Sentiment Analysis of Flipkart Product Reviews using Natural

International Journal of Recent Technology and Engineering (IJRTE)

ISSN: 2277-3878 (Online), Volume-12 Issue-2, July 2023

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

Sentiment Analysis of Flipkart Product Reviews

using Natural Language Processing

S Kiruthika, U

Sneha Dharshini, K R Vaishnavi, R V Vishwa Priya

Abstract: In this contemporary world, people depend more on

ecommerce sites or applications to purchase items on-line. People

purchase items on-line based upon the scores and evaluates

offered by individuals that purchased items previously which

identifies the success or failing of the item. Furthermore, business

suppliers or manufacturers identify the success or failing of their

item by evaluating the evaluates offered by the clients. In current

system, a number of techniques were utilized to evaluate a dataset

of item evaluates. It likewise provided belief category formulas to

use a monitored discovering of the item evaluates situated in 2

various datasets. The proposed speculative methods examined the

precision of all belief category formulas, and ways to identify

which formula is more precise. Additionally, the existing system

unable to spot phony favorable evaluates and phony negative

reviews with discovery procedures. One of the most popular works

was done "Bad" and "Outstanding" seed words are utilized by

him to determine the semantic positioning, factor smart shared

info technique is utilized to determine the semantic positioning.

The belief positioning of a file was determined as the typical

semantic positioning of all such expressions. Semantic Positioning

of context independent viewpoints is identified and the context

reliant viewpoints utilizing linguistic guidelines to infer

positioning of context unique reliant viewpoint are thought about.

Contextual info from various other evaluates that discuss the exact

same item function to identify the context indistinct-dependent

viewpoints were drawn out.

Keywords: Semantic positioning, linguistic guidelines, context

indistinct-dependent viewpoints.

I. INTRODUCTION

Device Discovering is stated as a subset of synthetic

knowledge that's primarily interested in the advancement of

formulas which permit a computer system to gain from the

information and previous experiences by themselves. The call

artificial intelligence was initially presented by Arthur

Samuel in 1959.

Manuscript received on 18 May 2023 | Revised Manuscript

received on 29 May 2023 | Manuscript Accepted on 15 July 2023

| Manuscript published on 30 July 2023.

*Correspondence Author(s)

S Kiruthika, Department of Computer Science and Engineering, Sri

Krishna College of Technology, Coimbatore, (Tamil Nadu), India. E-mail:

kiruthika.s@skct.edu.in, ORCID ID: 0000-0003-1946-473X

U Sneha Dharshini, Department of Computer Science and Engineering,

Sri Krishna College of Technology, Coimbatore, (Tamil Nadu), India. E

mail:

19tucs223@skct.edu.in, ORCID ID: 0009-0006-2139-2455

K R Vaishnavi*, Department of Computer Science and Engineering, Sri

Krishna College of Technology, Coimbatore, (Tamil Nadu), India. E-

mail:

19tucs246@skct.edu.in, ORCID ID: 0009-0003-7063-9408

R V Vishwa Priya, Department of Computer Science and Engineering,

Sri Krishna College of Technology, Coimbatore, (Tamil Nadu), India. E

mail:

19tucs258@skct.edu.in, ORCID ID: 0009-0002-8081-

2756

Sciences Publication (BEIESP). This is an open access article under the CC

BY-NC-ND license http://creativecommons.org/licenses/by-nc-nd/4.0/

With the assistance of example historic information, which

is called educating information, artificial intelligence

formulas develop a mathematical design that assists in

production forecasts or choices without being clearly

configured. Artificial intelligence brings computer system

scientific research and stats with each other for producing

anticipating designs. Artificial intelligence constructs or

utilizes the formulas that gain from historic information. The

more the info, the greater will be the efficiency. Classic

artificial intelligence is frequently classified by how a

formula learns to ended up being more precise in its forecasts.

There are 4 fundamental methods: monitored discovering,

without supervision discovering, semi-supervised

discovering and support discovering. The kind of formula

information researchers decides to utilize depends upon what

kind of information they wish to anticipate. Monitored

discovering, likewise called monitored artificial intelligence,

is specified by its use labelled datasets to educate formulas

that to categorize information or anticipate results precisely.

