Research On Conceptual Modeling Method Based On FCA Theory

Conceptual Modeling is an important part of the development of requirements engineering and the most important predecessor of software engineering.With the increasing complexity of information systems,developers are required to correctly describe the knowledge and rules in the relevant fields to build conceptual models based on the sufficient understanding of functional requirements and application contexts.However,the state of concept evolves continuously during the formalization process,which makes the final conceptual modeling results different from the original knowledge understanding.At the same time,the increasing demand for information reuse and interoperability has led to an increase in the logic complexity of information systems,which has increased the difficulty of domain knowledge analysis and conceptual modeling for developers.Therefore,there is an urgent need for the support of efficient automated conceptual modeling methods to help developers,domain experts and even end users clearly express the abstract concepts and domain knowledge of information systems.The role of conceptual modeling is to abstract the real-world before the implementation of the information system,which is the bridge between the real-world and the development of information systems.Researchers have conducted extensive research on conceptual modeling and proposed a variety of theoretical analysis and conceptual modeling methods.However,the existing methods still have certain deficiencies in practice,and the outstanding performances are as follows:(1)Existing methods lack effective processing of large-scale data and uncertainty problems in complex environments.Existing methods less consider the influence of uncertain factors in the Order Relations metric,so the measurement results will be too strict and lose generalization ability.(2)Existing methods lack effective treatment of inconsistencies.Existing methods usually require domain experts to participate in the modeling process,but there is no complete interactive iterative framework support,and the resulting inconsistency is difficult to eliminate;(3)Existing methods rely too much on the domain knowledge of the designer,and the knowledge acquisition is less automated,and the validity of the conceptual model cannot be automatically verified.In addition,the existing methods may cause poor modeling results due to many uncertain factors(such as implicit information and fuzzy information).In order to make up for these shortcomings and improve the effectiveness of knowledge discovery,this paper conducts a conceptual modeling method based on Formal Concept Analysis(FCA)related theory for the above problems.The specific research work is as follows.1.Concept Lattice-based Relation Granulation and algebraic structure are proposed to provide theoretical support for knowledge reduction in conceptual modeling.Binary Relations in the real-world has more general Binary Relations than just Equivalence Relations,Tolerance Relations,or Order Relations.In this case,this paper proposes a new Rough Set Theory(RST)model from the perspective of sub-relationship based on the concept lattice suitable for analyzing and dealing with binary relations.In this model,a general binary relationship can be decomposed into several sub-relationships,which can be used as Granules to study algebraic structures.This algebraic structure organizes all relational granulation results in the form of concept lattices.In addition,the introduction of Granular Computing(GrC)to further transform complex information systems into relatively simple computational processes is a new attempt and exploration of concept lattice and RST,GrC fusion,and also from the perspective of relationship granularity.The extended form conceptual analysis theory provides a theoretically supported algebraic structure for conceptual modeling.2.This paper propose a fusion theory of concept lattice and inclusion degree and its application method.Quantitative analysis of Uncertain Ordered Relations under the framework of fusion of concept lattice and inclusion degree,and relevant the results are further applied to conceptual modeling of information systems.The core purpose of the fusion theory is to highlight a certain degree of certainty in the order relationship.From the perspective of concrete realization,the research scope of the order relationship has undergone a process of expansion to convergence.Some order relations first extend to fuzzy relations or uncertain relationships,and then the threshold parameters can be set to make the fuzzy or uncertain relationship converge to a certain degree of certainty.Obviously,by appropriately extending the research scope of the order relationship,it can flexibly meet the actual needs,and also help to improve the robustness and generalization ability of the knowledge acquisition model.In addition,based on fusion theory,this approach also provides solutions for Algebraic Structure,Reduction,Core,and Dependency issues in ordered information systems.3.This paper proposes a method for maintaining the consistency of conceptual models based on Constrained Lattices.Aiming at the problem of knowledge representation gap in the conceptual modeling process based on FCA theory,a formal model integrating expert knowledge into concept lattice structure is proposed.First,the original lattice is revised by aligning a set of attribute dependencies with a series of implications provided by the concept lattice.Then,the change trajectory is provided by establishing a constraint lattice using an Extent Projection.On this basis,a modeling method based on FCA constraint lattice theory to compensate for this gap is proposed to maintain the consistency of the conceptual model.This method not only provides a way for domain experts to revise the conceptual model,but also retains the original lattice and the final the trajectory of the change between the constraints.Through these changes,experts can access how concepts in practice are associated with the concept of automatic data publishing.Finally,the effectiveness of the consistency model based on constraint lattice maintenance concept model is verified by an example.4.This paper proposes an application framework of extended FCA theory in conceptual modeling practice.Taking structured data sets as an example,through FCA and related fusion theory analysis,it effectively discovers the concept of knowledge completion in data set.This paper proposes that the practice framework is different from previous research.First,although most of the research focuses on the similarity of concepts,this paper proposes a comprehensive framework that provides a structured and systematic description of the conception process.Secondly,based on the FCA of related fusion theory,the paper studies how to solve the problem of implicit and fuzzy information.Third,most existing studies only discuss the one-to-one mapping between concepts,but the proposed method can deal with more complex problems,that is,an ontology concept may have multiple mapping relationships with another ontology.Finally,this paper applies the proposed technology to the data set samples for verification,and studies how the proposed method can effectively improve data interoperability.The experimental results show that the technique described in this paper provides an effective mechanism to coordinate the data of the scattered source and support the development of ontology concepts.Compared with the knowledge engineering method,the vocabulary accuracy and classification accuracy have been substantially improved,and the basic knowledge structure of the field can be better adapted and respected.In summary,this paper proposes a formal modeling method for conceptual modeling based on FCA.New methods and theoretical tools are provided to improve the validity and consistency of information system conceptual modeling,key concepts such as concept lattice and inclusion fusion theory.