Ebook: Knowledge-Based Software Engineering
As knowledge-based software engineering matures and increasingly automates the software engineering life cycle, software engineering resources are shifting towards knowledge acquisition and the automated reuse of expert knowledge for developing software artifacts. This book summarizes the work and new research results presented at the Tenth Joint Conference on Knowledge-based Software Engineering (JCKBSE 2012), held on the island of Rhodes, Greece, in August 2012.The biennial Joint Conference on Knowledge-Based Software Engineering brings together researchers and practitioners to share ideas on the foundations, techniques, tools, and applications of knowledge-based software engineering theory and practice.Topics addressed include theoretical foundations, practical techniques, software tools, applications and/or experience reports in knowledge-based software engineering. This book is published in the subseries Knowledge-Based Intelligent Engineering Systems (KBIES).
This volume summarizes the works and new research results presented at the Tenth Joint Conference on Knowledge-based Software Engineering (JCKBSE 2012), which took place in August 23–26, 2012 on the island of Rodos, Greece. JCKBSE is a well established international biennial conference that focuses on the applications of Artificial Intelligence on Software Engineering. The Tenth Joint Conference on Knowledge-based Software Engineering (JCKBSE 2012) was organized by the Department of Informatics of the University of Piraeus and it was the second time that a JCKBSE took place in Greece.
This year, pretty much like every year, the majority of submissions originated from Japan, while Greece was second. The submitted papers were reviewed by at least two reviewers. Finally, 35 papers were accepted for presentation at the JCKBSE2012 and inclusion in its Proceedings. The papers accepted for presentation in JCKBSE2012 address topics such as the following:
Architecture of knowledge-based systems, intelligent agents and softbots
Architectures for knowledge-based shells
Automating software design and synthesis
Decision support methods for software engineering
Development of multi-modal interfaces
Development of user models
Development processes for knowledge-based applications
Empirical/evaluation studies for knowledge-based applications
Intelligent user interfaces and human-machine interaction
Internet-based interactive applications
Knowledge engineering for process management and project management
Knowledge management for business processes, workflows and enterprise modeling
Knowledge technologies for semantic web
Knowledge technologies for service-oriented systems, Internet of services and Internet of things
Knowledge technologies for web services
Knowledge-based methods and tools for software engineering education
Knowledge-based methods and tools for testing, verification and validation, maintenance and evolution
Knowledge-based methods for software metrics
Knowledge-based requirements engineering, domain analysis and modeling
Methodology and tools for knowledge discovery and data mining
Ontologiespatterns in UML modeling
Program understanding, programming knowledge, modeling programs and programmers
Software engineering methods for Intelligent Tutoring Systems
Software life cycle of intelligent interactive systems
Software tools assisting the development
In JCKBSE2012, we had Prof. Nikolaos Bourbakis as Keynote Speaker. Prof. Bourbakis (IEEE Fellow) is an OBR Distinguished Professor of IT and the Director of the Assistive Technologies Research Center (ATRC) at Wright State University, Ohio, USA. He pursues research in Applied AI, Machine Vision, Bioinformatics & Bioengineering, Assistive Technologies, Information Security, and Parallel- Distributed Processing funded by USA and European government and industry. He has published more than 330 articles in refereed International Journals, book-chapters and Conference Proceedings, and 10 books as an author, co-author or editor. He has graduated 17 Ph.Ds and 37 Master students. He is the founder and the EIC of the International Journal on AI Tools, the Editor-in-Charge of a Research Series of Books in AI (WS Publisher), the Founder and General Chair of several International IEEE Computer Society Conferences, Symposia and Workshops, an Associate Editor in several IEEE and International Journals and a Guest Editor in 18 journals special issues. His research work has been internationally recognized and has won several prestigious awards. Some of them are: IBM Author recognition Award 1991, IEEE Computer Society Outstanding Contribution Award 1992, IEEE Outstanding Paper Award ATC 1994, IEEE Computer Society Technical Research Achievement Award 1998, IEEE ICTAI 10 years Research Contribution Award 1999, IEEE Symposium on BIBE Outstanding Leadership Award 2003, ASC Award for Assistive Technology 2005, University of Patras Degree of Recognition 2007. The title of his talk was “An Intelligent Interface for Human-Machine Interaction for the Blind”.
We would like to thank the authors for choosing JCKBSE2012 as the forum for presenting the results of their research. We would also like to thank the reviewers for taking the time to review the submitted papers rigorously. For their efforts in undertaking the organization of JCKBSE2012, we would like to thank Profs. George A. Tsihrintzis and Th. Panagiotopoulos. For putting together the website of JCKBSE2012 and for managing the conference administration system, we would like to thank Mr. Ari Sako. Finally, we would like to thank the IOS Press personnel for their wonderful job in producing this Proceedings.