As input information is fed into the design, it changes it

weights up till the design have been equipped properly. This

happens as section of the go across recognition procedure to

guarantee that the design prevents overfitting or underfitting.

Monitored discovering assists companies refix for a range of

real- world issues at range, such as categorizing spam in a

different folder from your inbox. Some techniques utilized in

monitored discovering consist of neural networks, naïve

bayes, direct regression, logistic regression, arbitrary

woodland, assistance vector device (SVM), and more.

Without supervision discovering, likewise called without

supervision artificial intelligence, utilizes artificial

intelligence formulas to evaluate and collection unlabeled

datasets. These formulas find concealed patterns or

information groupings without the required for human

treatment. Its capability to find resemblances and distinctions

in info make it the suitable service for exploratory

information evaluation, cross-selling techniques, client

segmentation, picture and pattern acknowledgment. It is

likewise utilized to decrease the variety of functions in a

design with the procedure of dimensionality reduction;

primary element evaluation (PCA) and singular worth

decomposition (SVD) are 2 typical methods for this. Various

other formulas utilized in without supervision discovering

consist of neural networks, k-means clustering, probabilistic

clustering techniques, and more. Semi-supervised

discovering provides a pleased tool in between monitored and

without supervision discovering. Throughout educating, it

utilizes a smaller sized labelled information readied to direct

category and function removal from a bigger, unlabeled

information establish.

Sentiment Analysis of Flipkart Product Reviews using Natural Language Processing

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

Semi-supervised discovering can refix the issue of having

actually insufficient labelled information (or otherwise

having the ability to pay for to tag sufficient information) to

educate a monitored discovering formula. Support

discovering deals with a feedback-based procedure, where an

AI representative (A software application element)

immediately check out its bordering by hitting & path, acting,

discovering from experiences, and enhancing its

performance. Representative obtains awarded for every great

activity and obtain penalized for each poor action; thus, the

objective of support discovering representative is to optimize

the benefits. In support discovering, there's no labelled

information like monitored discovering, and representatives

gain from their experiences just. The support discovering

procedure resembles an individual being; for instance, a kid

learns various things by experiences in his daily life. An

instance of reinforcement learning is to play a video game,

where the Video game is the atmosphere, relocations of an

representative at each action specify specifies, and the

objective of the representative is to obtain a high rack up.

Representative gets comments in regards to penalty and

benefits. Belief evaluation (or viewpoint mining) is an all-

natural language refining (NLP) method utilized to identify

whether information is favorable, unfavorable or neutral.

Belief evaluation is frequently carried out on textual

information to assist companies check brand name and item

belief in client comments, and know client requirements.

Belief evaluation is the procedure of spotting favorable or

unfavorable belief in text. It is frequently utilized by

companies to spot belief in social information, evaluate brand

name credibility, and know clients. Since people reveal their

ideas and sensations more freely compared to before, belief

evaluation is quick ending up being an important device to

check and know belief in all kinds of information.

Immediately evaluating client comments, such as viewpoints

in study reactions and social networks discussions, enables

brand names to discover what makes clients pleased or

annoyed, to ensure that they can customize services and

products to satisfy their customers' requirements. All-natural

language refining (NLP) describes the branch of computer

system scientific research and more particularly, the branch

of synthetic knowledge or AI interested in providing

computer systems the capability to know text and talked

words in a lot similarly humans can. NLP integrates

computational linguistics—rule-based modelling of human

language with analytical, artificial intelligence, and deep

discovering designs. With each other, these innovations allow

computer systems to procedure human language through text

or articulate information and to ‘understand' its complete

implying, total with the audio speaker or writer's intent and

belief. NLP owns computer system programs that equate text

from one language to another, react to talked commands, and

summarize big quantities of text quickly also in actual time.

There is a likelihood you have communicated with NLP

through voice-operated GPS systems, electronic aides,

speech-to-text dictation software application, customer

support chatbots, and various other customer benefits.

