Search

Arsenite [As(III)] is much more toxic and difficult in removal than arsenate [As(V)]. Therefore, a method for removing As(III) is more urgently needed than for As(V). The precondition for better As(III) removal is to oxidize it to As(V). This study investigates the oxidation ability of ferrate, Fe(VI), under different pH conditions, different initial concentrations of As(III) and different dosages of Fe(VI). The results show that Fe(VI) is an effective oxidizer for As(III) under basic condition. The pH value of tested water is a significant factor affecting the oxidation efficiency of As(III) by Fe(VI). Although theoretical redox analyses suggest that Fe(VI) is a better oxidizer under acidic condition, the experimental results show that basic condition is more efficient than acidic condition for As(III) oxidation by Fe(VI) due to the acid-catalyzed redox reaction between Fe(VI) and H2O. Initial As(III) concentration does not play an important role in affecting the oxidation efficiency. The results in this study also reveal that the final product of Fe(VI) is Fe(III) and the needed mole ratio of Fe(VI):As(III) for complete As(III) oxidation under basic condition is 2:3. [ABSTRACT FROM AUTHOR]

In this paper, we present a novel approach for the managing the clarity of queries which users post to a knowledge portal. The approach is based on the (i) measuring the ambiguity of a query and (ii) suggesting the refinement of the original query, based on various ambiguity analyses: structure of the query, knowledge repository and user' behaviours. The approach is mainly based on tracking users' interactions with the portal. Since the approach requires more semantic information about users' querying activities (for example, how many results are retrieved), which are not contained in the traditional Web server log, we have developed the new, client-based, method for tracking users' activities which is presented in the paper as well. The short evaluation study shows the benefits of applying our approach in a real-world portal. [ABSTRACT FROM AUTHOR]

This paper describes a semantic portal through which knowledge can be gathered, stored, secured and accessed by members of a certain community. In particular, this portal takes into account companies and research institutes participating in the E.U. funded thematic network called OntoWeb. Ontology-based annotation of information is a prerequisite in order to offer the possibility of knowledge retrieval and extraction. The usage of well-defined semantics allows for the knowledge exchange between different OntoWeb community members. Thus, members are able to publish annotated information on the web, which is then crawled by a syndicator and stored in the portal's knowledge base. The backbone of the portal architecture consists of a knowledge base in which the ontology and the instances are stored and maintained. In addition, ontology-boosted query mechanisms and presentation facilities are provided. [ABSTRACT FROM AUTHOR]

While knowledge is viewed by many as an asset, it is often difficult to locate particular items within a large electronic corpus. This paper presents an agent based framework for the location of resources to resolve a specific query, and considers the associated design issue. Aspects of the work presented complements current research into both expertise finders and recommender systems. The essential issues for the proposed design are scalability, together with the ability to learn and adapt to changing resources. As knowledge is often implicit within electronic resources, and therefore difficult to locate, we have proposed the use of ontologies, to extract the semantics and infer meaning to obtain the results required. We explore the use of communities of practice, applying ontology-based networks, and e-mail message exchanges to aid the resource discovery process. [ABSTRACT FROM AUTHOR]

We explore the application of Spatial Data Mining, the partially automated search for hidden patterns in georeferenced databases, to the analysis of ecological data. A version of the subgroup mining algorithm is described that searches for deviation patterns directly in a spatial database, automatically incorporating spatial information stored in a GIS into the hypothesis space of a data mining search. We discuss results obtained on a multi-relational biodiversity data set recorded in Niger. Vegetation records are georeferenced and associated with a set of environmental parameters. Further data provide information on climate, soil conditions, and location of spatial objects like rivers, streets and cities. The subgroup mining finds dependencies of a plant species on other species, on local parameters and non-local environmental parameters. [ABSTRACT FROM AUTHOR]

