SOCIAL NETWORK BOTS IN NATURAL COMMUNICATION: PROBLEMS OF DIAGNOSTICS

The article deals with the issue of automated devices presence in natural dialogue as new, robotic participants of modern public communication. The possibility of studying speech communication employing bots on the basis of the methodology of the theory of linguistic personality is considered. Within the specified communicative conditions it is proposed to use the term social network bot - i.e. automated software connected to the platform, which is used by bots to interact with real users. There are difficulties in identifying a robotic user in terms of its similarity to man and its imitation of natural human speech. The ways of creating conflict-prone content by robotic users are identified. A set of linguistic features, applicable to characterizing a hypothetical social network bot, is proposed. . The analysis of bot personotexts has allowed to establish specific traits of social network bots' accounts, organization of their text messages, their communicative actions, regularly used language means. It proves the possibility of establishing stable, regular, superpersonal features that allow to diagnose the traits of artificial bot-communication. The conclusion is drawn on the impossibility to study the signs of speech communication of social network bots as opposed to the human factor in the language, because it is the features of natural human speech that are modelled and simulated by modern software devices.