Abstract:
Document Level Sentiment Classifier is a web based application which is following the concepts of data mining. This application aims to analyze the orientation of online documents i.e. reviews regarding cell phones on the basis of sentiments present in it. User will be able to search his/her desired cell phone which he/she is opting to buy. The searched results will then target various websites to extract and scrap real time data on which further operations (Such as Data cleaning, data storage etc.) are applied. Classification algorithm Naive Bayes is used to categorize data into positive & negative sentiments. Dashboard displays visual representation of the analyzed results. Vision of this project is to develop an application that will facilitate the users in finding opinions and sentiments of people about a cell phone. This can help individual users in making informed decision about buying a cell phone by knowing how that is perceived in the market. The application will also help phone-manufacturers in getting valuable feedback from the market and orientation of the buyers towards their product for further improvement. Our application targets online cellphone reviews to analyze cellphone user's orientation and keeps on updating it regularly. Users can visually see the graphical representation of results enabling them to make wise and correct decision before buying such products. This web-based application is free of cost thereby relieving people from financial burden of expensive licensed applications. Users will just have to make an account once or sign in via Facebook or Gmail. Instead of spending time on going through numerous websites to get an overview of your chosen cellphone, this application will provide you the best solution about the product you are looking for.
