Speaker:         Radosław Piotr Katarzyniak
Wrocław University of Technology and Science, Poland

Title:               Cognitive-oriented approach to linguistic summaries production in ambient intelligence and smart environments

Linguistic summaries are a well known concept of advanced data analysis, related mainly to the classic knowledge management field. From the theoretical point of view linguistic summaries constitute a subset of a broader class of natural controlled languages. Many of these languages have already been effectively used by human end users in various practical contexts. Customer relationship management is one of the most interesting practical examples. It is interesting to note that for the latest years linguistic summaries have become an important tool for another important class of applications: ambient intelligence and smart environments. Obviously, it has mainly resulted from extended popularity of Big Data and Internet of Things.

The new contexts in which linguistic summaries are used are, however, substantially different to those in which they used to be appllied in the past. The main difference results from a higher level of required functional autonomy of computational constituting ambient intelligence and smart environments. In particular, in previous solutions a linguistic summary production was usually launched and controlled by a human end user, herself. In modern contexts this functionality needs to be assigned to an artificial computational component. In consequence, existing models of classic linguistic summaries need to be tailored to new functional requirements.

Making artificial components to imitate human like language production, even for simple controlled languages, is not an easy task. It is almost always related to an original and case specific implementation of advanced concepts of cognitive linguistics in complex technical systems. The required autonomous production of linguistic summaries is a prominent example of such situation. Some initial steps to an effective integration of basic cognitive concepts with models of linguistic summaries have already been made by some prominent authors, when they applied a new concept of linguistic protoform to their research and development projects. However, the issue appears more complex and calls for further technical application of another advanced theoretical concepts such as cognitive semantics, as well as for better mapping of natural cognitive systems organization onto computational components design. The lecture is dedicated to outline an original cognitive-oriented approach to effective implementation of human – like production of linguistic summaries in modern computational systems.