Statistical Analysis of a Japanese Emotion Corpus for Natural Language Processing
In this paper, we build a Japanese emotion corpus and perform statistical analysis on it. We manually entered in about 1,200 example dialogue sentences. We collected statistical information from the corpus to analyze the way emotion is expressed in Japanese dialogue. Such statistics should prove useful for dealing with emotion in natural language. We believe the collected statistics accurately describe emotion in Japanese dialogue.
KeywordsNatural Language Processing Machine Translation Emotion Word Emotion Category Natural Language Processing Technique
Unable to display preview. Download preview PDF.