An Experience Factory is an infrastructure for organizational learning in software development that includes an Experience Base as an organizational memory. We introduce a system architecture how such an infrastructure can be technically supported based on Case-Based Reasoning (CBR) technology. As a first instantiation of this architecture we present the CBR-PEB application, a publicly accessible WWW-based experience base for CBR-system development. Based on first experiments, some results about the evaluation of the success of this application are described.

Experience management has a high relevance for the industry as recent studies between 2001 and 2005 show (e.g., KPMG 2001, FhG-WM 2005). In the Fraunhofer study, experience management was the top item among the challenges regarding knowledge and information. “Experience base” was the top item for planned usage/installation among the IT support for knowledge management. Core technologies for realizing experience management systems that come from the field of artificial intelligence are related to decision making, knowledge acquisition and extraction tasks – e.g. case-based reasoning, ontologies, machine learning, and natural language processing. An important motivation for implementing experience management are increasing demands in industry towards process improvement approaches such as CMM Level 5, Six Sigma, etc.: All these approaches aim at better understanding, stabilizing, standardizing, and optimizing processes and decisions in order to achieve a better and more repeatable product quality, e.g., by automation at a more fine-grained level, for production lines as well as for business processes.

An Experience Factory is an infrastructure for organizational learning in software development that includes an Experience Base as an organizational memory. We introduce a system architecture showing how such an infrastructure can be technically supported based on Case-Based Reasoning (CBR) technology. As a first instantiation of this architecture, we present the CBR-PEB application, a publicly accessible WWW-based experience base about CBR applications and tools.

IT-based living assistance systems focusing on the support of people with special needs in their daily routine have to continuously monitor and assist them in an appropriate way. To this end, we have developed a Monitoring and Assistance component including a hybrid reasoner that is able to adapt planned and running treatments according to the current situation and context. In this paper, we explain the underlying approaches followed in the reasoner, describe the reasoning technologies used for this task and its subtasks, and present some first evaluation results.

Challenges in context-sensitive applications are not limited to reasoning. Rapid changes in the environment, which are reflected by the context information, pose a particular challenge. During the run of a single CBR cycle, a lazy adaptation can be used to deal with this challenge of rapidly changing context. In an ambient intelligence system, there is also the need to modularize the knowledge according to the overall system. We present a model for CBR in such a setting. The model consists of an extended/modified CBR process, a knowledge model, and an architecture pattern for embedding a CBR module into an ambient intelligence system. Our focus is on the context-aware adaptation of the actions to be executed when a certain situation is recognized. The technical feasibility of our model has been evaluated with an application in the area of ambient assisted living for elderly people.

Our society needs and expects more high-value services. Such "knowledge-intensive" services can only be delivered if the necessary organizational and technical requirements are fulfilled. In addition, the cost-benefit analysis from the service provider point of view needs to be positive. Continuous improvement and goal-directed (partial) automation of such services is therefore of crucial importance. As a contribution to this we describe our current research vision for (partially) automated support of knowledge work(ers) based on intelligent information systems focusing on the use of experience. For the implementation of such a vision we base on the integration of approaches from artificial intelligence and software engineering. A "deep" integration of case-based reasoning and experience factory is a first successful step in this direction [33,28]. We envision the further integration of software product-lines and multi-agent systems as the next one.

Current case acquisition and case base maintenance techniques implement quality assurance for cases through reviews or by analyzing case properties before making a case available for retrieval. Since reviews of cases with much textual data, in particular, cannot be fully automated, this is done by a maintenance team. With limited resources, this maintenance team becomes a bottleneck. To reduce this bottleneck, our idea is to move parts of the case acquisition and maintenance tasks to the user. With this, the maintenance team needs to do case maintenance only for cases with an ‘out of range’ quality. Obviously, this idea requires changes to the CBR process and system design. This is described by our experience feedback loop design concept, which is presented in this paper. This design concept contains a userfeedback-driven case base maintenance technique as its core element. This design concept has been validated positively in real-world projects for intra-organizational CBR systems with tight integration into business processes and into the respective tools.

Today's software developments are faced with steadily increasing expectations: software has to be developed faster, better, and cheaper. At the same time, application complexity increases. Meeting these demands requires fast, continuous learning and the reuse of experience on the part of the project teams. Thus, learning and reuse should be supported by well-defined processes applicable to all kinds of experience which are stored in an organizational memory. In this paper, we introduce a tool architecture supporting continuous learning and reuse of all kinds of experience from the software engineering domain and present the underlying methodology.

The benefits of an organizational memory are ultimately determined by the usefulness of the organizational memory as perceived by its users. There-fore, an improvement of an organizational memory should be measured in the added perceived usefulness. Unfortunately, the perceived usefulness has many impact factors. Hence, it is difficult to identify good starting points for improvement. This paper presents the goal-oriented method OMI (Organizational Memory Improvement) for improving an organizational memory incrementally from the user's point of view. It has been developed through several case studies and consists of a general usage model, a set of indicators for improvement potential, and a cause-effect model. At each step of the general usage model of OMI, the indicators are used to pinpoint improvement potential for increasing the perceived usefulness and asking the user for specific improvement suggestions where feasible.

In the past many experience factory case studies and experiments have been carried out. We summarize some development steps and research results that, from our perspective, are important. We especially focus on the integration of experience factory and case-based reasoning and report on the respective benefits and impacts of such a seamless integration for building (more) autonomous and automated knowledge-based information systems, which will be of increasing importance in the future. It is our goal to build software-agent-enacted experience factories that improve case bases using maintenance and learning methods.

According to recent surveys, software developers still perceive current inspection tools as insufficient. One reason might be that most of the existing tools focus their support on organizational aspects of the inspection. We present a knowledge-based tool that provides intelligent support in the defect detection step. This tool is embedded in two learning loops: one for the inspection process itself and one loop that goes across the whole software lifecycle.

Abstract: Das Management der Erfahrung von Mitarbeitern ist eine in der Studie „Wissen und Information 2005“ identifizierte HauptHerausforderungen von Organisationen. Dieser Beitrag stellt die fur das Thema Erfahrungsmanagement relevanten Ergebnisse vor und leitet Herausforderungen ab, diesen Erfahrungsaustausch mittels IT zu unterstutzen. Das Management der Erfahrung von Mitarbeitern ist eine in der Studie „Wissen und Information 2005“ identifizierte HauptHerausforderungen von Organisationen. Dieser Beitrag stellt die fur das Thema Erfahrungsmanagement relevanten Ergebnisse vor und leitet Herausforderungen ab, diesen Erfahrungsaustausch mittels IT zu unterstutzen.