However, NLP likewise plays an expanding function in

business services that assistance improve company

procedures, enhance worker efficiency, and streamline

objective crucial company procedures.

II. LITERATURE REVIEW

P. Kalaivani and N. L. Shanmuganathan et alia.,[1] has

suggested that big quantities of information are offered in the

internet. This paper research researches on-line item

evaluates utilizing belief evaluating methods. Particularly,

the paper contrasts 3 monitored artificial intelligence

approaches, Naive Bayes and KNN for Belief Category of

Evaluates. Empirical outcomes specifies that SVM method

surpassed the Naive Bayes and KNN methods, and the

educating dataset had a a great deal of evaluates, SVM

method got to accuracies of a minimum of 80%. The

objective of examine is to assess the efficiency for belief

category in regards to precision, accuracy and remember in

this examine, in this paper, we contrasted 3 monitored

artificial intelligence formulas of SVM, Naive Bayes and

KNN for belief category of the item evaluates which contain

1000 favorable and 1000 unfavorable refined evaluates. The

speculative outcomes reveal that the SVM method surpassed

compared to the Naive Bayes and KNN methods and the

educating dataset had a a great deal of evaluates, the SVM

method got to accuracies of greater than 80%.

Bilal Sabari and Saidah Saad et alia.,[2] has suggested

Viewpoint Mining (OM) or Belief Evaluation (SA) can be

specified as the job of spotting, drawing out and categorizing

viewpoints on something. The procedure of info removal is

extremely important since it's an extremely helpful method

however likewise a difficult job. That imply, to essence belief

from an item in the web-wide, have to automate opinion-

mining systems to do it. The current methods for belief

evaluation consist of artificial intelligence (monitored and

unsupervised), and lexical-based methods. Thus, the primary

objective of this paper provides a study of belief evaluation

(SA) and viewpoint mining (OM) methods, different methods

utilized that belong in this area.

Vishal S. Shirsat and Sachin N. Deshmukh et alia.,[3] has

suggested that Belief Evaluation and Viewpoint Mining is a

many prominent areas to evaluate and discover

understandings from text information from different

resources like Twitter and google, Twitter, and Amazon.com,

and so on. It plays an important function in allowing business

to work proactively on enhancing business technique and

acquire an in- deepness understanding of the buyer's

comments regarding their item. It includes computational

examine of habits of a private in regards to his purchasing rate

of passion and after that mining his viewpoints regarding a

company's company entity. This entity can be visualized as

an occasion, private, article or item experience. In this paper,

Dataset has drawn from Amazon.com which includes

evaluates of Video cam, Laptop computers, Smartphones,

tablet computers, TVs, video clip monitoring. After

preprocessing we used artificial intelligence formulas to

categorize evaluates that declare or unfavorable. This paper

wraps up that, Device Discovering Methods provides finest

lead to categorize the Items Evaluates. Naïve Bayes obtained

precision 98.17% and Assistance Vector device obtained

precision 93.54% for Video cam Evaluates.

International Journal of Recent Technology and Engineering (IJRTE)

ISSN: 2277-3878 (Online), Volume-12 Issue-2, July 2023

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

Sundus Hassan and Muhammad Rafi et alia.,[4] has

suggested that the task of identifying of files inning

accordance with their web content is called text classification.

Lots of experiments have been brought bent on improve text

classification by including history understanding to the file

utilizing understanding repositories like Word Web, Open up

Job Directory site (OPD), Wikipedia and Wikitology. The

arise from the previous paper plainly suggest Wikitology is

much much far better compared to various other

understanding bases. This paper contrasts Assistance Vector

Device (SVM) and Naive Bayes (NB) classifiers under text

enrichment with Wikitology. The validated outcomes with

10-fold go across recognition and revealed that NB provides

an enhancement of +28.78%, on the various other hand SVM

provides an enhancement of +6.36% when compared to

standard outcomes. Naive Bayes classifier is much far better

option when outside enriching is utilized with any type of

outside data base.

