Semantic Appraisal Resources in the J.R. Martin and P.R.R. White Theory and its Use in the Computational Approach to Sentiment Analysis: The Written Tweets in the Saudi colloquial as a Model

Abstract

Objectives: The study aimed to present the semantic appraisal resources in the theory of J.R. Martin and P.R.R. White, and focus on the irrealis variable for discovering challenges in the computer approach to evaluating opinions and sentiment to improve it.

Methods: The study employed the descriptive approach in presenting the conceptual apparatus of semantic Appraisal resources, and applying one of the variables related to irrealis affect on written tweets in the Saudi spoken Arabic.

Results: The results showed that relying on linguistic analysis and its hypotheses is a basis for describing the semantic resources and possibilities of appraisal more accurately. The classification of emotions into realis/irrealis that was mentioned in the field of emotion according to Martin and White reclassified the polarity of emotions (desires - hopes - fears). The results indicated that the computational approach faces challenges in the Arab model, which are evident in ontologies that need to feed concepts with cultural details and multiple local expressions to analyze emotions accurately.

Conclusions: This study examined the potential of functional systemic linguistics in identifying assessment resources that support the analysis of a speaker's views, feelings and attitudes in social practices, a cumulative culture that makes the linguistic sign increasingly reflective in the use of both aspects of language: connotations and procedures, which is difficult for the machine to contain.