Search

Purpose: Sales and procurement are the main boundary spanning functions of an organization – each with a specific focus and partly different views and objectives. They are often considered as two sides of a coin that struggle with one another for relative competitive advantage. The digitalization of procurement functions and the introduction of enterprise resource systems have led to a seeming data abundance. However, the results especially in the area of artificial intelligence are not yet satisfactory in practical application. In addition, few academic works are steered towards procurement. In fact, some expect that procurement is less likely to benefit from the application of methods of artificial intelligence (AI) emphasizing the potential benefits in functions such as finance, production, marketing and sales. Why is that? What can we do about it? Methodology: Explanatory study considering the needed decisions and available data of procurement in contrast to sales and marketing. The manuscript is structured in three sections based upon the “Memorandum of Design-Orientated Business Informatics” with analysis, draft, evaluation and diffusion (Österle et al., 2010). Findings: There is a need for research on the purchasing-marketing interface, not just for AI but also for AI applications and analytics in general. Procurement scholars and managers must speed up data and analytical development, especially since our negotiation partners are benefiting from the rapid development of AI technology. Five propositions have been derived from a master thesis analyzing the needed decision and available data as well as during discussions at the 2021 European Research Seminar to facilitate practical application for management and direct further research. These are more perceived value, more data, better technological solutions, more skills and training, and different role in the value chain. Originality: This is an original work of in-progress research.

In recent years, a considerable number of process model matching techniques have been proposed. The goal of these techniques is to identify correspondences between the activities of two process models. However, the results from the Process Model Matching Contest 2015 reveal that there is still no universally applicable matching technique and that each technique has particular strengths and weaknesses. It is hard or even impossible to choose the best technique for a given matching problem. We propose to cope with this problem by running an ensemble of matching techniques and automatically selecting a subset of the generated correspondences. To this end, we propose a Markov Logic based optimization approach that automatically selects the best correspondences. The approach builds on an adaption of a voting technique from the domain of schema matching and combines it with process model specific constraints. Our experiments show that our approach is capable of generating results that are significantly better than alternative approaches.

We introduce a decision-theoretic framework based on Description Logics (DLs), which can be used to encode and solve single stage multi-attribute de- cision problems. In particular, we consider the background knowledge as a DL knowledge base where each attribute is represented by a concept, weighted by a utility value which is asserted by the user. This yields a compact representa- tion of preferences over attributes. Moreover, we represent choices as knowledge base individuals, and induce a ranking via the aggregation of attributes that they satisfy. We discuss the benefits of the approach from a decision theory point of view. Furthermore, we introduce an implementation of the framework as a Protégé plugin called uDecide. The plugin takes as input an ontology as background knowledge, and returns the choices consistent with the user’s (the knowledge base) preferences. We describe a use case with data from DBpedia. We also provide empirical results for its performance in the size of the ontology using the reasoner Konclude.

Process model matching refers to the automatic detection of semantically equivalent or similar activities between two process models. The output of process model matchers is the basis for many advanced process model analysis techniques and, therefore, must be as accurate as possible. Measuring the performance of process model matchers, however, is a difficult task. On the one hand, it is hard to define which correspondences are actually correct. On the other hand, it is challenging to appropriately take the output of matchers into account, because they often produce confidence values between zero and one. In this paper, we propose the first evaluation procedure for process model matchers that addresses both of these challenges. The core idea is to rank both the computed and the desired correspondences based on their confidence values and compare them using the Spearman’s rank correlation coefficient. We perform an in-depth evaluation in which we apply the new evaluation procedure and illustrate how it helps gaining interesting insights.

Creating mappings between ontologies is a common way of approaching the semantic heterogeneity problem on the Semantic Web. To fit into the landscape of Semantic Web languages, a suitable, logic-based representation formalism for mappings is needed. We argue that such a formalism has to be able to deal with uncertainty and inconsistencies in automatically created mappings. We analyze the requirements for such a formalism, and we propose a novel approach to probabilistic description logic programs as such a formalism, which tightly combines normal logic programs under the well-founded semantics with both tractable ontology languages and Bayesian probabilities. We define the language, and we show that it can be used to resolve inconsistencies and merge mappings from different matchers based on the level of confidence assigned to different rules. Furthermore, we explore the semantic and computational aspects of probabilistic description logic programs under the well-founded semantics. In particular, we show that the well-founded semantics approximates the answer set semantics. We also describe algorithms for consistency checking and tight query processing, and we analyze the data and general complexity of these two central computational problems. As a crucial property, the novel tightly integrated probabilistic description logic programs under the well-founded semantics allow for tractable consistency checking and for tractable tight query processing in the data complexity, and they even have a first-order rewritable (and thus LogSpace data complexity) special case, which is especially interesting for representing ontology mappings. © 2012 Springer Science+Business Media B.V.