Value does not lie in knowledge itself. The real value lies in the potential of that knowledge and thus in: - The ability to identify and remove blockages between individuals and in the system that hinder interaction; - The ability of knowledge workers to find what they are looking for and indeed through serendipity what they do not know they are looking for; - The ability to reuse the knowledge through smart ‘packaging'; - -The courage and pride of workers to show communicate and sell their knowledge via personal and digital channels; - The creativity to identify new contexts in which knowledge is relevant together with the ability to realise these contexts in the organisation - these are in fact the essentials in the innovation process. The main conclusion of this article is that one should speak of the value of knowledge potential rather than the value of knowledge. The value added is dependent on the identification of the transfer to and the activation of knowledge in the various contexts where that knowledge is being used. The value of knowledge therefore is not the intrinsic property of an organisation but is dependent on the environment and the objectives of the measurement. Most organisations focus their KM (ICT) programmes on knowledge transfer and thus on the knowledge connectivity factor. We believe that more emphasis should be put on the knowledge relevance indicator which is directly related to choosing a ‘pull' strategy rather than a ‘push' strategy. Finally the knowledge activation factor deserves closer attention since this determines the actual realisation of the value and is influenced by complicated environmental parameters which are often related to human and cultural influences. - We the authors of this article are working on making the qualitative relationships described above between the elements in the formula and the environmental parameters more quantitative. This will help in the assessment of the key value drivers and the most efficient and effective ways to realise the potential value of knowledge. We hope that this article will spark interaction with our peers so that together we will be able to take the next step in demystifying the true value of knowledge and make more transparent the return on investments knowledge management activities have to offer. [ABSTRACT FROM AUTHOR]

This paper addresses some of the characteristics of learning in knowledge-based work environments. The role that learning mechanisms play in the firm's ability to create, transfer and enhance knowledge is discussed as well as sustaining learning and competitiveness. The work involves a field study of a software development firm [SDF] that utilizes agent-based technologies and other "cooperative decision making" tools, a platform architecture, and "extreme programming". [ABSTRACT FROM AUTHOR]

An important issue for the management team of R&D departments is the selection of a portfolio of projects in periods of excess demand. Standard methods mostly focus on project selection on the basis of expected returns. In many cases other strategic aspects are important e.g. at the department level the development of intellectual capital or at the company level achieving comparative advantage. In this paper we develop a tool for portfolio selection, explicitly taking into account the measurement and balancing of these strategic factors. The indicators of an intellectual capital scorecard are periodically measured against a target and constitute the input of a linear programming model. From the optimal portfolio computed by the model, clear objectives for management can be derived. Our method is illustrated in an industrial case study. [ABSTRACT FROM AUTHOR]

This paper presents a methodology to define areas within the costing profession which require training, based on the expert categorisation of essential knowledge types for the vncommercial and engineering activities within cost estimating (CE2). Ten knowledge types (comprising of both commercial and engineering) were identified as being essential for CE2. A lateral transfer of these cost-knowledge types is required across the CE2 existing experts, and for the professions' novices. If the knowledge used within the costing domain is successfully transferred across it, the comprehension of the interacting cost experts, and the communication between them will increase: Resulting in the consistent production of faster, more accurate estimates. A training analysis (DIF) was performed per knowledge type, in order to determine which required formal training; and which would suffice with informal. The relevant knowledge was then grouped into training modules for development. The primary focus for this training development is within the aerospace and automotive industries. [ABSTRACT FROM AUTHOR]

Software development is knowledge intensive work. Therefore in this kind of education it is necessary not only to teach theories and/or methodologies according to textbooks but also for students themselves to acquire knowledge through exercises, especially group learning because software development in practical setting is done in the form of project. Knowledge through exercises is acquired by a learner or learners' group through problem-solving, or by interactions between learners' group and the teacher side (teacher and teaching assistants). This paper deals with software inspection as one interaction between learners' group and the teacher side, describes a support system of the interaction and the results of application of the system to an actual software engineering class in a university. According to the results, the author also describes design on a knowledge sharing environment for software engineering education. [ABSTRACT FROM AUTHOR]