Maria Virvou and Saeko Matsuura
JCKBSE2012 General and Program co-Chairs
In a requirements elicitation using ontology, the quality of elicited requirements fully depends on the quality of requirements ontology. In this paper, a verification method of the correctness of requirements ontology is proposed in order to improve the quality of requirements ontology. We provide a rule description language to specify properties that requirements ontology should satisfy. Then, by checking whether the rules are consistent with requirements ontology, we verify the correctness of the ontology. Both the verification method and a prototype system based on the method will be illustrated with examples.
Question Answering (QA) is a task of answering natural language question with an adequate sentence. This paper proposes the method to evaluate candidate answer sentences using two methods, i.e., the Web relevance score and the translation probability. The experiments revealed the proposed method was significantly better than the original methods when MRRs were compared.
A lot of whether or not a software project is successful depends on management capability of project managers. Therefore, EVA (Earned Value Analysis) has received a lot of attention recently as a method for managing a project in an integrated fashion by introducing a unified metric called EV (Earned Value) and quantitatively measuring and analyzing the cost and schedule of a project. Project managers can use EVA to estimate SEAC (Schedule Estimate At Completion) and EAC (Estimate At Completion). However, there are several constraints inherent in a software development plan, and any software development plans have to satisfy the constraints. Therefore, SEAC and EAC are estimated by using EVA and by taking account of the constraints,, and the estimation result by the former is compared with the estimation result by the latter. As a result, it turns out that EVA quite often generates inaccurate estimations for SEAC and EAC. In this way, this paper clarifies problems of a method for estimating SEAC and EAC by EVA.
A business modeling tool called Conceptual Data Modeling (CDM) was proposed by the MASP Association in Japan. This tool creates a diagram called the “Organization Model.” This diagram shows information flow between organizations, and derives ideal state. However, the diagram doesn't have business details, quantitative data and qualitative data. Therefore, it is difficult for a person without business knowledge to derive ideal state. This paper proposes a new analysis method using CDM and Theory of Constraints (TOC). In particular, “Assumptions” obtained by the Conflict Resolution Diagram (CRD) of TOC are mapped to the Organization Model. By the proposed method, we can derive many candidates in greater depth for business process engineering, compared with the conventional CDM.
Goal Structuring Notation (GSN) is used to describe assurance and safety cases for critical software intensive systems. Although it provides simple graphical notations, it is not clear how to describe and validate GSN graphs. In this paper, a review method for GSN is proposed based on a matrix relationship of GSN graph. The proposed method is also exemplified to reduce ambiguity of GSN.
In this paper, we address the recommendation process as a one-class classification problem based on content features and a Negative Selection (NS) algorithm that captures user preferences. Specifically, we develop an Artificial Immune System (AIS) based on a Negative Selection Algorithm that forms the core of a music recommendation system. A NS-based learning algorithm allows our system to build a classifier of all music pieces in a database and make personalized recommendations to users. This is achieved quite efficiently through the intrinsic property of the NS algorithm to discriminate “self-objects” (i.e. music pieces of user's like) from “non self-objects”, especially when the class of non self-object is vast when compared to the class of self-objects and the examples (samples) of music pieces come only from the class of self-objects (music pieces of user's like). Our recommender system has been fully implemented and evaluated and found to outperform state of the art recommender systems based on support vector machines-based methodologies.
We present a concept-based image image acquisition system which embodies the ability to extract a certain subset of images that are representatives of a user defined query concept. The system incorporates SVM - based incremental learning into the relevance feedback procedure in order to refine the discrimination function separating the images pertaining to the user defined query concept (relevant) and the rest images in the database (irrelevant).
Towards building new, friendlier human-computer interaction and multimedia interactive services systems, new computer techniques are applied which aim at revealing information about the affective state, cognitive activity, personality, intention and psychological state of a person. In this paper we conduct a survey to recent developments and current state-of-the-art in the field of body analysis by the use of non-intrusive optical systems. Markerless based human motion capture describes the activity of analyzing and expressing human motion using image processing and computer vision techniques. In this paper, we categorize these approaches in the following four categories: Model-based methods, Region-based methods, Active contour based methods and Feature-based methods and present the advantages and disadvantages of each category separately.
Information systems are now developed as open systems which mutually depend on each other. Although assurance cases are expected to confirm dependability of open systems, it is difficult to clearly describe dependency among co-related systems, because systems are not explicitly represented in structured trees of assurance cases. By using the case study of an elevator control system, we evaluated the d* framework which can describe the dependability of inter-related open systems.
