Assessment software based on semantic and fuzzy logic

Abstract

This article presents an intelligent system designed for objective and effective assessment of student assignments in the educational process. The proposed approach has the function of automatic processing and text extraction of documents in PDF and DOCX formats, and evaluates the semantic similarity between assignment conditions and student responses through deep learning models based on the Sentence Transformer architecture. In addition to semantic proximity metrics, the system also analyzes multidimensional linguistic features such as content coverage, internal consistency, and structural connectivity. A probabilistic classification mechanism is implemented to determine the probability of content generated by artificial intelligence. A decision-making system based on fuzzy logic theory is used to combine all assessment components into a single final score, which ensures interpretability and transparency of assessments. The result of this research makes a practical contribution to the field of automated educational technologies and creates a reliable basis for supporting decisions in large-scale educational environments.