Nave Bays Algorithm and three unsupervised techniques called DE, PSO and K-Means
The results are validated using different three benchmark labeled datasets data sets and on the different feature sets We have also performed
feature selection using genetic algorithm and validated results using the features selected by the GA Experimental results shows that supervised
techniques have outperformed supervised techniques on one dataset while for the two datasets supervised techniques have outperformed
unsupervised techniques
Keywords: Sentimental Analysis, Feature Extraction, Feature Selection, Swarm Intelligence.
__________________________________________________*****_________________________________________________
I. INTRODUCTION
3.1 Twitter-sanders-apple The following results were obtained after applying non
Sanders Analytics have collected this dataset for Apple Corp. supervised techniques:
on four separate topics: Apple, Microsoft, Twitter and Google.
It consists of a total of 479 reviews, out of which 163 are Table 3: Accuracy obtained using PSO, K Means and DE
positive and 316 are negative. (for Dataset 1):
PSO DE K Means
3.2 Amazon Movie Reviews 93 features 75.36 75.36 52.6
This dataset contains movie reviews collected from amazon 73 features 65.97 65.97 75.36
website. Positive reviews are labeled as „pos‟ and negative
reviews as „neg‟. There are a total of 8544 reviews. 3998 are 53 features 65.97 65.13 66.17
labeled as positive and remaining 4546 as negative. 33 features 65.34 68.47 65.35
After GA 75.36 75.36 50.93
3.3 Amazon Food Reviews
Food product reviews have been collected from amazon
List of features selected by GA :
website and after being classified as negative or positive have
575
IJFRCSCE | November 2017, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 3 Issue: 11 573 – 577
_______________________________________________________________________________________________
Clout, Authentic, Tone, Function, Shehe, Auxverb, Adj,
Number, Anx, Social, Friend, Female, Male, Insight, Discrep,
See, Affiliation, Achieve, Focuspast, Relativ, SemiC
REFERENCES
[1] Michael Crawford, Taghi M. Khoshgoftaar, Joseph D.
Prusa, Aaron N. Richter and Hamzah Al Najada, “Survey of
review spam detection using machine learning
techniques”,2011
[2] Hamad Alhammady. “Weighted Naive Bayesian Classifier”,
2007
[3] Luiz F. S. Coletta, Nadia F. F. da Silva, Eduardo R.
Hruschka,Estevam R.Hruschka Jr.” Combining
Figure 5: Result of Non Supervised Techniques on DS 2
Classification and Clustering for Tweet Sentiment
Analysis”,2014.
576
IJFRCSCE | November 2017, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________
International Journal on Future Revolution in Computer Science & Communication Engineering ISSN: 2454-4248
Volume: 3 Issue: 11 573 – 577
_______________________________________________________________________________________________
[4] Akshi Kumar, Renu Khorwal* and Shweta Chaudhary: “A
survey on Sentiment Analysis using Swarm Intelligence”,
2016
[5] Bo Pang and Lillian Lee: “Thumbs Up? Sentiment
Classification using Machine Learning Techniques”, 2002
[6] Henrique Siqueria and Favia Barros: A feature extraction
process for Sentiment Analysis of Opinions on services
[7] Muhammad Zubair Asghar, Aurangzeb Khan, Shakeel
Ahmad, Fazal Masud Kundi: “A review of feature
extraction in sentiment analysis”, 2014
[8] Bingwei Liu*, Erik Blasch, Yu Chen, Dan Shen*, and
Genshe Chen*: “Scalable Sentiment classification for Big
Data Analysis using Naive Bayes Classifier, 2013
[9] Avinash Chandra Pandey ∗, Dharmveer Singh Rajpoot,
Mukesh Saraswat: “Twitter sentiment analysis using hybrid
cuckoo search method”
[10] Twitter-sanders-apple:(2015). http://boston.lti.cs.cmu.edu/
classes/95- 865- K/HW/HW3/ .
[11] movie_pang: http:// www.cs.cornell.edu/ people/ pabo/
movie-review-data/
[12] amazon fine food review https:// www.kaggle.com/ snap/
amazon-fine-food-reviews
[13] PSO http://www.swarmintelligence.org/tutorials.php
[14] SVM:https:// en.wikipedia.org/ wiki/ Support_vector_
machine
[15] DE_algorithm: https://en.wikipedia.org/ wiki/ Differential
_evolution
[16] K-means:https://sites.google.com/site/dataclustering
algorithms/k-means-clustering-algorithm
[17] Naïve Bayes Classifier: https://en.wikipedia.org/wiki/
Naïve_Bayes_classifier
577
IJFRCSCE | November 2017, Available @ http://www.ijfrcsce.org
_______________________________________________________________________________________