Stanford CoreNLP is the Natural Language Processing (NLP) pipeline that allow analysing text at paragraph, sentence and word levels. Its outcomes can be used for extracting core elements of functional characteristics of the Topological Functioning Model (TFM). The TFM elements form the core of the knowledge model kept in the knowledge base. The knowledge model ought to be the core source for further model transformations up to source code. This paper presents research on main steps of processing Stanford CoreNLP application results to extract actions, objects, results and executors of the functional characteristics. The obtained results illustrate that such processing can be useful, however, requires text with rigour, and even uniform, structure of sentences as well as attention to the possible parsing errors.

Topological Functioning Model specifies functional and structural characteristics of a system in the holistic manner. Cause-and-effect relations link cause and effect functional characteristics of the system, illustrating causality in it. The Unified Modelling Language (UML) provides its own relationship kinds among elements. Traditionally, a use of UML relationships depends on analyst’s experience in UML and knowledge about the system. However, after TFM transformation meaning of cause-and-effect relations in UML model is not always clear. The paper summarizes research results on this matter and provides mapping guidelines from TFM causal relations to often used UML relationships. These guidelines can be applied in further (manual or automated) refinement of UML diagrams.

BPMN (Business Process Model and Notation) models are used to specify business knowledge in the language that is familiar for business people. They consist of multiple process diagrams that highlight different aspects of interaction among participants. Verification of BPMN models is important since graphical fragmentary presentation could be a source of errors such as incompleteness, deadlocks, livelocks, incorrect terminations etc. We consider verification of model completeness. The model is transformed to the topological functioning model (TFM) in order to check completeness of inputs, outputs and functioning cycles of the entire specified system. The proposed approach is dedicated to the verification of the model at the beginning of analysis, and it could be supplemented by other methods at the design stage. This approach is more dedicated to analysis of the whole system, than to the verification of the concrete fragment work.

According to requirements provided by customers, the description of to-be functionality of software systems needs to be provided at the beginning of the software development process. Documentation and functionality of this system can be displayed as the Topological Functioning Model (TFM) in the form of a graph. The TFM must be correctly and traceably validated, according to customer’s requirements and verified, according to TFM construction rules. It is necessary for avoidance of mistakes in the early stage of development. Mistakes are a risk that can bring losses of resources or financial problems. The hypothesis of this research is that the TFM can be validated during this simulation of execution of the UML activity diagram. Cameo Simulation Toolkit from NoMagic is used to supplement UML activity diagram with execution and allows to simulate this execution, providing validation and verification of the diagram. In this research an example of TFM is created from the software system description. The obtained TFM is manually transformed to the UML activity diagram. The execution of actions of UML activity diagrams was manually implemented which allows the automatic simulation of the model. It helps to follow the traceability of objects and check the correctness of relationships between actions.

System behaviour usually is modelled with logical operators and triggering conditions on control flows between processes, activities, tasks, or events. This allows branching control flows in order to increase model comprehensibility. In case of Topological Functioning Model (TFM), where system's functionality is represented by causal relations among functional characteristics, combinations of causes as well as triggered effects may be quite complex. Therefore, specification of them must not decrease apprehensibility of the TFM, while keeping its accuracy, compactness and level of abstraction. Additionally, this specification must also be modifiable and transformable. In this paper we discuss and refine a concept of a cause-and-effect relation and a logical relation in the TFM. Then, we analyze specification means used in BPMN, UML Activity Diagrams, EPCs, flowcharts, Petri Nets and Decision Models and assess which of them are more appropriate for using or integrating with the TFM. The more suitable means will increase the accuracy of specification of logical relations and system behaviour in the TFM. As a result, it would be possible to eliminate human participation in transformations from the TFM to models at the lower level of abstraction.

It is difficult and time-consuming to migrate a legacy system to some new platform or integrate it with other software system manually. High-level abstract models (domain models) of the existing software system must be got for further merging with new domain models. TFM4MDA (Topological Functioning Modeling for Model Driven Architecture) is an approach for software development from the high level of abstraction to the lower levels. The formal TFM (Topological Functioning Model) for software system analysis can be obtained stepwise from the low levels using RE (Reverse Engineering) techniques. The algorithm for transformation from UML sequence diagrams to the TFM is suggested in this research. It is based on the previous research results. Additional information about other approaches such as MDRE (Model-Driven Reverse Engineering) and ADM (Architecture Driven Modernization) is overviewed in order to use it for further analysis and full formalization of the transformation considered in our work.