Twitter Sentiment Analysis For Product Reviews To Gather Information Using Machine Learning Technique
Twitter Sentiment Analysis For Product Reviews To Gather Information Using Machine Learning Technique
Twitter Sentiment Analysis For Product Reviews To Gather Information Using Machine Learning Technique
10(03), 669-674
RESEARCH ARTICLE
TWITTER SENTIMENT ANALYSIS FOR PRODUCT REVIEWS TO GATHER INFORMATION USING
MACHINE LEARNING TECHNIQUE
company's specific products via social media sites such as Facebook and Twitter. This study attempts to increase the
algorithm's accuracy by using the right kernel and setting the SVM hyper-parameters as key aspects in SVM
accuracy, as well as having a large number of training sets to broaden the hyper plane of vectors and create strong
support vectors. Sentiments will be acquired using the Twitter API, then submitted to a pre-processor, which will
remove any extraneous words. The pre-processed sentiments will then be transformed to SVM format, and the
structured data will be used as the data Train. We know that sentiment is a complicated mix of emotions and
viewpoints that serves as a foundation for action or judgement. In general, these are the feelings that users have had
concerning a reputable company's products. The examination of comments and ideas left on social media sites such
as blogs and social networks is known as sentiment analysis. Not only will the words be reviewed, but all of the user
comments and attitudes toward a brand will be examined as well. Our study will take into account words as
variables, such as context from Twitter or Facebook status. We shall attempt to propose a better machine learning
algorithm solution.
Architecture:-
670
ISSN: 2320-5407 Int. J. Adv. Res. 10(03), 669-674
must distinguish between important and irrelevant objects. For example, "BestBuy" is the name of a store, not a
competitive product.
Integration: -
Because we need to match the five pieces of information in the quintuple, integrating the about tasks is also difficult.
Comparative opinion:
A comparative view is a preferred relationship between two or more objects based on some of their common
characteristics. "Coke tastes better than Pepsi," for example, is frequently stated using the comparative or superlative
form of an adjective or adverb. Due to a lack of space
Related Work:-
A lot of work has been done in the topic of "Sentiment analysis" by a number of researchers in recent years. Work in
the field has been going on since the turn of the century. It was originally designed for binary classification, which
categorises thoughts or reviews into two categories: positive or negative. Paper [3] uses an unsupervised learning
system to predict review based on the average semantic orientation of a phrase containing adjective and adverb,
determining whether the phrase is positive or negative and classifying it as thumbs up or thumbs down. Some
sentiment assessments, such as [4], are based on a review of the product's user summarising system. The product
feature in [4] uses a filtering approach based on latent semantic analysis (LSA) to identify opinions. A comparison
of positive and negative statements is used in Paper [5.] It pulls data from the web and labels the word set manually,
which takes a lot of time and work. The author of [6] employed a rule-based system for sentiment analysis of
Chinese documents based on Baseline and SVM, which extracts the overall document polarity of specific words
using a sentiment word dictionary and adjusts it based on context information. The polarity of a word is computed in
another work [7] by adding up the polarities of all the words in the phrase, which might be positive or negative
671
ISSN: 2320-5407 Int. J. Adv. Res. 10(03), 669-674
depending on the sentence structure. Lakshmi and Edward [8] recommended preprocessing the data to improve the
original sentence's quality structure. For sentiment analysis, they used the LSA approach and cosine similarity. For
sentiment categorization, Basant Agarwal et al. [9] used the phrase pattern technique. It extracts contextual and
grammatical information from the document using part of speech-based rules and dependency relationships. The
author of [10] aimed to present aspect-based opinion polls based on unlabeled free form textual customer reviews
that do not compel customers to respond to the questions. M. Karamibekr and A.A. Ghorbani [11] suggested a
method for sentiment classification of a document in the social domain based on verbs as a key opinion term.
