Abstract

The project has been developed as an emotion-based android application, meant for identifying the feelings, sentiment and attitude of the user. Our app identifies six emotions: happy, sad, angry, bored fear and neutral. The project is based on the principle of identification of human emotions from user’s input text and playing videos that are appropriate for enhancing their current emotional state. The application takes the user’s written statement as input via an android device and analyzes emotions based on it. Once the emotion is detected and classified, a video suitable to the mood of the user will be presented.

Background Theory

The project we have taken refers to the use of text analysis, computational linguistics to systematically identify, extract, and study effective states and subjective information. Our project is mainly applied to classification of emotions. Emotions, in the field of psychology, are defined as a subjective but conscious experience which is primarily comprised of various biological reactions, different mental states, as well as physiological expressions. Emotions can be influenced by both external and internal factors, such as an outward event or personality. Other factors which can affect emotions include physical factors, such as injury, as well as biological factors such as hormones. Hormones, for example, may cause emotions to be more spontaneous, out of control, or intense. Because emotions are subjective, they are sometimes difficult for psychologists to measure; typically, psychologists will use biological and mental factors to help measure emotional states. Our project identifies feelings, emotions of a person or a group of people by analyzing the written statement and various questionnaires of that person or group. After analyzing and identifying, it categorizes these state of emotions of subjected population. It can be viewed as a tool to identify the attitude of the writer as it finds the hidden emotional state of the person who has written the text. So our project polarity of the statement and identifies the positivity or negativity of the particular scenario in which that statement was written and has to be analyzed. It also gives feedback based on that particular statement or scenario. Further, it provides recommendations to the user based on their current mood. Thus we will approach each sentence and extract meaning and match them with pre-trained structures to find how much it matches those pre-trained classification.

Problem statement

Loneliness and anxiety are two of the biggest issues in today’s world. Every day there are more and more people who feel alone. They see it as a problem, but they don’t know how to solve it or live with it without it causing them harm. A lot of times it begins with a vague fear of loneliness. In recent years, cases involving depression and anxiety are very serious. A mental disorder could be life threatening. Thus, we seek to contribute to the public mental health analysis by providing a platform that will identify a person who could potentially suffer from mental disorders. This could also help to identify potential suicidal victims as usually it is due to their mental imbalance.

Objective

• To determine the current mood of the user and provide video recommendations based on that

Scope of Project and Applications

The goal of human emotion recognition is to automatically classify user’s temporal emotional state based on input data and provide necessary recommendations. With the growth of Internet, service providers collect more and more information about their users. Based on these data, content, layout and ads are displayed according to the user’s profile. Adding information about the emotions of users could provide more accurate personality models of the users.

Authors

Aayusha Shrestha, Aakriti Dhakal, Astha Adhikari, Elina Karanjit