Many techniques for the advanced analysis of process models build on the annotation of process models with elements from predefined vocabularies such as taxonomies. However, the manual annotation of process models is cumbersome and sometimes even hardly manageable taking the size of taxonomies into account. In this paper, we present the first approach for automatically annotating process models with the concepts of a taxonomy. Our approach builds on the corpus-based method of second-order similarity, different similarity functions, and a Markov Logic formalization. An evaluation with a set of 12 process models consisting of 148 activities and the PCF taxonomy consisting of 1,131 concepts demonstrates that our approach produces satisfying results.

In this paper we study whether state-of-the-art techniques for multi-domain and multilingual entity linking can be ported to the clinical domain. To do so, we compare two known entity linking systems, BabelFly and TagMe, that leverage on Wikipedia and DBpedia, with the standard clinical semantic annotation and disambiguation system, MetaMap, over the SemRep clinical word sense disambiguation gold standard. We show that BabelFly and especially TagMe, while achieving decent precision on clinical annotation, outmatch MetaMap's F1-score.

Process model matching refers to the automatic identification of corresponding activities between two process models. It represents the basis for many advanced process model analysis techniques such as the identification of similar process parts or process model search. A central problem is how to evaluate the performance of process model matching techniques. Often, not even humans can agree on a set of correct correspondences. Current evaluation methods, however, require a binary gold standard, which clearly defines which correspondences are correct. The disadvantage of this evaluation method is that it does not take the true complexity of the matching problem into account and does not fairly assess the capabilities of a matching technique. In this paper, we propose a novel evaluation method for process model matching techniques. In particular, we build on the assessment of multiple annotators to define probabilistic notions of precision and recall. We use the dataset and the results of the Process Model Matching Contest 2015 to assess and compare our evaluation method. We find that our probabilistic evaluation method assigns different ranks to the matching techniques from the contest and allows to gain more detailed insights into their performance.

We present MEL ++ (M denotes Markov logic net- works) an extension of the log-linear description log- ics EL ++ -LL with concrete domains, nominals, and in- stances. We use Markov logic networks (MLNs) in or- der to find the most probable, classified and coherent EL ++ ontology from an MEL ++ knowledge base. In particular, we develop a novel way to deal with concrete domains (also known as datatypes) by extending MLN’s cutting plane inference (CPI) algorithm.

One of the major problems of large scale, distributed and evolving ontologies is the potential introduction of inconsistencies. In this paper we survey four different approaches to handling inconsistency in DL-based ontologies: consistent ontology evolution, repairing inconsistencies, reasoning in the presence of inconsistencies and multi-version reasoning. We present a common formal basis for all of them, and use this common basis to compare these approaches. We discuss the different requirements for each of these methods, the conditions under which each of them is applicable, the knowledge requirements of the various methods, and the different usage scenarios to which they would apply.

As metadata become of increasing importance to the Web, we will need to start managing such metadata. We argue that there is a strong need for metadata management. We introduce the Spectacle Workbench for verifying semi-structured information and show how it can be used to validate, aggregate and visualize the metadata of an existing information system. We conclude that the possibility to verify and aggregate metadata is an added value with respect to contents-based access to information and that it is possible to generate it on the basis of Web contents. ?? 2004 Elsevier B.V. All rights reserved.

