Sentiment Clustering with Topic and Temporal Information from Large Email Dataset

Sentiment analysis with features addition to opinion words has been an appealing area in recent studies. Some research has been conducted for finding relationship between sentiments, topics and temporal sentiment analysis. Nevertheless, Email sentiment analysis received relatively less attention due to the complexity of its structure and indirectness of its language. This paper introduces a systematic framework for sentiment clustering using topic and temporal features for large Email datasets. Interesting Email and sentiment distribution patterns are summarized and discussed with empirical results.