G. S. Brar and A Sharma et alia.,[5] has suggested that Belief

Evaluation is a brand-new topic in Research study and works

in lots of various other areas. In Contemporary World, A big

quantity of textual information is gathered utilizing studies,

remarks, and evaluates over the internet. All the gathered

information is utilized to enhance services and products

offered by both personal companies and federal governments

worldwide. This Paper consists of belief evaluation of item

evaluates utilizing feature-based viewpoint mining and

monitored artificial intelligence. In this paper, the primary

concentrate is to identify the polarity of evaluates utilizing

nouns, verbs, and adjectives as viewpoint words. Evaluates

will be Categorized into 2 various classifications favorable

and unfavorable. Evaluates of Open up Item Data source is

utilized as resource information establish and All-natural

Language Refining Toolkit for Section of Speech Tagging.

Nishajebaseeli and Kirubhakaran Ezra et alia.,[6] has

suggested that the web ends up being an important location

for trading concepts, on-line discovering, evaluates for a

services or product or items. It makes difficult to document

and know the individual feeling since evaluates over the web

are offered for millions for a services or product. Belief

evaluation is an arising for research study to gather the

subjective info in resource product by using All-natural

Language refining, Computational Linguistics and text

analytics and classified the polarity of the viewpoint or belief.

This paper offers a general study regarding belief evaluation

or viewpoint mining relates to item evaluates. In this literary

works study it's seen that for choice production procedure

regarding item, solution, item, social problems, belief

evaluation or viewpoint mining play extremely important

function.

Bing Liu and Junsheng Cheng et alia.,[7] has suggested that

The Internet has ended up being an outstanding resource for

collecting customer viewpoints. There are currently various

Website including such viewpoints, e.g., client evaluates of

items, online discussion forums, conversation teams, and

blog sites. This paper concentrates on on-line client evaluates

of items. It makes 2 payments. Initially, it suggests an unique

structure for evaluating and contrasting customer viewpoints

of contending items. A model system called Viewpoint

Observer is likewise executed. The system is such that with a

solitary glimpse of its visualization, the individual has the

ability to plainly see the staminas and weak points of each

item psychological of customers in regards to different item

functions. This contrast works to both prospective clients and

item producers. For a prospective client, he/she can see an

aesthetic side-by-side and feature-by function contrast of

customer viewpoints on these items, which assists him/her to

choose which item to purchase. For an item producer, the

contrast allows it to quickly collect advertising knowledge

and item benchmarking info. 2nd, a brand-new method based

upon language pattern mining is suggested to essence item

functions from Pros and Disadvantages in a specific kind of

evaluates. Such functions develop the basis for the over

contrast. Speculative outcomes reveal that the method is

extremely efficient.

Ahmad Abdel-Hafez and Yue Xu et alia.,[8] has suggested

that with the extensive of social networks sites in the web,

and the big variety of individuals taking part and producing

unlimited variety of components in these sites, the require for

customization enhances significantly to ended up being a

requirement. Among the significant problems in

customization is constructing users' accounts, which depend

upon lots of elements; such as the utilized information, the

application domain name they objective to offer, the

depiction technique and the building approach. Just lately,

this of research study was a concentrate for lots of scientists,

and thus, the suggested techniques are enhancing really

rapidly. This study objectives to review the offered individual

modelling methods for social networks sites, and to

emphasize the weak point and stamina of these techniques

and to offer a vision for future operate in individual modelling

in social networks sites.

III. EXISTING SYSTEM

In current system, a number of techniques were utilized to

evaluate a dataset of item evaluates. This paper likewise

provided belief category formulas to use a monitored

discovering of the item evaluates situated in 2 various

datasets. Our speculative methods examined the precision of

all belief category formulas, and ways to identify which

formula is more precise. Additionally, the system was unable

to spot phony favorable evaluates and phony unfavorable

evaluates with discovery procedures. In the current system

investigates in the document-based viewpoint mining are

discussed listed below. One of the most popular works was

done "Bad" and "Outstanding" seed words are utilized to

determine the semantic positioning, factor smart shared info

technique is utilized to determine the semantic positioning.

