SENTIMENT ANALYSIS APPROACH IN NATURAL LANGUAGE PROCESSING FOR DATA EXTRACTION

https://iaeme.com/Home/journal/IJCET 27 editor@iaeme.com
International Journal of Computer Engineering and Technology (IJCET)
Volume 13, Issue 2, May – August 2022, pp. 27-32, Article ID: IJCET_13_02_004
Available online at https://iaeme.com/Home/issue/IJCET?Volume=13&Issue=2
ISSN Print: 0976-6367 and ISSN Online: 0976–6375
DOI: https://doi.org/10.17605/OSF.IO/GD7YH
© IAEME Publication
SENTIMENT ANALYSIS APPROACH IN
NATURAL LANGUAGE PROCESSING FOR
DATA EXTRACTION
Nikhil Vijayrao Khander1 and Dr. Jitendra Sheethlani2
1
Research Scholar, Computer Application/Science Department, SSSUTMS,
Sehore, Madhya Pradesh, India
2
Professor, Computer Application/Science Department, SSSUTMS,
Sehore, Madhya Pradesh, India
ABSTRACT
The study of sentiment analysis and opinion mining examines how people's opinions,
sentiments, assessments, attitudes, and emotions are expressed in written language. In
addition to being heavily researched in data mining, web mining, and text mining, it is
one of the most active research fields in natural language processing. Applications for
sentiment analysis include analysing the effects of events in social networks and
examining consumer views of goods and services. With the expansion of social media,
including reviews, forum conversations, blogs, microblogs, Twitter, and social
networks, sentiment analysis is becoming more and more important. For measuring
sentiments with a large volume of opinionated data captured in digital form for analysis,
techniques like supervised machine learning and lexical-based approaches are
available.
Key words: Sentiment Analysis (SA), Natural Language Processing (NLP), Opinion
Mining, Training Set, Classifier.
Cite this Article: Nikhil Vijayrao Khander and Jitendra Sheethlani, Sentiment Analysis
Approach in Natural Language Processing for Data Extraction, International Journal of
Computer Engineering and Technology 13(2), 2022, pp. 27-32.
https://iaeme.com/Home/issue/IJCET?Volume=13&Issue=2
1. INTRODUCTION
Sentiment Analysis (SA), referred to as temper extraction [1], is a blooming hobby region as
an software of Natural Language processing (NLP). Mood Extraction automates the choice
making executed through human. It additionally classifies the polarity of textual content in
phrases of superb, poor or neutral (surprise). Based on polarity, a schooling set is ready and in
addition classifier is applied to categorise the opinions as superb or poor. Social community
revolution performs a decisive function in accumulating statistics containing public opinion. To
acquire subjective and real reaction from the collected statistics, public evaluations are

Sentiment Analysis Approach in Natural Language Processing for Data Extraction
extracted through functions extractor [20]. Figure 1 indicates the format of predicting the hidden
statistics with regards touser‟s intensions, likeliness and taste.
Figure 1 Layout for Sentiment
Analysis of Data for Social Networks
For this purpose sentiment analysis follows various approaches as discussed below:
Subjective lexicon: It unearths nature of phrase on the idea of a rating assigned to every phrase
in a listing of words. Nature of phrase can • be positive, bad or objective. The primary strategies
used for producing a subjective lexicon are gadget translation, the use of phrase internet or Bi-
lingual dictionaries [2] [15].
Using N-Gram modeling: An N-Gram version (unigram, bi-gram, tri-gram or mixture of these)
is designed for categorization on the idea of precise education facts. This version is skilled via
way of means of the use of the education facts and then, its performance is examined the use of
the examined facts
Machine learning: By extracting the characteristics from the text and learning the model,
supervised or semi-supervised learning [13] is carried out. The idea of machine learning [3][10]
can be implemented using a variety of built-in tools, such as Naive Bayes and Support Vector
Machine [17].
Subjectivity analysis is another name for sentiment analysis. It is the computer analysis of
the emotions, opinions, and sentiments expressed in text, such as blogs, newspaper editorials,
product, movie, and book reviews, etc.
2. SENTIMENTANALYSIS
Sentiment evaluation classifies the polarity of a given textual content of the file, sentence or
element degree expressing the opinion as high-quality, poor or neutral [5]. The sentiment
evaluation may be carried out at one of the following levels: • Document-Level Sentiment
Classification: In file degree sentiment evaluation foremost venture is to extract informative
textual content for inferring sentiment of the complete file [6]. Two foremost tactics for file-
degree sentiment evaluation consist of supervised getting to know and unsupervised getting to
know. The supervised method assumes that there's a finite set of instructions into which the file
is assessed and schooling facts to be had for every class. The most effective case consists of
instructions viz. high-quality and poor. Unsupervised tactics are primarily based totally on
figuring out the Semantic Orientation (SO) of precise terms in the file for file-degree sentiment