Organizational defensive routines and lack of ability to early identify and involve critical competence and knowledge, is a main cost driver in developments projects today. We argue that through systematic cross functional risk assessment (PPA- potential problem analysis), culture and competence will be strengthened towards open, involving and proactive handling of own risk factors and potentials. The project not only has to handle its technical risk, but also continuously be able to provide the best organisational preconditions for reaching own defined objectives. The focus is on barriers and effective interventions. The theory and method is put extra to the test, through a cross, - country, -linguistic, and -cultural attempt to handle common risk between a customer and supplier. [ABSTRACT FROM AUTHOR]

This paper describes our ongoing work on developing a holistic approach for the introduction and improvement of KM in an organization. The described methodology does not focus on a specific area but considers all layers of an organization, the organizational, the process and the technology layer. The overall approach will consist of two main components: a KM Measures Repository and a KM Maturity Model. The KM Measures Repository contains best practice measures for each layer. The KM Maturity Model allows an organization to describe itself. Both components will be made "machine executable" and included in a self assessment tool which uses different KM technologies. [ABSTRACT FROM AUTHOR]

This paper addresses the area that is at the intersection of Knowledge Management and Electronic Commerce. This area refers to the exchange and trade of explicit and implicit knowledge at an inter-organisational level. Electronic knowledge marketplaces are currently emerging to address the opportunities and risks found in the purchase and selling of knowledge at the business-to-business (B2B) environment, the need for supporting long-lasting relationships of knowledge exchange and the requirement for facilitating virtual community contexts where knowledge seekers can find suitable knowledge providers and knowledge providers can advertise and sell their available knowledge. The paper describes the business challenges associated with the design of Internet-based knowledge marketplace. INKASS, a European IST project, has stimulated this work and has provided real-life verification on the arguments raised and on the positions adopted herein. [ABSTRACT FROM AUTHOR]

Data mining is typically applied to large databases of highly structured information in order to discover new knowledge. In businesses and institutions, the amount of information existing in repositories of text documents usually rivals or surpasses the amount found in relational databases. Though the amount of potentially valuable knowledge contained in document collections can be great, they are often dificult to analyze. Therefore, it is important to develop methods to efficiently discover knowledge embedded in these document repositories. In this paper we describe an approach for mining knowledge from text collections by applying data mining techniques to metadata records generated via automated text categorization. By controlling the set of metadata fields as well as the set of assigned categories we can customize the knowledge discovery task to address specific questions. As an example, we apply the approach to a large collection of product reviews and evaluate the performance of the knowledge discovery. [ABSTRACT FROM AUTHOR]

Web-based Intelligent Systems Environment (WISE) offers generic, modular, flexible and scalable system solutions for information retrieval, extraction, fusion, knowledge discovery and intelligent decision support, using heterogeneous, distributed data and knowledge sources in information rich, open environments. WISE consists of multiple, interacting, software modules, connected if and when needed, allowing for the development of complex, multiparadigm, intelligent applications. Two illustrative examples of WISE application are described in more detail - a web-based system for water remediation technology assessment and selection, and an intelligent investment advisory system, FIDES. The former one has opted predominantly for multicriteria decision making algorithms (MCDM), particularly outranking methods, while the rule-based systems, and fuzzy rule-based systems in particular, have been added in financial and economic decision making in the latter one. WISE proved flexible enough to satisfy the needs of both, albeit completely different web-based intelligent systems. [ABSTRACT FROM AUTHOR]