The belief positioning of a file was determined as the typical

semantic positioning of all such expressions. Semantic

Positioning of context independent viewpoints is identified

and the context reliant opinions using linguistic guidelines to

infer positioning of context unique reliant viewpoint are

thought about. Contextual info from various other evaluates

that discuss the exact same item function to identify the

context indistinct-dependent viewpoints were drawn out.

Sentiment Analysis of Flipkart Product Reviews using Natural Language Processing

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

Drawbacks of Existing System

• The current system to identify the success or failing

of the item would certainly be based upon the scores

provided by the clients to the item.

• In this, people might unknown regarding the stopped

working functions of the item and the context where

people offer the scores.

• It ends up being challenging for business

neighborhoods to create the item or work after the

stopped working functions to conquer the loss, if the

item is a failing on the market.

• Moreover, the current system utilizes monitored

discovering where the information is qualified to

anticipate the result.

IV. PROPOSED SYSTEM

In suggested system, the paper wish to prolong this examine

to utilize various other datasets such as Amazon.com dataset

or eBay dataset and utilize various function choice

techniques. Additionally, this paper might use belief category

formulas to spot phony evaluates utilizing different devices

NLP Methods after that we'll assess the efficiency of our deal

with a few of these devices. The without supervision

dictionary-based method is utilized in this system. WorldNet

is utilized as a thesaurus to identify the viewpoint words and

their basic synonyms and antonyms. The suggested work is

carefully relating to the Mining and Summarizing Client

Evaluates. Provided the summary of the suggested system

‘Document centered Belief Positioning System'. Individual

and critic evaluates of the items were gathered and used as an

input to the system. The system classifies each file as

favorable, unfavorable and neutral and provides the overall

variety of favorable, unfavorable and neutral variety of files

individually in the outcome. The outcome produced by the

system useful for the individuals in choice production, they

can quickly determine the number of favorable and

unfavorable files exist. The polarity of the provided files is

identified on the basis of most of viewpoint words.

4.1 Advantages of Proposed System

• The belief evaluation of the item can be precisely

evaluated.

• The suggested system utilizes without supervision

discovering technique where the device immediately

learns from the information provided as input.

• It assists in identifying the success or failing of the

items in business domain name.

• It assists in enhancing business by fixing the stopped

working functions in the item utilizing feature-based

belief evaluation.

Without supervision discovering is a artificial intelligence

standard for issues where the offered information includes

unlabeled instances, implying that each information factor

includes functions (covariates) just, without an connected tag.

The objective of without supervision discovering formulas is

discovering helpful patterns or architectural residential or

commercial homes of the information. Instances of without

supervision discovering jobs are clustering, measurement

decrease, and thickness estimation.

The wish of without supervision discovering is that with

mimicry, which is an essential setting of discovering in

people, the artificial intelligence formula is qualified to

develop a small interior depiction of the information. When it

comes to a generative job, such depiction can work as well as

required for the formula to produce creative web content from

it. As opposed to monitored discovering where information is

tagged (labelled) by a professional, e.g., as a "sphere" or

"fish", without supervision techniques exhibition self-

organization that catches patterns as possibility densities or a

mix of neural function choices. Various other discovering

standards in the guidance range are support discovering

where the device is provided just a numerical efficiency rack

up as assistance, and semi-supervised discovering where a

smaller sized part of the information is labelled.

4.2 Parts-Of-Speech (Pos) Tagging

A POS label (or part-of-speech label) is a unique tag

designated to every token (word) in a message corpus to

suggest the section of speech and frequently likewise various

other grammatic classifications such as tense, number

(plural/singular), situation and so on. POS tags are utilized in

corpus searches and in text evaluation devices and formulas.

A collection of all POS tags utilized in a corpus is called a

tagset. Tagsets for various languages are generally various.

They can be totally various for unrelated languages and really

comparable for comparable languages, however this is not

constantly the guideline. Tagsets can likewise most likely to

a various degree of information. Fundamental tagsets might

just consist of tags for one of the most typical components of

speech (N for noun, V for verb, A for adjective and so on.).

