Main objective of this graduation thesis is clarification of informations about human emontions and its recognition and classification on basis of generally known methods. Among generally known methods belongs classification based on mimic and pantomimic expression basis and on voice tone basis. This graduation thesis is primary focused on human emotions recoginition and classification on text messages analysis. Specifically anger and joy. A Czech database in lemma form was crated for this method and it contains emotion functional words. It was made by translation of English emotion functional words and it is divided to emotion key words, emotion modification words and emotion phrases. Also was made special database of neutral words, which are removed from emotions classification except emotion phrases comparation. A degrees were assigned to emotion key words and phrases and this assignment was made by more people. That guarantee better objectivity of the database. Values of single emotion words and phrases might be influenced by emotion modification words, words that contains gradation, negation or punctuation. Output of this graduation thesis is application in JAVA language. That application after insertion input text message compare single words with all available databases and count resultant emotion rate for joy and anger emotions. Evaluation involve also single words, sentences and paragraphs. Application was designed in NetBeans with Swing GUI Builder.
| Identifer | oai:union.ndltd.org:nusl.cz/oai:invenio.nusl.cz:219189 |
| Date | January 2011 |
| Creators | Buday, Ondřej |
| Contributors | Burget, Radim, Smékal, Zdeněk |
| Publisher | Vysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií |
| Source Sets | Czech ETDs |
| Language | Czech |
| Detected Language | English |
| Type | info:eu-repo/semantics/masterThesis |
| Rights | info:eu-repo/semantics/restrictedAccess |
Page generated in 0.0017 seconds