Recently, many system development projects have been using a screen prototype. On requirements definition phase, screen prototyping is effective to decrease some troubles, which are misunderstanding, lack of requirements, and conflict among requirements and so on. On the other hand, there are some problems caused by using a screen prototype. A screen prototype has been mainly used to review system's user interface requirements. Therefore, it could be difficult to review other type of requirements (i.e. requirement about interaction-process between system and user and decision rule which system executes). This paper proposes a system prototype generation tool which transforms three requirements artifacts into system prototype. Three requirements artifacts are business process, business rule and user interface. We also adapt this tool to the requirements definition phase of system development project. As a result of the case study, we introduce that the proposed tool is helpful to review requirements precisely and detect requirements errors.
Quality requirements such as security, usability and performance should be defined in the early stage of the software development because they give large effects to the costs and efforts of the following stages. In addition, we have to confirm such quality requirements are correctly considered in the intermediate software artifacts such as design documents to avoid unnecessary redo of work. In this paper, we propose a method for confirming quality requirements considerations in a design document. The method enables software designers to avoid the oversight and the overabundance about the design related to quality requirements. To evaluate the method, we applied the method to several commercial software systems, and confirmed the method worked well.
We have developed a quality model of an Information System Procurement Specifications (ISPSs). The model is called the ISPS-Q Model. Our aim is to improve the quality of ISPSs for the improvement of procurement operations. The government and/or municipal offices request comments on an ISPS from the public before they open the ISPS to companies. The public comments are referred to in an attempt to further improve the quality of the ISPS. In order to evaluate the efficiency of the ISPS-Q Model, we obtained the public comments and their target ISPS for analyzation. As a result, we could clarify that most of the public comments were written with ambiguous words and/or incomplete statements. Furthermore, we could no regard these ambiguous and/or incomplete comments as a means to improve the quality of the ISPS. In this regard, the ISPS-Q Model is essential to the improvement of the quality of ISPSs. The ISPS-Q Model was developed by adding fair and deleting modifiable to IEEE Std. 830-1998, since fair is a property that the government requires of every ISPS, and that the ISPSs would never be changed after their procurement. We applied the model to a real ISPS and evaluated the quality of the ISPS. We found 19 problematic statements in the ISPS. Multiple companies submitted 109 comments regarding the same ISPS. In this paper, we discuss the differences between our comments and the public comments made by the companies. After that, we define new quality characteristics and reconstruct the ISPS-Q Model. We will close this paper by showing the revised ISPS-Q Model.
In order to improve the quality of software, we review specifications refined from the specifications of the previous design process. Nevertheless, errors and defects of specifications are sometimes detected during testing processes. According to our research, we've found that the internal design specifications have more defects than other specifications. Most of those defects were embedded during the refining process from the external design to the internal design. The aim of our research is to detect those defects by reviewing the internal design specifications. One of the difficulties of reviews is that reviewers trace an item in the external design specifications to an item in the internal design specifications. In order to solve the difficulties, we defined two metamodels and a relationship model between those specifications by analyzing the templates of design specifications. The metamodels represent the structure with items of external design specifications and internal design specifications. The relationship model represents dependency between items in an external design specification and an internal design specification. The metamodels and the relationship model were evaluated by review checklists. This paper shows those models and the result of our evaluation. We conclude that the review checklist derived from the models can list up review perspectives more effectively than conventional checklists.
The full-text retrieval system which is available in Internet-incapable places is reported. To use the retrieval service in a quick and efficient manner, there are properties which the system should require. Portability, accuracy and timeliness are defined as requirements for practical use and we constructed the system for satisfying all these properties. Moreover we supplied the registration support and implemented a unique bookmark feature in order that the system can fulfill the properties named installation-facility and reusability. Evaluation experiments made sure that the added features were useful for the information retrieval. Since the system is developed for supporting literature searching by humanities researchers, it does not only accept a variety of document files but split Excel files by row into the records at the time of registration. Application of the service to source code management is also discussed.
This preliminary study describes a baseline State Transition Model (STM) of the personal motivation process, which should be extended for monitoring and controlling the Personal Software Process (PSP) training course. First, we present an approach for formalizing the motivation process when introducing a business method into an organization. The baseline STM is then discussed for three kinds of PSP course at the Kyushu Institute of Technology, where a stable state indicates successful or unsuccessful completion of a course and thus an unstable state is one of the main risk factors for the course. Various characteristics of trainees are clarified to extend the baseline STM.
Web pages are used for various purposes. End users must sequentially follow page links to find necessary content, requiring significant time and effort. Our method supports the location of content by analyzing link types. This facilitates end users' access to necessary information on web pages.