It's, nevertheless, more typical to enter into more information

and compare nouns in singular and plural, spoken

conjugations, tenses, element, articulate and a lot more.

Private scientists may also establish their very own really

specific tagsets to fit their research study requirements.

POS tags make it feasible for automated text refining devices

to consider which section of speech each word is. This helps

with using linguistic requirements along with stats. POS tags

are likewise utilized to browse for instances of grammatic or

lexical patterns without specifying a concrete word. POS

tagging is frequently likewise described as annotation or POS

annotation.

V. PROPOSED METHODOLOGY

Item evaluates are the viewpoints or feedbacks of clients for

a specific item. Lots of on-line companies set up an

evaluation area on their site to permit clients to price and

evaluate the item they bought. An item evaluate assists

various other individuals obtain a remove concept of the item

previously buying it. The customers can check out the

evaluates and make their mind remove, and choose whether

the item deserves buying or otherwise. If the eCommerce

sites have not included an item evaluate area on your

eCommerce site, just due to being afraid unfavorable

evaluates, the business owners are losing a big variety of

prospective clients.

International Journal of Recent Technology and Engineering (IJRTE)

ISSN: 2277-3878 (Online), Volume-12 Issue-2, July 2023

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

Item evaluates are probably one of the most helpful methods

to remove shoppers' issues concerning an item. A bulk of

individuals are affected by item evaluates in their buy.

Regardless of whether you're a prominent brand name or

simply began, item evaluates do play an essential function in

your eCommerce company as for reliability is worried.

Reliability is just one of the essential aspects that choose the

success of your brand name over time. Vendors frequently

neglect the significance of item evaluates. The significant

concentrate stays on developing the website appearance and

enhancing the check-out web page however absolutely

nothing truly issues if you're not obtaining great evaluates on

your items. The significance of item evaluates can be

comprehended by that 90% of the customers check out on-

line evaluates previously purchasing and 88% of the

customers will be triggered to take an activity after reviewing

favorable evaluates.

Fig 1. Architecture of the proposed system

In the over representation (Fig.1) system streams where the

different components are being sent out to information

refining and by different formulas, we are adjusting the

Device Discovering design and we lastly obtain the efficiency

evaluation.

A. Implementation

This phase has to do with the summary of different

components and their data source and the outcome develop

and their application in our suggested system. The different

components provide in the system are explained listed below.

5.1 DATA PRE-PROCESSING

URL and Hash tags: Because of optimal of 140 characters

restriction of information, the individual share some

associated info on the subject utilizing URL and hashtags.

Tweets including such kind of info should be managed. The

suggested system gets rid of all the URL and hashtags from

the tweets. Reduce situation: Individual tweet might include

top situation and reduce situation or might be provide

implying to words if utilized unevenly. To decrease the

uncertainty, the suggested system additional procedures the

information by transforming all the tweets to reduce situation

letters. Determining punctuations: Punctuations and white

areas are determined and gotten rid of to prevent repetitive

functions and various other disputes.

5.2 DATA EXTRACTION

There are different mining methods utilized for information

extractions; it can be either file degree, expression degree or

sentence degree. Throughout information removal, the device

utilized can be of monitored or without supervision method.

Monitored technique consists of artificial intelligence

methods like Naïve Bayes (NB), Optimal Entropy (ME), and

Assistance Vector Devices (SVM).

5.3 STOP WORDS

There are many words which are basically removed by the

tokenizers which does not add any meaning to the sentences.

These words which don't include any type of implying to the

sentence are called stop words and should be eliminated from

the sentences for effective natural language processing. These

are typically utilized words like prepositions, conjunctions,

articles and pronouns from any language which is removed in

natural language processing.

5.4 STEMMING

Stemming is a method utilized to essence the base develops

of words by eliminating affixes from them. It's much like

reducing down the branches of a tree to its stems. Online

search engines utilize stemming for indexing words. That is

why instead of keeping all types of a word, an online search

engine can keep just the stems. By doing this, stemming

decreases the dimension of the index and enhances retrieval

precision.