Goal of the research presented in this paper is to generate an expertise document analyzing the potential of a document and workflow management system (DMS/WMS) for a specific customer. Addressee of the expertise is on the one hand the possible client who receives decision support for the investment decision in the DMS/WMS (like the expected ROI, quantitative and qualitative benefits, or potential risks), and on the other hand the software vendor himself who obtains detailed technical information about the potential project (achievable contribution margin, technical problems in similar projects, or code snippets used in comparable situations). Moreover, the paper presents the relevant part of the corresponding business process of a vendor of DMS/WMS, a description of the deficits of the current sales staff to analyze the potential of complex software systems (focus on DMS/WMS), and a concept together with a prototypical implementation of an expertise system. Since in a project-oriented distribution of a DMS/WMS many pieces of internal and external knowledge can be useful, the design of an ontology of the domain DMS/WMS which is realized in form of a Semantic Network is described. This domain knowledge serves as a means to store experiences, e.g. from past projects. Additionally, a rule base is presented in order to process business rules. [ABSTRACT FROM AUTHOR]

Distributed Knowledge Management is an approach to Knowledge Management based on the principle that the multiplicity (and heterogeneity) of perspectives within complex organizations should not be viewed as an obstacle to knowledge exploitation, but rather as an opportunity that can foster innovation and creativity. Despite a wide agreement on this principle, most current KM systems are based on the idea that all perspectival aspects of knowledge should be eliminated in favor of an objective and general representation of knowledge. In this paper we propose a peer-to-peer architecture (called KEx), which embodies the principle above in a quite straightforward way: (i) each peer (called a K-peer) provides all the services needed to create and organize "local" knowledge from an individual's or a group's perspective, and (ii) social structures and protocols of meaning negotiation are defined to achieve semantic coordination among autonomous peers (e.g., when searching documents from other K-peers). [ABSTRACT FROM AUTHOR]

The objective of this paper was to explore the role of a Knowledge Management professional in today's blue chip companies, to examine if and how the KM professionals differ from each other, and if Knowledge Management professionals on the whole match the earlier proposed models of an ‘ideal' KM professional. The findings suggested that employing organisations did not have a clear idea of the KM role. This ambiguity was picked up the newly appointed KM professionals, who as a result felt uncertain and insecure in their KM role. Further, the findings supported earlier studies, in what the participants perceived the KM role to be (2/3 change agent, 1/3 information systems technology) although they perceived themselves primarily as management consultants/entrepreneurs rather than technicians/managers. [ABSTRACT FROM AUTHOR]

The past years have seen an explosion in the number of virtual knowledge communities and community membership is assuming an increasingly important role in the working lives of their members. Many communities have a number of common, key constituent elements. This paper examines the characteristics of real and virtual communities, identifies key elements and proposes a model for analysing and evaluating them over time. Focusing on the interaction and knowledge exchange activities in a virtual community of freelance translators, an empirical study of life in a thriving knowledge community is presented and recommendations made for continued success. [ABSTRACT FROM AUTHOR]

In a distributed knowledge management system, knowledge is firstly produced and then delivered to a person or community of users that is interested in it. Knowledge creation or production is a set of cooperative tasks that need to be coordinated. A multiagent architecture is introduced for this aim, where knowledge-producing agents are arranged into knowledge domains or marts, and a distributed interaction protocol is used to consolidate knowledge that is generated. The knowledge that is produced in this way is used as the source data to dynamically build user communities that can drive the delivery of knowledge amongst users. [ABSTRACT FROM AUTHOR]

The research project ServiceWorld1 aims at developing a virtual community to support knowledge management in the field of after-sales service for machine and plant construction industries. Based on an existing service management reference model, a service community reference model and a service knowledge model were developed. Relevant services were identified and implemented in a community software prototype. First experiences of the pilot implementation at a heavy weight machine manufacturer will be available by the end of the year 2002. [ABSTRACT FROM AUTHOR]

In this paper, we elaborate on the idea that the processes of managing individual and organizational knowledge are situated in the context of work and emerge from the workplace practices. We address the relationship between Business Process-orientation and work practices as they represent the formal and the situated perspectives on work respectively. We examine the mediating role of information and communication technologies in the creation of collective knowledge within communities of practice from a Business Processoriented viewpoint. The paper proposes a series of design implications and presents a prototype solution for a knowledge management system. [ABSTRACT FROM AUTHOR]

