Development and Analysis of a Methodology for Automatic Generation of a Knowledge Base with Stable Templates of Communication in Healthcare

The language and cultural barriers between doctor and patient can be the cause of misunderstandings, and as a result deterioration in the quality of medical services. Even though there are systems for supporting doctor-patient interaction as well as various translators, such as Google Translator, the existing solutions either do not specialize in the medical field or are significantly limited in capabilities, provide questionable accuracy and are unreliable. Therefore, the health care industry needs a system that will deliver more comprehensive and qualitative results. Accordingly, the purpose of the research is generation of a knowledge base built upon textual medical templates and investigation of its ability to obtain the basis of a medical support system. During the process of developing a knowledge base, the method of comparison analysis is applied to study the means of the audio data processing. Thus, it were received the highest quality text transcripts of dialogues between patient and doctor. Furthermore, applying the Inter Systems platform methods, namely analytic and rule-based, the pipeline of linguistic processing and preparation of the medical knowledge base is implemented. Its purpose is to extract information from text templates. At the final stage, it was obtain graphs based on the medical terms. The obtained results justify the hypothesis that the possibility of implementing the given approach as a foundation of the support system, this achievement will expand the scope of the research. Consequently, this will provide and opportunity to proceed to the development and exploration of the solution that can solve the problems of the linguistic and cultural barriers in the doctor-patient interaction.