Abstract: Sentiment analysis is an important research domain of affective computing,and researchers focus on the context in this field. Because the thinking mode influences the formation of language, this paper researched and compared the Chinese thinking modes and western thinking modes. I}he former contains "spiral graphic mode"，"concreteness"and "scattered view", and the latter owns "straight line mode"，"abstractness" and "focus view". A position-related method was proposed to quantify the "spiral graphic mode" and "straight line mode"，and then according to the "concreteness" and "abstractness",an approach was implemented based on part of speech and grammar. Finally a view-window method was applied to analyze the "scattered view" and "focus view". The experimental results show that the performance of sentiment analysis including the thinking mode factors is better than that not including.

Key words: Sentiment analysis,Chinese and western thinking mode, View-window