Sentiment classification, Opinion mining, Product Review

Fuzzy Natural Logic (FNL) is introduced as a model that could be useful in the area of sentiment analysis. FNL is a formal theory of human reasoning that includes mathematical models of the semantics of natural language expressions with regard to the vagueness phenomenon. The most elaborated constituent of FNL is the theory of evaluative linguistic expressions. To capture their semantics, it uses a single scale for computing extension of any evaluative expression that might be relevant for sentiment analysis. Therefore, it provides a more fine-grained classification of opinion and sentiments than dichotomous models which only distinguish between ‘positive’ and ‘negative’ values. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.