Nikhil Vijayrao Khander and Jitendra Sheethlani
evaluation. If the common SO of those terms is above a few predefined threshold then the file
is assessed as high-quality and in any other case it's far deemed poor [7].
• Sentiment classification at the sentence level is a more fine-grained level than sentiment
classification at the document level, and it allows for the classification of sentences into
three categories: positive, negative, and neutral. Different procedures are used to handle
various sentence kinds. Conditional sentences, question phrases, and sarcastic
statements are examples of sentences that require special approaches [6]. Sarcasm
mostly occurs in political circumstances and is very difficult to spot [7].
• Aspect-Based Sentiment Analysis: The two methods mentioned above can be used with
either the entire document or a single sentence. Entities frequently have a large number
of aspects (attributes), and each aspect has a unique opinion.
• Aspect-based sentiment analysis, also called feature-based sentiment analysis, focuses
on the recognition of all sentiment expressions within a given document and the aspects
to which they refer [7].
3. APPLICATIONS ARE AS OF SENTIMENT ANALYSIS
In order to ascertain the subjective nature of the data, sentiment analysis or opinion mining is
primarily used. The following are the domains where sentiment analysis is applied:
• Assist with decision-making: Making decisions is a crucial aspect of starting over. It
includes decisions like "which car to buy," "which cafe to attend," and "which tourist
attraction to see." Sentiment Analysis analyses the reviews left by previous buyers of a
particular product, and it then gives the user the best possible response [16].
• Improving Product Quality: There are numerous manufacturing companies for every
product, which creates fierce rivalry. Sentiment analysis is used by businesses to analyse
products more effectively. Customer feedback and comments are used to raise the
calibre of the product. Additionally, this idea inspires the creation of novel items [11].
• Recommendation Systems: It is furnished to the customers for imparting their views.
This machine additionally affords the improvement of a awesome corpus. There are
several web sites with an in-constructed advice machine. These kinds of web sites are
typically associated with the books, music, on-line media, and movie industry.
Recommendation machine additionally continues a few crucial facts of person like
private facts likes and dislikes preceding records and his friend’s facts to offer greater
suggestions [12].
• Business Strategies: Developing a method for commercial enterprise isn't always the
paintings of an individual, however a group paintings. This group consists of the better
authorities, experts, developers, junior body of workers and the maximum crucial is the
clients. Now, the difficulty arises, the way to speak with the clients for his or her
assistance. Sentiment evaluation used the reaction of the clients, their wishes and needs
to generate a destiny method and cowl the preceding flaws.
• Business Intelligence: Sentiment analysis is used to search the web for opinions and
reviews of these opinions from different Blogs, Amazon, tweets, etc. It also helps in
Brand analysis or competitive intelligence, new product perception, product and service
benchmark and market forecasting[12].
• Political SA: It has several packages and opportunities viz. studying trends, figuring out
ideological bias, focused on marketing and marketing or messages, gauging reactions,
etc. It is likewise beneficial in assessment of public critiques and perspectives or
discussions of policy.