SentiFul is a sentiment lexicon created by Paper [12], which uses and expands it through synonyms, antonyms,
hyponyms, derivation, and compounding. They suggested a strategy for identifying four types of affixes based on
their role in sentiment features: propagation, weakening, reversal, and intensification. These methods attribute
polarity to sentiments, which aids in the expansion of the lexicon and hence improves sentiment analysis. For
sentiment analysis, a lot of work has been done where academics have researched and implemented soft-computing
methodologies, primarily fuzzy logic and neural works. [13] and [14] are two instances of works that use fuzzy logic
to solve problems. [13] makes a significant addition by using a fuzzy domain sentiment ontology tree extraction
technique. This algorithm creates a fuzzy domain sentiment ontology tree based on reviews, which involves
extracting sentiment words, product attributes, and feature relationships, and accurately predicting the polarity of
reviews. The authors of [14] developed a membership function-based fuzzy inference system. They defined and
standardised the procedure of quantifying the strength of reviewer's judgments in the presence of adverbial
modifiers by constructing membership functions. They used the approach to study adverbial modifier trigram
patterns.
Research Issues:-
Sentiment and subjectivity classification, which regarded the topic as a text classification problem, was the first step
in the field's research. Sentiment categorization determines whether a document (for example, product reviews) or a
sentence reflects a favourable or negative impression. A sentence's subjective or objective nature is determined by
its subjectivity categorization. Many real-world applications, on the other hand, necessitate a more in-depth
examination because the user frequently wants to know what perspectives have been voiced. For example, one
would wish to discover what characteristics of a product have been appreciated and condemned by users from a
product review. To introduce the overall problem, let's use the following iPhone review segment as an example
(each sentence has a number for easier reference):
"(1) I recently purchased an iPhone. (2) It was a lovely phone. (3) I really liked the touch screen. (4) The audio
quality was also excellent. (5) My mother, on the other hand, was furious with me since I didn't notify her before I
bought it. (6) She also said the phone was overpriced and requested that I return it to the store."
What we want to get out of this review is the answer to that question. The first thing we notice is that this review
contains a variety of viewpoints. Three favourable opinions are expressed in sentences (2), (3), and (4), whereas
negative opinions are expressed in sentences (5) and (6). Then we see that each opinion has a certain objective on
which it is expressed. The opinion expressed in line (2) is about the iPhone as a whole, whereas the opinions
expressed in sentences (3) and (4) are about the iPhone's "touch screen" and "voice quality," respectively. The
sentiment expressed in line (6) is about the price of an iPhone, but the sentiment expressed in phrase (5) is about
"me," not iPhone. This is a crucial point. In an app, the user may be interested in opinions on specific targets but not
all (for example, "myself" is improbable). Finally, we may notice the sources of information or the people who have
the opinions. In sentences (2), (3), and (4), the source or bearer of the opinions is the review author ("I"), however in
words (5) and (6), it is "my mum." We can define sentiment analysis or opinion mining with this case in mind.
672
ISSN: 2320-5407 Int. J. Adv. Res. 10(03), 669-674
online reviews and compress the information acquired are the focus of research and development. Several
companies currently offer tools that track public attitudes on a wide scale by providing graphical summaries of
blogosphere trends and opinions. The development of opinion-tracking tools is crucial for business. There are also
various technologies available to assist organisations in extracting and analysing information from blogs concerning
large-scale customer patterns.
Conclusion:-
We introduced a set of machine learning approaches with semantic analysis for identifying sentences and product
reviews based on Twitter data in this research. The main goal is to analyse a huge number of reviews using a pre-
labeled Twitter dataset. The nave byes approach provides us a better result than maximal entropy, and SVM is
subjected to a unigram model, which gives us a better result than just utilising it. When the semantic analysis
WordNet is combined with the above approach, the accuracy increases to 89.9% from 88.2%. The training data set
can be expanded to improve the feature vector-related sentence identification process, as well as WordNet for
review summarization. It may provide a better visual representation of the content, which will be beneficial to the
users.
673
ISSN: 2320-5407 Int. J. Adv. Res. 10(03), 669-674
References:-
[1] J. Xie and B. K. Szymanski. Community detection using a neighborhood strength driven label propagation
algorithm. In IEEE Network Science Workshop 2011, pages 188-195, 2011.
