Semantic Models of Agglutinative Languages
Dzavdet Sh. Suleimanov, Гильмуллин Ринат Абрекович, A. Ya. Fridman, Damir Mukhamedshin · 2024
This paper considers a decentralized semantic model of agglutinative languages, the Tatar one in particular, as a general means to organize cognitive interactions between any entities: human to human, human to computer and so on. Such an opportunity is justified by the following unique features of the Tatar language: morphological abilities of affixal morphemes, multiple linkage in ellipsis and recursion, etc., which provide ample knowledge processing options aimed at organizing efficient communications. The proposed techniques look promising for building “general common intelligence” and constitute a core part of the planned designing and exploiting an integral multi-level semantically inspired model of agglutinative natural languages within a flexible software system intended for solving various linguistic problems, which arise in the course of cognitive and communicative activities of intelligent entities (they can be both humans and/or computer systems). The developing modeling system significantly differ from other ones in the following aspects: the structure of the language model leans on the semantical, rather than morphological and/or syntactical properties of concepts; the model is managed as a decentralized system by means of some proactive modules, which adhere to the semiotic models by D.A. Pospelov and V.N. Polyakov; communications are created and supported according to the principles of the situational approach in order to assure the relevance and completeness of the data involved; the system functions in the environment of the Tatar language that we consider the most suitable and corresponding to our goals.