Smart Prediction Guide Using Sentiment Analysis

Project Title

Smart Prediction Guide Using Sentiment Analysis

Project Area of Specialization Software EngineeringProject Summary

Everyday a lot of data is generated on different social media platforms that contains the sentiments and opinions of people .it is very difficult to analyze that data manually and to make conclusion based on that raw data. So, there is a need of an automated system that can read the sentiments of users and give us ratings. Our Sentiment analysis web app will allow users to post reviews and stores them to rate movies based on user sentiments. The system will analyze this data to check for user sentiments associated with each comment. This application will partition the user comments to check for sentimental keywords and predicts user sentiment associated with it. Once the keywords are found it associates the comment with a sentiment rank. The system will then gather all comments for a particular movie and then calculates an average ranking to score it.  

Project Objectives

An important part of this informative era has been to find out the opinion of other people. Through sentiment analysis we can collect rich data on attitude and opinion in real time, without compromising reliability, validity, and generalizability. Our aim is to gather feedback on attitudes and opinion as they occur without having to invest in lengthy and costly market research activities by using sentiment analysis. Given a text containing multiple features and varied opinions, our objective is to extract expression of the opinion describing a target feature and classify it as positive, negative or neutral. Our objective is to use sentiment analysis for evaluating attitudes and opinion on movie rating websites using machine learning.

Project Implementation Method

Methodology

We developed sentiment analysis system using the standard machine learning approach.      

IMDB Review Dataset:

We obtained dataset of 50,000 reviews from KAGGLE website Dataset has 50000 rows and 2 column ,First column showing the review and the next column showing the sentiment.

Preprocessing

Before applying any machine learning algorithm to our data, we will first preprocess it in order to clean it. Following are some preprocessing steps:

1.Removing Html Tags

2.Removing special characters

3.Converting everything to lowercase

4.Removing stop words

5.Steming the words

Creating Bag Of Words (BOW)

1.Tokenize the text.

2.Identify Unique words.

3.count occurrence of each unique word in the text.

Training and Test Split:

we will split our data into two sets for training and testing. We will train our model on 80% of available data and the rest of the data will be used for testing.

Model Selection:

Following are some models that we will use for training.

Naïve Bayes

K- Nearest Neighbour

Random Forest

SVM

Model taining:

We will train all these models on available datasets and will check their performance using different performance metrics.

Performance Metrics:

We will compare the results of different models on the basis of some performance metrics such as: Precision, Recall, F-score, and Accuracy. Other performance metrics for multiclass classification include average measures such as micro, macro and weighted f1 score can also be used.

Website Development:

Till here model development has completed, next step in our project is to make a website and deploying model on website for prediction. First step is to develop front end of website using html, css, javascript and other web technologies. Next step is to develop backend using flask and django framework.

Benefits of the Project

Sentiment analysis has endless application and can be applied to any industry, from finance and retail to hospitality and technology. Following are some of important application of sentiment analysis in business:

1. Social Media Monitoring
2. Brand Monitoring
3. Voice of customer (VoC)
4. Customer Service
5. Market Research

Technical Details of Final Deliverable

Our final deliverable will contain an application that will perform sentiment analysis of movie rating websites. It will allow viewers to post comments on different movies and it will automatically analyze sentiments of viewers and assign rating based on viewer’s sentiments as it is easy to analyze a movie through a rating rather than reading comments and then deciding what kind of movie do viewers like the most.

Final Deliverable of the Project Software SystemCore Industry EducationOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	70000
RAM 16GB (KINGSTONE) 3200MHZ	Equipment	2	10000	20000
SSD Adata XPG SX8200 Pro 512GB PCIE GEN3X4 M.2 2280 Solid State	Equipment	2	20000	40000
Battery latitude e5270 47 wh	Equipment	1	10000	10000