A Hybrid Framework for Detecting Non-basic Emotions in Text
| Autor: | Abid Hussain Wani, Rana Hashmy |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Disappointment
business.industry Computer science Emotion classification Supervised learning Boredom computer.software_genre Field (computer science) Identification (information) medicine Word2vec Artificial intelligence medicine.symptom business Set (psychology) computer Natural language processing |
| Zdroj: | Communications in Computer and Information Science ISBN: 9789811359910 |
| DOI: | 10.1007/978-981-13-5992-7_14 |
| Popis: | The task of Emotion Detection from Text has received substantial attention in the recent years. Although most of the work in this field has been conducted considering only the basic set of six emotions, yet there are a number of applications wherein the importance of non-basic emotions (like interest, engagement, confusion, frustration, disappointment, boredom, hopefulness, satisfaction) is paramount. A number of applications like student feedback analysis, online forum analysis and product manual evaluation require the identification of non-basic emotions to suggest improvements and enhancements. In this study, we propose a hybrid framework for the detection and classification of such non-basic emotions from text. Our framework principally uses Support Vector Machine to detect non-basic emotions. The emotions which go undetected in supervised learning are attempted to be detected by using the lexical and semantic information from word2vec predictive model. The results obtained utilizing this framework are quite encouraging and comparable to state-of-the-art techniques available. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|