RockIt is a maximum a-posteriori (MAP) query engine for statistical relational models. MAP inference in graphical models is an optimization problem which can be compiled to integer linear programs (ILPs). We describe several advances in translating MAP queries to ILP instances and present the novel meta-algorithm cutting plane aggregation (CPA). CPA exploits local context-specific symmetries and bundles up sets of linear constraints. The resulting counting constraints lead to more compact ILPs and make the symmetry of the ground model more explicit to state-of-the-art ILP solvers. Moreover, RockIt parallelizes most parts of the MAP inference pipeline taking advantage of ubiquitous shared-memory multi-core architectures. We report on extensive experiments with Markov logic network (MLN) benchmarks showing that RockIt outperforms the state-of-the-art systems Alchemy, Markov TheBeast, and Tuffy both in terms of efficiency and quality of results., To appear in proceedings of AAAI 2013

In this paper we present the MultiFarm dataset, which has been designed as a benchmark for multilingual ontology matching. The MultiFarm dataset is composed of a set of ontologies translated in different languages and the corresponding alignments between these ontologies. It is based on the OntoFarm dataset, which has been used successfully for several years in the Ontology Alignment Evaluation Initiative (OAEI). By translating the ontologies of the OntoFarm dataset into eight different languages - Chinese, Czech, Dutch, French, German, Portuguese, Russian, and Spanish - we created a comprehensive set of realistic test cases. Based on these test cases, it is possible to evaluate and compare the performance of matching approaches with a special focus on multilingualism. © 2012 Elsevier B.V. All rights reserved.

trojahn2010c; International audience; This paper reports the first effort into integrating OAEI and SEALS evaluation campaigns. The SEALS project aims at providing standardized resources (software components, data sets, etc.) for automatically executing evaluations of typical semantic web tools, including ontology matching tools. A first version of the software infrastructure is based on the use of a web service interface wrapping the functionality of a matching tool to be evaluated. In this setting, the evaluation results can visualized and manipulated immediately in a direct feedback cycle. We describe how parts of the OAEI 2010 evaluation campaign have been integrated into this software infrastructure. In particular, we discuss technical and organizational aspects related to the use of the new technology for both participants and organizers of the OAEI.

Udgivelsesdato: December 2009 Research in cooperative query answering is triggered by the observation that users are often not able to correctly formulate queries to databases such that they return the intended result. Due to lacking knowledge about the contents and the structure of a database, users will often only be able to provide very broad queries. Existing methods for automatically refining such queries based on user profiles often overshoot the target resulting in queries that do not return any answer. In this article, we investigate methods for automatically relaxing such over-constrained queries based on domain knowledge and user preferences. We describe a framework for information access that combines query refinement and relaxation in order to provide robust, personalized access to heterogeneous resource description framework data as well as an implementation in terms of rewriting rules and explain its application in the context of e-learning systems.

Problems with large monolithical ontologies in terms of reusability, scalability and maintenance have led to an increasing interest in modularization techniques for ontologies. Currently, existing work suffers from the fact that the notion of modularization is not as well understood in the context of ontologies as it is in software engineering. In this paper, we experiment on applying state-of-the-art tools for ontology modularization in the context of a concrete application: the automatic selection of knowledge components to be used for Web page annotation and semantic browsing. We conclude that, in a broader context, an evaluation framework is required to guide the choice of a modularization tool, in accordance with the requirements of the considered application.

The Resource Description Framework (RDF) is a step towards the support for integrated and uniform access to information sources. It is designed to standardize the definition and use of metadata descriptions of Web-based resources. It is complemented with RDF Schema (RDFS) that lets developers define an extensible, object-oriented type system for RDF data models. While RDF is targeted towards representing metadata, it can represent the underlying data as well. Together, RDF and RDFS provide a sound basis for the capture of domain knowledge in a form that lends itself for automatic processing. Since the Web is inherently distributed, RDF querying systems should also be able to handle the distribution of multiple, autonomous RDF repositories. A sound approach is needed for querying distributed RDF sources to disclose the information they contain. In this chapter we examine several architectures for querying distributed RDF (meta)data and point out a number of issues that will have to be dealt with in order to provide a viable solution. The discussion traces the development of distributed applications that attempt to exploit the promises of interoperability of the Semantic Web and have to deal with the requirements that arise from that. The chapter focuses on the problems associated with the introduction of RDF (meta)data in the distributed information system arena.

Just as the industrial society of the last century depended on natural resources, today's society depends on information. A lack of resources in the industrial society hindered development just as a lack of information hinders development in the information society. Consequently, the exchange of information becomes essential for more and more areas of society: Companies announce their products in online marketplaces and exchange electronic orders with their suppliers; in the medical area patient information is exchanged between general practitioners, hospitals and health insurances; public administration receive tax information from employers and offer online services to their citizens. As a reply to this increasing importance of information exchange, new technologies supporting a fast and accurate information exchange are being developed. Prominent examples of such new technologies are so-called Semantic Web and Peer-to-Peer technologies. These technologies address different aspects of the inherit complexity of information exchange. Semantic Web Technologies address the problem of information complexity by providing advanced support for representing and processing information. Peer-to-Peer technologies, on the other hand, address system complexity by allowing flexible and decentralized information storage and processing.