Knowledge management can support industrial research processes. Indeed, these industrial research processes are situated between a system of operational units, requiring and using research results, and a system of technological providers and academic laboratories. The knowledge flow between these systems is exposed to certain industrial constraints. The research center needs to optimize internal and external knowledge flows. In this purpose we identified a knowledge typology of industrial researchers. Based on a research process model, the knowledge typology and a theoretical knowledge management model we propose a portal solution. This portal solution allows manipulating information and knowledge more flexible in order to assure a certain dynamic for the knowledge management system. In future work we will integrate a knowledge evaluation process to focus on this dynamical aspect. This article discusses the introduction of a knowledge management system for industrial research processes at the EADS Corporate Research Center. [ABSTRACT FROM AUTHOR]

This article describes the PROMOTE approach, to define and implement a service-based Knowledge Management System (KMS) that has been developed during an EC-funded project. The aim is to define a modelling language that is used to analyse, document and implement a KMS on the basis of so-called Knowledge Management Processes (KMPs). KMPs define the knowledge interaction between knowledge workers in a process-oriented manner and consists of activities that are supported by knowledge management key actions (KA) like searching, categorising or storing information. These knowledge management key actions are supported by IT-based knowledge management services (KM-Service). The prototype of PROMOTE® is briefly mentioned to discuss the KMP-models and the service based KMS. [ABSTRACT FROM AUTHOR]

Next-generation business solutions in the domain of collaborative knowledge work require the integration of process modeling, groupware, and workflow technologies. This paper aims at discussing the fundamentals required for improving process-awareness for group collaboration. Prospective process-aware collaborative tools are required to record, map and manage processes involved in knowledge work: from creating ideas to their development into profit contributors. Development activities involve distributed teams working across organizations, time, and locations. We present acase study and an implementation of a collaborative knowledge flow solution, enabling the traceability of collaborative work activities for team members. This leads to increased efficiency of New Product Development teams as well as to faster timeto- market of product and services. [ABSTRACT FROM AUTHOR]

‘Interestingness' measures are used to rank rules according to the ‘interest' a particular rule is expected to evoke in a user. In this paper, we introduce an aspect of interestingness called ‘item-relatedness' to determine interestingness of item-pairs occurring in association rules. We elucidate and quantify three different types of item-relatedness. Relationships corresponding to item-relatedness proposed by us are shown to be captured by paths in a ‘fuzzy taxonomy' (an extension of the concept hierarchy tree). We then combine these measures of item-relatedness to arrive at a total-relatedness measure. We finally demonstrate the efficacy of this total measure on a sample taxonomy. [ABSTRACT FROM AUTHOR]

Normative documents are used to record knowledge about "what" and "how" regarding goals or purposes the organisation must achieve, and products or services the organisation must produce or provide. In cases, when normative documents are complex and must be understood by individuals that have not been involved in their development, graphical representations of the document contents, and of the relationships between the documents may serve as facilitators of better comprehension of the documents. [ABSTRACT FROM AUTHOR]

Today's web search engines return very large result sets for query formulations consisting of few specific keywords. Results are presented as ranked lists containing textual description of found items. Such representations do not allow identification of topical clusters, and consequentially make it difficult for users to refine queries efficiently. In this paper, we present WebRat, a framework for web-based retrieval, clustering and visualisation which enables parallel querying of multiple search engines, merging of retrieved result sets, automatic identification of topical clusters and interactive visualisation of the result sets and clusters for query refinement. This framework is lightweight in the sense that it consists of a small, platform-independent component which can be easily integrated into exisiting Internet or Intranet search forms without requiring specific system environments, server resources or precalculation efforts. The WebRat system extends existing approaches to web search result visualisation in many aspects: Found results are added incrementally as they arrive, labelling is performed in 2-dimensional space on clusters the user can see and rendering is optimised to provide sufficient performance on standard office machines. The WebRat framework has been used to implement a variety of applications: We have provided enhanced web search capabilities for users doing scientific research. Overview and refinement capabilities have been implemented for the environmental domain. Finally, abstracts generated on the fly by a knowledge management system have been used to provide topical navigation capabilities to developers searching for technical information in mailing list archives. [ABSTRACT FROM AUTHOR]

