A Hybrid Approach for Emotion Extraction
Journal Title: International Journal for Research in Applied Science and Engineering Technology (IJRASET) - Year 2017, Vol 5, Issue 4
Abstract
With the growth of internet the amount of content that gets produced everyday has increased significantly. Many different types of textual/image documents are produced on varying types of platforms such as social media, news, blogs etc. This paper presents a hybrid approach for emotion extraction from the text documents. With such large amounts of textual content being produced every day, one cannot ignore the importance of mining and extracting opinions, sentiments and also emotions that are embedded in the text. The task of opinion mining, extraction and sentiment analysis has been extensively used in many fields. This is because emotions form an important part of any communication and opinion and have profound effect on the sentiments that the text represents. This makes emotion recognition or extraction from text as well as other media of communication such as speech images videos a crucial task. Emotion analysis is the task of making the computers makes sense of emotions. Emotion extraction is focused on delivery of the exact emotional state of a user at the time when the document was being prepared. In this paper, the main focus of the proposed work is to improve emotion extraction and the study of different approaches and methods that have already been developed and deployed for use. A number of approaches have been developed for emotion extraction and each of them has some strengths and weaknesses. Many machine learning techniques have been used such as Support Vector Machines, Naive Bayes classifier, Vector Space Model and others. Many others combine machine learning techniques with other rule-based techniques or keyword-based techniques. Our approach combines the SVM with k-NN algorithm while still employing the rule based technique to take the advantage of all while avoiding their weaknesses. Our algorithm will be implemented and evaluated using a standard benchmark GRIMM’s data set. The new approach will be evaluated on the basis of accuracy of the emotions that are assigned to the sentences in a document.
Authors and Affiliations
Mr. Ahefaz Pothiyawala, Mrs. Kavita R. Singh
Keywords
Related Articles
- EP ID EP23751
- DOI http://doi.org/10.22214/ijraset.2017.4044
- Views 357
- Downloads 7
Review of Railway Track Fault Finding System
In India so many transport facilities is available. In which railway played vital role to transport of Indian population. The Indian population is on second position in world. So, transportation problem is reduced becau...
Restructuring of Indian Power System with Micro Grid
Indian power system is a vertical power system, where power generating station is at the top and customers are at bottom. Most of the electricity produced today is generated in large generating stations, which is then t...
Factors Influencing Voice Onset Time (VOT): Voice Recognition
In this paper we study the methods to evaluate the features which might estimate the closeness of acoustic signals. For measuring Voice onset time, we were considered 3 stop consonants in word- initial position and two...
E-Commerce: Merchandise Suggestion using Microblogging Information of Consumer
The Web makes superb open doors for organizations to give customized online administrations to their clients. Recommender frameworks intend to consequently create customized recommendations of items administrations to c...
Optimization of Naïve Bayes Data Mining Classification Algorithm
As a probability-based statistical classification method, the Naïve Bayesian classifier has gained wide popularity; however, the performance of Naive Bayes classification algorithm suffers in the domains (data set) that...