Automatic conversion of communication media

This thesis explores the feasibility of creating a personal communication support system capable of making automatic decisions within user-defined limits regarding communication terminal selection and medium conversion. This system alleviates the need for users to anticipate every potential communication scenario, allowing them to concentrate on their primary communication goals and preferences while the system manages the details. Conceptually, it functions as an 'electronic secretary.' A prototype has been developed that evaluates potential outcomes of media conversions and automatically recommends the most appropriate conversion and delivery method based on user reachability, technical quality, and content semantics. The work introduces a new model for media handling and conversion, discussing the theoretical impacts of media conversion on information semantics. It also proposes a generic model for telecommunication services and employs a novel approach to assess the quality of concatenated conversion modules. Adapted decision-making methods and algorithms for selecting and configuring software and hardware modules are developed, supporting complex media conversions. The system features a generic interface for integrating third-party products like Text-to-Speech and Speech Recognition, facilitating further enhancements. It prioritizes both technical quality and content semantics for evaluation, adheres to the Reference Mode