Takeo Hayase

A three-view model for developing object-oriented frameworks

This paper describes a three-view model for developing object-oriented frameworks. We propose a new methodology based on this model, and demonstrate its effectiveness using an example of practical industrial applications. This model can lead software engineers to a framework that has high reliability, portability, and maintainability. These quality factors of applications are especially important in a domain that has repeatability and changeability of hardware and software environment. The three-view model consists of a domain analysis view, a layer view, and a mechanism view. The domain analysis view is used to clarify all information and domain knowledge by using a new concept that we call Domain Reference Models (DRM), which reference models for modeling domain-specific objects, so that a framework has high reusability. The layer view is used to divide a framework into three layers that are piled up vertically: an infrastructure layer, a generic layer, and a domain layer. Because software engineers can replace a part of the framework for restriction on implementation, the framework has high portability. The mechanism view is used to decide which mechanism of whitebox frameworks or blackbox frameworks. By using this guideline, the framework has high maintainability. We applied our approach to the framework development for an industrial monitoring applications. By developing a prototype, we have a prospect of withdrawing the development costs of the framework by five or six times of application deployment. In this application, we estimate several ten times of application deployment. Therefore, it is effective for software engineers to develop a framework based on the three-view model.

Data Mining of Generalized Association Rules Using a Method of Partial-Match Retrieval

This paper proposes an efficient method for data mining of generalized association rules on the basis of partial-match retrieval. A generalized association rule is derived from regularities of data patterns, which are found in the database under a given data hierarchy with enough frequencies. The pattern search is a central part of data mining of this type and occupies most of the running time. In this paper, we regard a data pattern as a partial-match query in partial-match retrieval then the pattern search becomes a problem to find partial-match queries of which answers include sufficient number of database records. The proposed method consists of a selective enumeration of candidate queries and an efficient partial-match retrieval using signatures. A signature, which is a bit sequence of fixed length, is associated with data, a record and a query. The answer for a query is fast computed by bit operations among the signatures. The proposed data mining method is realized based on an extended signature method that can deal with a data hierarchy. We also discuss design issues and mathematical properties of the method.

Process Mining Approach Based on Partial Structures of Event Logs and Decision Tree Learning

Process mining techniques are able to improve processes by extracting knowledge from event logs commonly available in todays information systems. In the area, it is important to verify whether business goals can be satisfied. LTL (Linear Temporal Logic) verification is an important means for checking the goals automatically and exhaustively. However, writing formal language like LTL is difficult, and the properties by which the users intentions are not reflected sufficiently have bad influence on the verification results. Therefore, it is needed to help writing correct LTL formula for users who do not have sufficient domain knowledge and knowledge of mathematical logic. We propose an approach for goal achievement prediction based on decision tree learning. It is conducted focusing on partial structures represented as event order relations of each trace. The proposed technique is evaluated on a phone repair process log.

Business Process Verification and Restructuring LTL Formula Based on Machine Learning Approach

It is important to deal with rapidly changing environments (regulations, customer behavior change, and process improvement etc.) to keep achieving business goals. Therefore, verification for business process in various phases are needed to make sure of goal achievements. LTL (Linear Temporal Logic) verification is an important method for checking a specific property to be satisfied with business processes, but correctly writing formal language like LTL is difficult. Lacks of domain knowledge and knowledge of mathematical logics have bad influence on writing LTL formulas. In this paper, we use LTL verification and prediction based on decision tree learning for verification of specific properties. Furthermore, we helps writing properly LTL formula for representing the correct desirable property using decision tree constrction. We conducted a case study for evaluations.