RELATIONAL PERSONALIZED DATALOGICAL DATA MODEL FOR DETERMINING THE COMPATIBILITY OF SURVEY DATA USING NEURAL NETWORK MEANS

Abstract

The article presents an approach to determining the compatibility of questionnaire data using neural network-based methods. The method is based on comparing the questionnaire data of clients and candidates through a frame-based representation model and textual analysis of descriptions.

The implementation of the approach was carried out through a marriage agency website, where a personalized relational data model was developed for storing client questionnaires and auxiliary information. This model consists of 23 interconnected tables. The main components of this model include data about clients and their requirements for a potential partner, which are used to find compatible candidatesThe results of the study showed that as the number of candidates in the database increases, the number of compatible pairs also grows. The parallel use of the frame-based model for comparing questionnaires and neural network-based analysis of textual descriptions significantly increased the effectiveness of determining the compatibility of questionnaire data. The developed approach ensures high accuracy in identifying compatible client pairs, enhancing the search process through the use of artificial intelligence technologies.

Furthermore, the application of this method allows the system to process large volumes of data efficiently, providing more reliable and accurate compatibility assessments. By combining both frame-based comparison and neural network analysis, the system can account for not only the structural features of the data but also the semantic content of the descriptions, which is essential for assessing compatibility in a deeper context. This dual approach ensures a more holistic understanding of potential matches, improving the user experience and outcomes in the matchmaking process. The proposed method opens new possibilities for personalized matchmaking systems, offering a scalable solution for web-based applications. Through the integration of AI technologies, this approach could be adapted and expanded for use in various domains where compatibility analysis is needed, making it a versatile tool for future applications in other fields beyond matchmaking.