[2] J. Xie and B. K. Szymanski. Towards linear time overlapping community detection in social networks. In
PAKDD, pages 25-36, 2012.
[3] J. Xie, B. K. Szymanski and X. Liu. SLPA: Uncovering Overlapping Communities in Social Networks via A
Speaker-listener Interaction Dynamic Process. In Proc. of ICDM 2011 Workshop, 2011.
[4] S. Gregory. Finding overlapping communities in networks by label propagation. New J. Phys., 12:103018, 2010.
[5] K. Wakita and T. Tsurumi. Finding community structure in mega-scale social networks. In WWW Conference,
pp. 1275-1276, 2007.
[6] M. E. J. Newman and M. Girvan. Finding and Evaluating Community Structure in Networks. Phys. Rev. E, 69,
pp. 026113, 2004.
[7] V. Blondel, J. Guillaume, R. Lambiotte and E. Lefebvre. Fast Unfolding of Communities in Large Networks. J.
Stat. Mech., 2008.
[8] S. White and P. Smyth. A spectral clustering approach to finding communities in graphs. Proc. of SIAM
International Conference on Data Mining, pp. 76-84, 2005.
[9] I. Leung, P. Hui, P. Lio, and J. Crowcroft. Towards real-time community detection in large networks. Phys. Rev.
E, 79:066107, 2009.
[10] Y.Singh, P.Kumar Bhatia, and Omprakash Sangwan. "A review of studies on machine learning techniques."
International Journal of Computer Science and Security 1, no. 1, 70-84, 2007
[11] P.D.Turney, "Thumbs up or thumbs down?: semantic orientation applied to unsupervised classification of
reviews." In Proceedings of the 40th annual meeting on association for computational linguistics, pp. 417-424.
Association for Computational Linguistics, 2002.
[12] L.Ramachandranand Edward F. Gehringer. "Automated assessment of review quality using latent semantic
analysis." In Advanced Learning Technologies (ICALT), 2011 11th IEEE International Conference on, pp. 136-
138, 2011
[13] J. Fernando S_anchez-Rada1, Marcos Torres1, Carlos A. Iglesias1, Roberto Maestre2, and Esther Peinado, ―A
Linked Data Approach to Sentiment and Emotion Analysis of Twitter in the Financial Domain‖, Retrieved 2014
[14] Jalaj S. Modha* Prof & Head Gayatri S. Pandi Sandip J. Modha, ― Jalaj S. Modha* Prof & Head Gayatri S.
Pandi Sandip J. Modha‖, International Journal of Advanced Research in Computer Science and Software
Engineering, 2013
[15] H.Saif, Y. He, and H. Alani. "Alleviating data sparsity for twitter sentiment analysis." CEUR Workshop
Proceedings (CEUR-WS. org), 2012.
[16] H. Saif, Y.He, and H. Alani. "Semantic sentiment analysis of twitter." In The Semantic Web–ISWC 2012, pp.
508-524. Springer Berlin Heidelberg, 2012.
[17] A.Kumar and T. M. Sebastian, ―Sentiment Analysis on Twitter‖, International Journal of Computer Science
Issues, Vol. 9, Issue 4, No 3, 2012
[18] P. Bo, L. Lee, and S. Vaithyanathan. "Thumbs up?: sentiment classification using machine learning
techniques." In Proceedings of the ACL-02 conference on Empirical methods in natural language processing-
Volume 10, pp. 79-86. Association for Computational Linguistics, 2002.
[19] A. Go, R. Bhayani, and L. Huang. "Twitter sentiment classification using distant supervision." CS224N
Project Report, Stanford, pp.1-12, 2009
[20] L. Zhang, R. Ghosh, M. Dekhil, M. Hsu, and B. Liu. "Combining lexicon-based and learning-based methods
for Twitter sentiment analysis." HP Laboratories, Technical Report HPL-2011 89, 2011
[21] A. Mudinas, D. Zhang, and M. Levene. "Combining lexicon and learning based approaches for concept-level
sentiment analysis." In Proceedings of the First International Workshop on Issues of Sentiment Discovery and
Opinion Mining, p. 5. ACM, 2012.
674