Issuu

International Research Journal of Engineering and Technology (IRJET)

e-ISSN: 2395-0056

Volume: 11 Issue: 10 | Oct 2024

p-ISSN: 2395-0072

www.irjet.net

Analysis of movies Reviews using Seq2Seq Pattern in Different Machine Learning Pradeep 1, Prof. Priyanka Bhatele2, Prof. Ajit Kumar Shrivastava3 M.Tech Scholar, Department of CSE, SISTec-R, Bhopal (M.P), India1 Assistant Professor, Department of CSE, SISTec-R, Bhopal (M.P), India2 Professor & Head of Department of CSE, SISTec-R, Bhopal (M.P), India2 -------------------------------------------------------------------------***-----------------------------------------------------------------------ABSTRACT Here, Authors did their Experiment and show their Experimental work. In this scenario they choose a dataset of movies where reviews are recorded. When we talked about sentiment Analysis corpora contains different dataset from where we selected movies data. As we know that in natural language processing, we have different categories like positive, negative and neutral. In this dataset we have only 2 categories positive and negative. As data scientist we have to extract the polarity against the given reviews by thousands of users. Here Author's used Seq2Seq and Word2Vector functionality for extraction the real and true values. As we know that for creating a model we need to pick different Machine Learning Algorithms. Here we have taken NaiveBayes with different flavour, Regression and support vector machine. After creating model, we applied different performance parameters and found that Voted Algorithms gives some better result in comparison to others. In terms of Accuracy this gives us nearly 90% of Accuracy.

users. Take an example carwala.com where every user is totally depends of reviews of a given Car or Model. The positive reviews can make positive to buy a car by many customers on daily basis. In movies section one best example is Box-Office which gives us clear cut message of any movie’s impression on initial 2-3 days. By these reviews we come to consideration whether movies are going to direction of Hit or flop. In many bloggings site we see that in day-to-day life at many topics Millions of pass their positive or Negative comments. which can be exploited for emotion. Many times, they contain more factual or relevant information and they are not considered to the desired extent. In day-to-day life we have different scenario that can be solved with the help of following mechanism. Predictive models: In this mechanism given model will predict the result for near future. The model needs to train itself by huge data available for that domain. Then they are able to predict a value for a given new data. The predictive model is very important for upcoming business Analysis.

Keywords: Sentiment Analysis, Plaintext Corpora, Tokens, Features, Word2Vector, Seq2Seq, Sklearn Performance parameters like Accuracy.

libraries,

Descriptive models: This model is using for deep analysis which is used further for doing segmentation-based analysis. Here Relationship is available with different parameters of given domain. By these data we can able to create an equation in mathematical form which gives us consolidate information in deep analysis. It is generally used to help to understand what the system is Right now and what it can do and how it is doing it.

I INTRODUCTION According In Recent Era Internet play vital role in every sector from Education to Entertainment, From Research to Agriculture, from space to sea at every place we found that presence of Internet is available. Day by social media is also playing very important roles in business analysis as well as product analysis. When we start exploring the social media from every corner of the world have some involvement in every Industries or sector. Now social media is replacing the news media as well as print media. when we are observing the social media nowadays reels play a vital role where people show their views in terms of reels, text messages, images or video's. If we observe a big player in E-Commerce like Amazon, Flipkart and many more all totally depends upon user's review, means reviews are driving force today in every business [1].

1.1 Machine Learning Techniques Learning algorithm play important roles in different industries like E-Commerce, Insurance, Education, Engineering etc. In this section we worked for sentiment analysis or finding polarity from movies reviews dataset. We know that we have number of algorithms to solve our problem out of them we are explain some algorithms:

Famous websites like imdb.com, Carwale.com and many more are available with millions of reviews of thousands of

ISO 9001:2008 Certified Journal | Page 66