• SA and Sociology: Idea propagation via corporations is anvitalidea in sociology.
Opinions and reactions to thoughts are applicable to adoption of recent thoughts and
studying sentiment reactions on blogs can supply perception to this manner e.g.
modelling agree with and have an effect on withinside the blogosphere the usage of
hyperlink polarity [18].
• SA and Psychology: It has potential to augment psychological investigations or
experiments with data extracted from natural an guage text.
4. CHALLENGES TO SENTIMENT ANALYSIS
• The computer examination of affect, opinions, and sentiments conveyed in text, such as
blogs, editorials, newspaper articles, and reviews of goods, TV shows, and movies is
known as sentiment analysis. The following are typical difficulties in sentiment analysis
research:
• Noise (abbreviations, slang): The amount of noise on the internet is always growing.
People frequently employ acronyms, slang, and emoticons for simplicity of usage, but
these can add to the complexity of language processing. For instance: tour is awesome.
Spelling variants result from mistakes of this nature. Awesome, for instance, can be
found as awesm, asumm, or alum.
• Unstructured Data: There is a lot of unstructured data on the web. Different forms of
the same entity can represent it [21]. Web documents, journals, books, health records,
industry-specific internal files, company logs, multimedia platforms, texts, videos,
audios, and photographs are just a few examples of the sources available online. As a
result, the complexity is increased by the variety of data sources and file types [18].
• Contextual Information: Actual sense of the text varies from domain to domain; this
property is referred as contextual property. So, based on the context, the behaviour of
the word changes.
• Word Sense Disambiguation: One word may have multiple meanings. This concept also
affects the polarity of the word. For example-In English word “good” have multiple
senses according to the usage in a particular sentence[15].
• Language Constructs: Different styles in a language lead to different challenges. Some
of the challenges while dealing with English language are as under:
• Word order: for figuring out the subjective nature of the text, preparations of phrases in
a sentence play an essential role. In English language, there's a hard and fast order set
via way of means of grammatical guidelines i.e. concern is accompanied via way of
means of verb that is similarly accompanied via way of means of object. This idea is
defined via way of means of examples in Table.
• Morphological Variations: The idea of morphological variables states that facts is fused
withinside the phrases. Table 1 suggests the verb „ate‟ which contains an awful lot facts
a side from simply the basis word. For example: Smith ate apple.