A quality indicator is a measure of medical service quality. To develop quality indicators that are useful in fair comparison of medical service quality in multiple hospitals, it is helpful to develop them in a proper representation framework. The authors introduce a representation system of quality indicators and show the usefulness of the system by using existing quality indicators. To this end, they organize problems of development of quality indicators from the viewpoint of knowledge science, and analyze expression patterns of assessment criteria of medical services. This paper shows that the representation system has a certain amount of ability to address the problems above, and sufficient power of expression to describe assessment criteria including quality indicators in the patterns above.
System assurance has become a great importance in many industrial sectors. Recently, assurance cases have been widely used for system assurance. Assurance cases are structured documents for arguing dependability/safety/etc. of systems. However, it is sometimes difficult to keep track of the description of assurance cases in accordance with system's actual behavior, particularly for failure mitigating actions. In this paper, we introduce a prototype mechanism for the problem. A preliminary result shows that the mechanism only consumes relatively small amount of resources of the target system.
The paper presents a framework for automatic inferring knowledge about reasons for the appearance of anti-patterns in the program source code during its development. Experiments carried out on histories of development of few open-source java projects shown that we can efficiently detect temporal patterns, which are indicators of likely appearance of future anti-pattern. The approach presented in this paper uses expert knowledge (formal description of anti-patterns) to automatically produce extra knowledge (with machine learning algorithm) about the evolution of bad structures in the program source code. The research can be used to build scalable and adaptive tools, which warns development teams about the fact that system architecture is drifting in the wrong direction, before this is reported by typical static source code analysis tools.
In the upstream software development processes, from requirements definition to design, one of the most important tasks is to make the requirements completely satisfied in the design process, therefore actually realizing the required system. Thus, a seamless connection between the requirements model and the design model is very effective to completely realize user requirements. KAOS  is one of the typical methodologies of Goal-oriented requirements engineering . It elicits and builds up the requirements systematically and exhaustively. ICONIX process  designs a system using UML diagrams. In the ICONIX process, requirements are reviewed using use case models, and then designed with class and sequence diagrams to realize the use cases through robustness diagrams. Based on these methodologies, this paper presents an approach to transfer as much information as possible from KAOS requirements models to preliminary design of system behaviors in ICONIX. Our approach allows the use case models and robustness diagrams to reflect the requirements of the KAOS model with the refinement patterns. We also conducted a case study where flight reservations in an international airline were modeled for the requirements definition and the preliminary design. We could find every requirements defined by the KAOS model in the use case models and the robustness diagrams. Our approach was confirmed to be feasible through this case study.
Prototyping is known as a method to effectively validate the requirements. Especially a prototype system has effect to validate the functional aspects such as operation steps, input/output data, relation between input and output on the external aspects of a system such as user interface. However, such prototype system provides weak support to check the correctness of internal aspects of a system such as business logic. Therefore, we have proposed a review method to check the correctness of the internal aspects based on requirements specification which is created by using model driven requirements analysis method employing automatic generation of prototype system. We aim to remove the defects such as the inconsistency among several different functions or the omission of data structure at an early stage of development. In this paper, the effectiveness of our review method was confirmed by applying this method into the development of Library Management System in a laboratory.
We developed the Software requirement specifications(SRS) editor system with semantic information. Since this system have the semantic information, the engineer who takes charge of software requirement specifications and the engineer who takes charge of software test can reach shared understanding. Also, This system have the function which can find out the of description, the function which reuses the inputted the description, the function which shows the relationship of the descriptions. This paper describes that the sentence of SRS can be divided into six parts, 5W1H (Actor(Who), Time(When), Location(where), Reason(Why), Object of target(What)), and SRS has the tag which 5W1H are written as semantic information. This paper also describes how to give the degree of requirement, and describes the methods to have the tag which the degree of requirement is written as semantic information. This paper describes the SRS editor system with semantic information, and describes the function which can find out the of description, the function which reuses the inputted the description, the function which shows the relationship of the descriptions.
To realize usability of software, consistency of GUI (Graphical User Interface) layouts is required. Consistency is realized by user interface guidelines, in many cases. However, there are many items in the guidelines, and it is difficult to apply all items to GUIs in terms of costs and schedule. In our method, we have proposed a method to realize consistency of GUIs. Concretely, source programs of developed GUIs are analyzed, and layout rules are extracted. Then, source programs of undeveloped GUIs are generated based on the extracted layout rules.
Model Driven Development is a promising approach to develop high quality software systems. We have proposed a method of model-driven requirements analysis using Unified Modeling Language (UML). The main feature of our method is to automatically generate a Web user interface prototype from UML requirements analysis model so that we can confirm validity of in-put/output data for each page and page transition on the system by directly operating the prototype. We have shown that the requirements analysis model has traceability to the final product by implementation experiment from the requirements analysis model. This paper proposes a mapping rule in which design information independent of each web application framework implementation is defined based on the requirements analysis model.