Sentiment Analysis: Linguistic Potential of Preprocessing Regimentation

Abstract

The article deals with the sentiment analysis regimentation as a relevant direction in automated natural language processing and its linguistic potential. Despite its impressive practical significance, the sentiment analysis still lacks reliable theoretical foundation. Although information technologies develop very fast, their fundamental foundations correlate with the linguistic system of knowledge. In fact, the methodological priority of the applied linguistics has no alternative with regard to the interdisciplinary specificity of the modern communication. The complex nature of this research made the authors appeal to the computer linguistics in order to provide a meta-description on the algorithmization and modeling of sentiment evaluation. The effectiveness of the relevant practice was conditioned by the optimal configuration of the procedure and an appropriate material evaluation. The preprocessing included identifying the meta-structure, defining its referentiality and level orientation, and choosing the analysis model. The authors described these main steps of the preprocessing algorithm, as well as the relevant practice. The study contributes to productive theoretical optimization of text sentiment analysis. In a broad context, the expedient disclosure of linguistic potential is relevant to the whole sphere of automated natural language processing.

Keywords

• sentiment analysis
• automated processing
• natural language
• linguistic specificity
• regimentation
• preprocessing
• model
• algorithm