Paper Title
A Survey on Emotion Detection in Social Media through Data Mining Techniques

The phenomenon that human beings go through in daily life due to changes happening in and around the environment like happiness, sadness, love, fear, anger, or hatred is called an emotion. Emoticons are used to depict a human emotion, and then used in sentiment analysis and other NLP task. In this paper, we have attempted to prove that while emoticons have a deep and strong effect in conveying a user emotions ,still the usage of all the emoticons aren’t clear. Emoticons are being used in order to convey an emotion via a virtual expression that any person wants to express. To understand the use such an application, there has to be an accurate emotion recognition system in real time. Here, an emotion detector is used for social media application that can run on high performance and it depends upon the content that has been shared by the user on his social media account. Through this paper we will analyze that there are four types of emoticons : Positive, Negative, Neutral and Objects. Thus, any sentiment polarity algorithm can be used with utmost care so that it can be decided which emoticons are dependent and which are not. During the analysis clustering of various emoticons and its K-means ratio will be analyzed and a fresh database will be prepared. First the data set is taken and analyzed on the basis of frequency of emoticons occurred on social media to demonstrate the prevalence. Analysis of the relationship between emoticons and sentiments are carried out. The result affirms that few emoticons are valid for expressing the sentiments by displaying the category of emoji it falls into, taking benefit of them in any sentiment analysis. With the help of emoticon detector it would become easy for nontechnical people and senior citizens to understand the use of each and every emoji during their textual conversation. Keywords - Emoticons, Emotion, Positive, Negative, Neutral, Objects, Hierarchical Clustering, K-Means, Cuckoo Search, Machine Learning, Mobile Applications, Psychological Behavior.