Sentiment Analysis of Flipkart Product Reviews using Natural Language Processing

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

5.5 POS TAGGING

This procedure tags words in accordance with components of

speech utilizing NLP. POS tagging classifications are

linguistically made by splitting the sentence grammatically

and each word in the sentence are tagged with the correct

parts of the speech like noun, adjective, verb, preposition and

so forth which gives the correct meaning of the sentences in

natural language processing.

5.6 DATA COMPARISON

Outcome of POS tagging is compared to unfavorable and

favorable dataset. If favorable worths surpass after that great

or else poor. If both favorable and unfavorable worths are

exact same intermediate is showed.

5.7 DATABASE DESIGN

The data source develop includes development of tables that

are stood for in physical data source as kept data. They have

their very own presence. Each table make up of rows and

columns where each paddle can be deemed document that

includes associated info and column can be deemed area of

information of exact same kind. The table is likewise

developed with some setting can have a null worth. without

redundancy and with normalized style.

VI. EXPERIMENTAL RESULTS

Fig 2: Enter the URL

International Journal of Recent Technology and Engineering (IJRTE)

ISSN: 2277-3878 (Online), Volume-12 Issue-2, July 2023

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

Fig.3: Click on Get Content

The above figure (Fig.2) shows the page in which the review URL has to be entered and Get Content button has to clicked as

shown in the figure (Fig.3)

Fig 4: Click on Stop words and Stemmer Button

Sentiment Analysis of Flipkart Product Reviews using Natural Language Processing

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

Fig 5: Final Result

The above figure (Fig.5) shows the number of positive and negative words the overall result.

VII. CONCLUSION AND FUTURE ENHANCEMENT

In this research study work, the paper provides Viewpoint

Mine, a structure which carries out probabilistic rational

thinking, for viewpoint mining, in issues relates to social

networks. The proposed system uses the structure on Twitter

information, providing a situation where the proposed work

wishes to anticipate whether an individual is meant to go to

Crete or otherwise with apparent applications for travel

bureau and in all domain names of the tourist market. Belief

Evaluation and Entity Acknowledgment techniques have

been put on automate essential jobs such as the development

of arbitrary variables, the development of guidelines and the

derivation of the proof establish. These jobs are the

fundamental functions of a probabilistic visual design like a

BN. After the conclusion of these automated jobs by the

proposed structure, it continues to the educating of the design

by utilizing formulas of ProbLog. Afterward, new Tweets can

be categorized inning accordance with the preferred result,

i.e., whether the individuals will go to Crete with some

possibility. The assessment of the obtained design is based

upon metrics that has any type of regression design. More

particularly, this paper utilize origin imply settle mistake,

imply outright mistake and imply squared mistake to

determine the typical of the mistakes of the designs obtained

by structure. The obtained metrics permit us in conclusion

that the obtained designs by the structure have great

efficiency. An essential function of the structure is its

capability to be adjusted really quickly to lots of subjects in

social networks to carry out viewpoint mining. Twitter was

utilized as instance for our examination; however, the

proposed method and structure can be likewise utilized for

other social media network such Twitter and google,

Instagram and so on. Additionally, the guidelines of the

obtained design are built in an effective method and

immediately. Lastly, the proposed structure sustains step-by-

step discovering so the obtained design can be enhanced.

DECLARATION

Funding/ Grants/

Financial Support

No, we did not receive.

Conflicts of Interest/

Competing Interests

No conflicts of interest to the

best of our knowledge.

Ethical Approval and

Consent to Participate

No, the article does not require

ethical approval and consent to

participate with evidence

Availability of Data and

Material/ Data Access

Statement

Not relevant

Authors Contributions

All authors having equal

contribution for this article.

International Journal of Recent Technology and Engineering (IJRTE)

ISSN: 2277-3878 (Online), Volume-12 Issue-2, July 2023

Published By:

Blue Eyes Intelligence Engineering

and Sciences Publication (BEIESP)

Retrieval Number: 100.1/ijrte.B77740712223

DOI: 10.35940/ijrte.B7774.0712223

Journal Website: www.ijrte.org

REFERENCES

1. P. Kalaivani and N.L. Shanmuganathan, “Sentiment classification of

movie reviews by supervised machine learning approaches” in

Artificial intelligence, vol.4, nos.4, Sept.2013, ISSN:0976-

5166.