The unstructured, heterogeneous and dynamic nature of the Web poses a new challenge to query-answering over multiple data sources. The so-called Semantic Web aims at providing more and semantically richer structures in terms of ontologies and meta-data. A problem that remains is the combined use of heterogeneous sources. In a dynamic environment, it is no longer realistic to assume that the involved data sources act as if they were a single (virtual) source, modelled as a global schema, as is done in classical data integration approaches. In this paper, we propose an alternative approach where we replace the role of a single virtual data source schema with a peer-to-peer approach relying on limited shared (or: overlapping) vocabularies between peers. Since overlaps between vocabularies of peers will be limited and the dynamic nature of the system prohibits the design of accurate mappings, query processing will have to be approximate. We provide a formal model for such approximate query processing based on limited shared vocabularies between peers, and we show how the quality of the approximation can be adjusted in a gradual manner. The result is a flexible architecture for query-processing in heterogenous and dynamic environments, based on a formal foundation. We present the approach and discuss it on the basis of a case study.

The benefits of modular representations are well known from many areas of computer science. In this paper, we concentrate on the benefits of modular ontologies with respect to local containment of terminological reasoning. We define an architecture for modular ontologies that supports local reasoning by compiling implied subsumption relations. We further address the problem of guaranteeing the integrity of a modular ontology in the presence of local changes. We propose a strategy for analyzing changes and guiding the process of updating compiled information.

Ontologies are shared models of a domain that encode a view which is common to a set of different parties. Contexts are local models that encode a party's subjective view of a domain. In this paper, we show how ontologies can be contextualized, thus acquiring certain useful properties that a pure shared approach cannot provide. We say that an ontology is contextualized or, also, that it is a contextual ontology, when its contents are kept local, and therefore not shared with other ontologies, and mapped with the contents of other ontologies via explicit (context) mappings. The result is Context OWL (C-OWL), a language whose syntax and semantics have been obtained by extending the OWL syntax and semantics to allow for the representation of contextual ontologies.

The increasing awareness of the benefits of ontologies for information processing has lead to the creation of a number of large ontologies about real world domains. The size of these ontologies and their monolithic character cause serious problems in handling them. In other areas, e.g. software engineering, these problems are tackled by partitioning monolithic entities into sets of meaningful and mostly self-contained modules. In this paper, we suggest a similar approach for ontologies. We propose propose an approach for automatically partitioning large ontologies into smaller modules based on the structure of the class hierarchy. The method is demonstrated on a part of the UMLS semantic network. Experiments with larger ontologies are available online at http://swserver.cs.vu.nl/partitioning/.

A number of sophisticated medical ontologies have been created over the past years. With their de-velopment the need for supporting the alignment of different ontologies is gaining importance. We proposed C-OWL, an extension of the Web Ontology Language OWL that supports alignment mappings between different, possibly incompatible ontologies on a semantic level. In this paper we report experiences from using C-OWL for the alignment of medical ontologies. We briefly review key concepts of the C-OWL semantics, explain the setting of the case study including some examples from the alignment and discuss the possibility of reasoning about the mapping based on the C-OWL semantics We conclude by arguing that C-OWL provides an adequate frame-work for aligning complex ontologies in the medical domain.

Peer-to-Peer systems are a new paradigm for information sharing and some systems have successfully been deployed. It has been argued that current Peer-to-Peer systems suffer from the lack of semantics. The SWAP project (Semantic Web and Peer-to-Peer) aims at overcoming this problem by combining the Peer-to-Peer paradigm with Semantic Web technologies. In the course of our investigations it turned out that the nature of Peer-to-Peer systems requires some compromises with respect to the use of semantic knowledge models. In particular, the notion of ontology does not really apply as we often do not find a shared understanding of the domain. In this paper, we propose a data model for encoding semantic information that combines features of ontology (concept hierarchies, relational structures) with a flexible description and rating model that allows us to handle heterogeneous and even contradictory views on the domain of interest. We discuss the role of this model in the SWAP environment and describe the model as well as its creation and access.

