EmSi-Interpreter: Module for emotion signs interpretation (LLC ImportRus)
Personal Knowledge Base Designer was used when prototyping one of modules of the HR-Robot application [11]. The main purpose of HR-Robot is to support decision-making when selecting candidates for vacancies and checking staff for motivation (research of the psychological situation in the team) on the basis of analyzing emotions. The system consists of following main modules: video processing, emotions signs detection, and emotions signs interpretation. This case deals with the development of a knowledge base prototype for emotions signs interpretation module called «EmSi-Interpreter».
The development process is similar to the Case Study 3.
A detailed description of the application is given in the paper:
Yurin A.Yu., Dorodnykh N.O. Creating Web Decision-Making Modules on the Basis of Decision Tables Transformations // Communications in Computer and Information Science. Modelling and Development of Intelligent Systems (MDIS 2020), 2021, Vol. 1341, P. 167-184. DOI: 10.1007/978-3-030-68527-0_11
The developing a prototype of knowledge bases for a module can be presented in the form of a diagram (Fig.1).
Fig.1 The development scheme for a PHP module using PKBD
Next, let’s consider the steps in brief.
Step 1. Conceptual models describing parts of the face and its main elements, which would be tracked in the process of determining emotions, were created as a domain model. A fragment of one of face models is shown in Fig.2, step 1.
Fig.2 Steps of development of the «EmSi-Interpreter» decision-making module
Step 2. Decision tables describing the structural aspect of the domain (domain models), as well as, the knowledge of psychologists, were developed. These tables contain information about combinations of signs describing emotions, for example, «fear» (Fig.2, step 2). In fact, each row of the table is a logical rule.
Next, we used PKBD, which provided the import of decision tables and their repre-sentation in the form of logical rules. In this example, the knowledge base segment for the «fear» emotion includes: 5 fact templates, 1 rule template, and 11 specific rules.
Step 3. Imported decision tables were refined in the RVML form (Fig.2, step 3).
Step 4. For this segment of the knowledge base, 250 lines of code were generated for PHP, and 453 lines were generated for Drools (Fig.2, step 4).