2. Saidah Saah and Bilal Saberi, “Sentiment analysis or Opinion mining:

A review” in University Kebangsaan Malaysia, vol.7, nos.5,

Halifax, Malaysia, October2017, doi:10.18517, ISSN:2088-5334.

[CrossRef]

3. Rajkumar S Jagdale, Vishal S. Shirsath, A. S. and Sachin Deshmukh,

“Sentiment Analysis on product reviews using machine learning

techniques” in book: Cognitive Informatics and Soft Computing,

pp.639-647, January 2019, doi:10.1007/978-981-13-0617-4_61.

[CrossRef]

4. Sundus Hassan, Muhammad Rafi and M. Shahid Shaikh, “Comparing

SVM and Naïve Bayes classifiers for text categorization with

wikitology as knowledge enrichment” in Habib University, February

2012, doi:10.1109/INMIC.2011.6151495. [CrossRef]

5. Nisha Jebaseeli A and E. Kirubakaran, “A Survey on Sentiment

Analysis of Product Reviews”, in International Journal of Computer

Applications, pp. 0975- 888, vol. 47, no.11, June 2012, doi:

10.1016/j.eswa.2020.113265. [CrossRef]

6. Bing Liu, Minqing Hu and Junsheng Cheng, “Opinion Observer:

Analyzing and

Comparing

opinion

the

web”,

the

International

conference,

pp.575,

May 2005, doi:

10.1145/1060745.1060797. [CrossRef]

7. Ahmad Abdel-Hafez and Yue Xu, “A Survey of user modelling in

social media websites”, of Computer and information science in the

Queensland university of Technology, vol.06, no.04, pp: 59-71,

September 2017, doi:10.5539/cis. v6n4p59. [CrossRef]

8. Z. Kastrati, F. Dalipi, A. S. Imran, K. P. Nuci, and M. A. Wani,

``Sentiment analysis of students' feedback with NLP and deep

learning: A systematic mapping study,'' Appl. Sci., vol. 11, no. 9, p.

3986, Apr. 2021, doi:10.3390/app11093986. [CrossRef]

AUTHOR PROFILE

S. Kiruthika, received the B.E in Computer Science and

Engineering from Sri Eshwar college of

Engineering,Coimbatore,Tamil Nadu,India in the year

2014 and received the M.E in Sri Krishna College of

Engineering and Technology ,Coimbatore,Tamil

Nadu,India in the year 2016.Her area of interest is Data

Science.

U. Sneha Dharshini is currently a final year student from

Sri Krishna College of Technology, Coimbatore, Tamil

Nadu, India, pursuing a bachelor’s degree in Computer

Science and Engineering and expected to be graduated in the

year 2023. Her research interests include machine learning,

artificial intelligence and networking security.

K. R.Vaishnavi is currently a final year student from Sri

Krishna College of Technology, Coimbatore, Tamil Nadu,

India, pursuing a bachelor’s degree in Computer Science

and Engineering and expected to be graduated in the year

2023. Her research interests include data science, machine

learning and big data analytics.

R.V.Vishwa Priya is currently a final year student from

Sri Krishna College of Technology, Coimbatore, Tamil

Nadu, India, pursuing a bachelor’s degree in Computer

Science and Engineering and expected to be graduated in

the year 2023. Her research interests include machine

learning, artificial intelligence and image processing.

Disclaimer/Publisher’s Note: The statements, opinions and

data contained in all publications are solely those of the

individual author(s) and contributor(s) and not of the Blue

Eyes Intelligence Engineering and Sciences Publication

(BEIESP)/ journal and/or the editor(s). The Blue Eyes

Intelligence Engineering and Sciences Publication (BEIESP)

and/or the editor(s) disclaim responsibility for any injury to

people or property resulting from any ideas, methods,

instructions or products referred to in the content.