Before implementing a PDM-system within a company, the internal processes of product and process development and the information handled herein should be organized well. To enable this organization, one should be able to see the bottlenecks. Therefore the working methods and the documents involved should be made transparent. That is why the development processes of three automotive suppliers are analyzed and documented in three representations, formatted according to a generic scheme, to gain transparency of the processes. Based on these representations a so-called induced model of product and process development is created. In this paper, the format of the representations will be shortly explained, the application method for the realization of the representations will be described and some of the results will be illustrated. Also the research problems that came up during the research will be described. [ABSTRACT FROM AUTHOR]

In an occupational environment every employee has to accept a role which she or he is obliged to execute depending on primarily objective criteria like the affiliation of an enterprise to a certain type, sector and life cycle. There is only little scope left for subjective preferences and resentments. The pragmatic research objective is to design frameworks to identify the information requirements of different roles. [ABSTRACT FROM AUTHOR]

Mobile devices, such as PDAs, evolve rapidly from digital calendars and address books to hosts of more complex functionality. Screen quality, processing power and mobile networks allow for full-blown mobile information systems. However, mobility and scaled-down technology lead to specific limitations in contrast to the usage of desktop computers. It is absolutely crucial to consider these restrictions to ensure the development of usable mobile applications. In this paper, we are going to derive basic requirements for mobile knowledge management and present a first approach to possible solutions within the area of project management work. Underlying technical and human constraints are drawn from a general discussion of mobile information access and processing. We will point out focal components and problems to address. Finally, our main concept of context-aware KM will be sketched, building on the idea of Knowledge Portals and prototyped as the "BuddyAlert" system. [ABSTRACT FROM AUTHOR]

The information society develops a novel and comprehensive vision of governance with knowledge as core part. For building KMS a mere transferring of concepts from the private sector to the public area will not suffice; approaches of their own are necessary. For this purpose, Public Administration in general as well as in typical applications is juxtaposed with main features of KMS. Focal question is how features of KMS technology can meet requests derived from administrative work. The outcome is a remarkable picture; it shows bright options and possibilities as well as shortfalls and indications for further development. [ABSTRACT FROM AUTHOR]

An increasing number of enterprises are getting interested in exploiting knowledge, tacit or explicit, lying outside their organizational borders and augmenting the knowledge network of their organizations. A first generation of knowledge e-marketplaces has arisen to provide the platforms for knowledge exchange and trading in an inter-organizational level. This paper develops a framework to evaluate the business models, roles, processes, and revenue models of knowledge trading platforms and provides a survey of six existing knowledge marketplaces based on this framework. Finally, a set of conclusions is drawn on what issues should be addressed in a knowledge marketplace in order to eliminate the risks and gain the trust of its targeted customers. [ABSTRACT FROM AUTHOR]

Understanding the way human interaction and knowledge transfer work is a fundamental issue in KM. IT supported/assisted KM strategies face the question of the extent to which knowledge can be codified, programmed into software systems, stored, extracted, distributed, organized and managed electronically. This paper examines the possibility of translating theorems rooted in socio-linguistics, epistemology, psychology, philosophy and other human sciences into a concept that fits the paradigms used in information theory. Information is defined as a >selection process< steered by four different types of commands: Selection requests, instructions, representation and interpretation modificators. >Selections< are assumed to be (true) knowledge under four conditions: Situational relatedness (situationedness of knowledge), supposition of reality-substantiality (sense making, belief, claim), conformity with patterns of social behaviour (culture, civilization, artificial world), integratability into the image of the relation between >self< and >world< (self-identity, Weltbild). Extraction, reconstruction of >knowledge< from protocols of thoughts and mental images is performed through the execution of internalized, learned selection/action patterns applied to individual life experience, and sense making understanding of a given situation. The translation of the logical structure of >situational understanding<, >sense making< and >selection process< into a conventional representation of functional relations between elements of an organizational situation (information processing in/of organizations by means of IT) makes >sense< if the choice between options (transparency) of selections (the content of negotiation and calibration) leads to desirable flexibility of individuals and organizations in a globalized world. [ABSTRACT FROM AUTHOR]