Nikhil Vijayrao Khander and Jitendra Sheethlani
Thus with the variations in the root word, there can be many words.
• Handling Spelling Variations: As in Punjabi language, one word can possess many
spellings, so this lead to high complexity. It becomes complex to process all the variants
a single word. This problem is also faced during training the model.
• Lack of resources: Lack of tools, resources, corpora lead to great struggle while doing
sentiment analysis for Indian languages.
5. CONCLUSION
Sentiment analysis can be used to examine opinions in blogs, articles, product reviews, social
media websites, and movie review websites where the viewpoints are expressed by a third party.
It is a crucial field to research and has various applications. It has significant commercial value
because businesses are interested in how their goods are viewed and because it allows potential
customers to learn more about the opinions of current users. For effective analysis, many feature
types and classification methods can be combined. The paper also discusses challenges and
applications of sentiment analysis. The field of sentiment analysis will become more popular
by offering adequate solutions to these problems.
REFERENCES
[1] “CaseStudy:AdvancedSentimentAnalysis”.Retrieved18October2013.
[2] BingLiu,“SentimentAnalysisandSubjectivity”.HandbookofNaturalLanguageProcessing,Se
condEdition,2010.
[3] “Sentiment Analysis on Reddit”. Retrieved10 October2014.
[4] G.Vinodhini,RM.Chandrasekaran,“SentimentAnalysisandOpinionMining:ASurvey”,Inter
national Journal of Advanced Research in Computer Science and Software
Engineering,2(6), June2012.
[5] B.Arti, Chandak M.B. and Z. Akshay,“Opinion Mining and Analysis: A Survey”,
International Journal on Natural Language Computing(IJNLC),2(3),June2013.
[6] Jagtap V.S .and Pawar Karishma,“AnalysisofdifferentapproachestoSentence-
LevelSentimentClassification”,InternationalJournal of Scientific Engineering and
Technology, 2(3),pp.164-170,April2013.
[7] F.Ronen, “Techniques and Applications for
SentimentAnalysis”,CommunicationsoftheACM,56(4),pp.82-89,April2013.
[8] P.RudyandTh.Mike,“Sentimentanalysis:Acombinedapproach”,InternationalJournalofInfor
matics,3,pp.143–157,2009.
[9] H.Taneja,S.DhuriaandK.Sukhija“NaturalLanguageProcessing:ABackboneforComputation
alLinguistics”, DHE Sponsored National Conference on Computational Sanskrit–Issues
andChallenges,pp.187-190,Dec.2013.
[10] ShabinaDhuriaandHarmunishTaneja,“ASurveyonSentimentAnalysisandOpinionMining”,I
nternationalConferenceonComputingSciences,pp.124-128,Nov.2013,Elsevier.
[11] K.R.Chowdhary,“NaturalLanguageProcessing”,M.B.M.EngineeringCollege,Jodhpur,India
April29,2012.
[12] XiaoyongLiu,“NaturalLanguageProcessing”,SchoolofInformationStudiesatSyracuseUnive
rsity.
[13] Pang B, Lee L and Vaithyanathan S. Thumbs up? Sentiment Classification using machine
learning techniques. In: Proceedings of EMNLP-02, 7th Conference on Empirical Methods

in Natural Language Processing, Philadelphia, PA, Association for Computational
Linguistics,pp.79–86,2002.
[14] BangBandLeeL,“Opinionminingandsentimentanalysis:FoundationsandTrendsinInformatio
nRetrieval”,pp.1–135,2008.
[15] MelvilleandWojciechGryc,“SentimentAnalysisofBlogsbyCombiningLexicalKnowledgewi
thTextClassification”2009.
[16] ZhaiZhongwu,LiuBing,XuHuaandXuHua,“ClusteringProductFeaturesforOpinionMining.”
WSDM‟11,Feb.2011.
[17] ChaovalitandLinaZhou,“MovieReviewMining:aComparisonbetweenSupervisedandUnsup
ervisedClassificationApproaches”,In:Proceedingsofthe38thHawaiiInternationalConference
onSystemSciences,2005.
[18] MikhailBautin,LohitVijayarenuandStevenSkiena,“Sentimentanalysisfornewsandblogs”,In:
ProceedingsoftheInternationalConferenceonWeblogs and Social Media (ICWSM),2008.
[19] ShabinaDhuriaandHarmunishTaneja,“OntologyEquippedNaturalLanguageProcessingforR
ealWorldApplications”,InternationalJournalofAdvanced
ResearchinComputerScienceandSoftwareEngineering,4(4),2014
[20] Pushpa R. Suri and Harmunish Taneja, “Web Objects Clustering Through Aggregation for
Enhanced Search Results”, International Journal of Scientific & Engineering Research,
2(8), August 2011
[21] Harmunish Taneja, “Object Oriented Information Computing over WWW”, International
Journal of Computer Science Issues, 7(3), May2010

SENTIMENT ANALYSIS APPROACH IN NATURAL LANGUAGE PROCESSING FOR DATA EXTRACTION

More Related Content

Similar to SENTIMENT ANALYSIS APPROACH IN NATURAL LANGUAGE PROCESSING FOR DATA EXTRACTION (20)

More from IAEME Publication (20)

Recently uploaded (20)

SENTIMENT ANALYSIS APPROACH IN NATURAL LANGUAGE PROCESSING FOR DATA EXTRACTION