The large-scale and almost ubiquitous availability of information has become as much of a curse as it is a blessing. The more information is available, the harder it is to locate any particular piece of it. And even when it has been successfully found, it is even harder still to usefully combine it with other information we may already possess. This problem occurs at many different levels, ranging from the overcrowded disks of our own PCs to the mass of unstructured information on the World Wide Web. It is commonly understood that this problem of information sharing can only be solved by giving computers better access to the semantics of the information. While it has been recognized that ontologies play a crucial role in solving the open problems, most approaches rely on the existence of well-established data structures. To overcome these shortcomings, Stuckenschmidt and van Harmelen describe ontology-based approaches for resolving semantic heterogeneity in weakly structured environments, in particular the World Wide Web. Addressing problems like missing conceptual models, unclear system boundaries, and heterogeneous representations, they design a framework for ontology-based information sharing in weakly structured environments like the Semantic Web. For researchers and students in areas related to the Semantic Web, the authors provide not only a comprehensive overview of the State of the art, but also present in detail recent research in areas like ontology design for information integration, metadata generation and management, and representation and management of distributed ontologies. For professionals in areas such as e-commerce (e.g., the exchange of product knowledge) and knowledge management (e.g., in large and distributed organizations), the book provides decision support on the use of novel technologies, information about potential problems, and guidelines for the successful application of existing technologies.

Current proposals for languages to encode terminological knowledge in intelligent systems support logical reasoning for answering user queries about objects and classes. An application of these languages on the World Wide Web, however, is hampered by the limitations of logical reasoning in terms ofefficiency and flexibility. In this paper we describe, how techniques from approximate reasoning can be used to overcome these problems. We discuss terminological knowledge and approximate reasoning in general and show the benefits ofapproximate reasoning using the example ofbuilding and maintaining semantic catalogues that can be used to query resource locations based on object classes.

Content-related metadata plays an important role in the effort of developing intelligent web applications. One of the most established form of providing content-related metadata is the assignment of web-pages to content categories. We describe the Spectacle system for classifying individual web pages on the basis of their syntactic structure. This classification requires the spe-cification of classification rules associating common pa-ge structures with predefined classes. In this paper, we propose an approach for the automatic acquisition of these classification rules using techniques from inducti-ve logic programming and describe experiments in ap-plying the approach to an existing web-based informa-tion system.

Content-related metadata plays an important role in intelligent information systems. Especially on the world-wide web meaningful metadata describing the contents of a web-site is the key to intelligent retrieval and access of information. Metadata description standards like RDF and RDF schema have been developed and work in progress addresses the use of ontologies to provide a logical foundation for metadata. However, the acquisition of appropriate metadata is still a problem. The main part of the paper is concerned with the specification of ontologies and metadata models. We describe the Spectacle approach, a knowledge-based approach for metadata validation and generation as well as tools related to the ontology language OIL. We conclude that the specification of ontologies and the generation of metadata models are processes that supplement each other and propose a method for semi-automatic generation of metadata models on the basis of ontologies.

DSMW is an extension to Semantic Mediawiki (SMW), it allows to create a network of SMW servers that share common semantic wiki pages. DSMW users can create communication channels between servers and use a publish-subscribe approach to manage the change propagation. DSMW synchronizes concurrent updates of shared semantic pages to ensure their consistency. It oers new collaboration modes to semantic wiki users and supports data ow-oriented processes., DSMW est une extension de Semantic MediaWiki (SMW) qui permet de créer un réseau de serveurs SMW partageant des pages communes. Les utilisateurs créent des canaux de communication entre les serveurs pour propager les changements. Le système synchronise les modifications concurrentes pour garantir leur consistance.

scharffe2012b; This report introduces a novel method for analysing web datasets based on key dependencies. This particular kind of functional dependencies, widely studied in the field of database theory, allows to evaluate if a set of properties constitutes a key for the set of data considered. When this is the case, there won't be any two instances having identical values for these properties. After giving necessary definitions, we propose an algorithm for detecting minimal keys and pseudo-keys in a RDF dataset. We then use this algorithm to detect keys in datasets published as web data and we apply this approach in two applications: (i) reducing the number of properties to compare in order to discover equivalent instances between two datasets, (ii) detecting errors inside a dataset.

shvaiko2012a; International audience; No abstract available.

shvaiko2010a; International audience; No abstract available.

shvaiko2009a; International audience; No abstract available.