This study produces and evaluates a process that aims to bring out the tacit competence within professionals so that others can use it. The main conclusion in this study is that using a seminar based on interviews for bringing out tacit competence works and can be recommended. The seminar has been developed gradually through prototyping. Two larger evaluation studies have been performed as well as smaller informal studies. The results from the two test seminars as well as results from interviews and questionnaires have been analysed and compared with established criteria to measure the success of the proposed process. The competence brought out during the test seminars has been found easy to understand and predicted to be useful. [ABSTRACT FROM AUTHOR]

Due to the recent trend of the management science and ebusiness, it is significant to respond customers' requirements rapidly than ever before. In order to achieve the rapid response, we have to manage enterprise models and to reflect the requirements on the models. This paper proposes a management framework of layered enterprise models. Proposed framework consists of a business model repository and a software repository that support enterprise modeling, workflow modeling, and application development. This framework helps us to develop web applications in incremental deployment of analysis, design, and implementation. We have implemented a prototype environment by JAVA. Each repository's contents are described by XML so that the repositories are interoperable. [ABSTRACT FROM AUTHOR]

Integration of heterogeneous data is one of the greatest challenges for versatile e-learning environments, since support for different multimedia data formats is often restricted or adaptions are necessary to fit strict requirements. Therefore, we examine the opportunities given by new metadata standards like MPEG-7 and XML for knowledge management in terms of automated processing, evaluation and presentation of e-content. In Germany's first interdisciplinary and collaborative research center on "Media and Cultural Communications", we are studying the influence of transcription, localization and (re-) addressing on e-learning environments. Exemplarily, we want to introduce our Virtual Entrepreneurship Lab (VEL) as an approach to comply with these tasks in a multimedia e-learning environment. [ABSTRACT FROM AUTHOR]

The End-User Access to Multiple Sources, the EAMS system integrates document collections in the internet (intranet) and relational databases by an ontology. The ontology relates the document with the database world and generates the items in the user interface. In both worlds, machine learning is applied. In the document world, a learning search engine adapts to user behavior by analysing the click-through- data. In the database world, knowledge discovery in databases (KDD) bridges the gap between the fine granularity of relational databases and the coarse granularity of the ontology. KDD extracts knowledge from data and therefore allows the knowledge management system to make good use of already existing company data. The EAMS system has been applied to customer relationship management in the insurance domain. Questions to be answered by the system concern customer acquisition (e.g., direct marketing), customer up and cross selling (e.g., which products sell well together), and customer retention (here: which customers are likely to leave the insurance company or ask for a return of a capital life insurance). Documents about other insurance companies and demographic data published in the internet contribute to the answers as do the results of data analysis of the company's contracts. [ABSTRACT FROM AUTHOR]

The effectiveness of metasearch data fusion procedures depends crucially on the properties of common documents distributions. Because we usually know neither how different search engines assign relevance scores nor the similarity of these assignments, common documents of the individual ranked lists are the only base of combining search results. So it is very important to study the properties of common documents distributions. One of these properties is the Overlap Property (OP) of documents retrieved by different search engines. According to OP, the overlap between the relevant documents is usually greater than the overlap between non-relevant ones. Although OP was repeatedly observed and discussed, no theoretical explanation of this empirical property was elaborated. This paper considers formal research of properties of the common documents distributions. In particular, sufficient and necessary condition of OP is elaborated and it is proved that OP should take place practically under arbitrary circumstances. [ABSTRACT FROM AUTHOR]