1 Abstract—The present approach deals with the identification of Emotions and classification of Emotional patterns at Phrase-level with respect to Positive and Negative Orientation. The proposed approach considers emotion triggered terms, its co-occurrence terms and also associated sentences for recognizing emotions. The proposed approach uses Part of Speech Tagging and Emotion Actifiers for classification. Here sentence patterns are broken into phrases and Neuro-Fuzzy model is used to classify which results in 16 patterns of emotional phrases. Suitable intensities are assigned for capturing the degree of emotion contents that exist in semantics of patterns. These emotional phrases are assigned weights which supports in deciding the Positive and Negative Orientation of emotions. The approach uses web documents for experimental purpose and the proposed classification approach performs well and achieves good F-Scores. Keywords—Emotions, sentences, phrases, classification, patterns, fuzzy, positive orientation, negative orientation. I. INTRODUCTION MOTIONS have been widely studied in psychology and behavioral sciences, as they are considered as an important element of human nature. It represents the psychological state of a person which is normally based on internal factors such as mental and physical status of a person and external factors say, social sensory feeling [15]. Identifying emotions from natural language texts has drawn the attention of several information processing communities since, it plays a vital role in human intelligence, decision making, social interaction, awareness, learning, creativity, etc. Analysis of the emotional content in text, determines opinions, attitudes, evaluations and inclinations. Also, researchers have focused in the field of human computer interaction namely facial expressions studies, recognition of emotions using sensors, opinion mining and market analysis, etc. In future, human-computer interaction is expected to emphasize the naturalness and effectiveness by integrating the models of human cognitive capabilities that includes emotional analysis and generation. Several efforts have been made by the natural language processing researchers to identify emotion at different level of granularities say word, sentence or document using reviews, news, question answering, information retrieval, etc. Nowadays, news websites and news channels have provided a new service that allows users to express their emotions after browsing news articles. This has focused on recognizing positive and negative orientations of a person with respect to interested articles. For each article, the readers Vadivel Ayyasamy is with the National Institute of Technology Trichy, India (e-mail: vadi@nitt.edu). express their emotions through voting for a set of predefined emotion labels/tags. In general, six emotions such as anger, fear, sad, disgust, happy, surprise [5] can be expressed in human beings and certain eventual situations can be expressed in certain possible manners. In many events, such expressions contain very little or sometimes no affect-related words and simply they describe the experiences which can be deciphered by the audience [6]. Existing models of emotion detection are able to find affect related direct expressions. The use of knowledge-rich dictionaries supports in classification. For instance, I am very happy, I am bit angry, etc., contain words like happy, angry with corresponding affective meaning in dictionaries. Instances containing negations, such as I am not happy conveys the definition of inverse emotions and corresponding rules to pass from one emotion to the other. Certain instances are more difficult to classify by dictionary-based models such as I am celebrating my 25th marriage anniversary, which can be labeled with joy. Such instances would perhaps be classifiable through a supervised system, which would know that the bigram marriage anniversary is associated for the sentences related to joy. A method is proposed by [3] in which the main idea is to obtain the knowledge of emotions that are related to different eventual concepts. In this process, the system learns that this specific bigram marriage anniversary relates to the joy emotion, and also it learns that the eventual concepts related to anniversary/parties/birthdays/marriages are related to emotion joy in general. These approaches solve the problem of indirectly mentioning an emotion by using the eventual concepts that are related to it instead. However, some instances would also fail to classify clearly the emotion expressed in more complex settings, such as If my husband hadn't expired, today we would be celebrating our 25th marriage anniversary, where the inverse emotion works and express the feel of sad. Thus, as it is observed that the presence of concepts in the text cannot be considered as a mark that the respective sentence directly contains that emotion. Thus, the events also play equally important role in emotion classification. The rest of the paper is organized as follows. The related work is presented in the next section and proposed work is presented in Section III. Experimental results are presented in Section IV. The paper is concluded in the last section. II. BACKGROUND Nowadays, research works focus on analyzing online users’ sentiment responses while they are exposed to news articles which are called as social emotions [1]. The first line of Neuro-Fuzzy Based Model for Phrase Level Emotion Understanding Vadivel Ayyasamy E World Academy of Science, Engineering and Technology International Journal of Computer and Information Engineering Vol:11, No:1, 2017 53 International Scholarly and Scientific Research & Innovation 11(1) 2017 scholar.waset.org/1307-6892/10006151 International Science Index, Computer and Information Engineering Vol:11, No:1, 2017 waset.org/Publication/10006151