Refine
Year of publication
Document Type
- Article (526)
- Conference Proceeding (162)
- Bachelor Thesis (149)
- Report (102)
- Working Paper (85)
- Book (73)
- Master's Thesis (66)
- Part of a Book (40)
- Periodical Part (38)
- Study Thesis (25)
- Course Material (18)
- Preprint (14)
- Doctoral Thesis (11)
- Review (7)
- Lecture (4)
- Other (3)
Is part of the Bibliography
- no (1323)
Keywords
- Bibliothek (44)
- Digitalisierung (38)
- Milchwirtschaft (36)
- E-Learning (27)
- Molkerei (27)
- Ausbildung (23)
- Euterentzündung (23)
- Informationsmanagement (22)
- Computersicherheit (21)
- Herbarium (20)
The present investigation was conducted to investigate the in-vitro activity of ethanolic extract of roots of Centaurea behens by using DPPH radical scavenging activity, nitric oxide radical scavenging activity, hydrogen peroxide radical scavenging activity, hydroxyl radical. Result suggests that the extract possess significant antioxidant activity as compared to the standard ascorbic acid and thus further in vivo investigation is required to evaluate the medicinal significance of the extract which can be used for assessing the possible therapeutic importance of the drug.
Wearable sensors in healthcare and sensor-enhanced health information systems: all our tomorrows?
(2012)
Wearable sensor systems which allow for remote or self-monitoring of health-related parameters are regarded as one means to alleviate the consequences of demographic change. This paper aims to summarize current research in wearable sensors as well as in sensor-enhanced health information systems. Wearable sensor technologies are already advanced in terms of their technical capabilities and are frequently used for cardio-vascular monitoring. Epidemiologic predictions suggest that neuro-psychiatric diseases will have a growing impact on our health systems and thus should be addressed more intensively. Two current project examples demonstrate the benefit of wearable sensor technologies: long-term, objective measurement under daily-life, unsupervised conditions. Finally, up-to-date approaches for the implementation of sensor-enhanced health information systems are outlined. Wearable sensors are an integral part of future pervasive, ubiquitous and person-centered health
care delivery. Future challenges include their integration into sensor-enhanced health information systems and sound evaluation studies involving measures of workload reduction and costs.
Mining geriatric assessment data for in-patient fall prediction models and high-risk subgroups
(2012)
Background: Hospital in-patient falls constitute a prominent problem in terms of costs and consequences. Geriatric institutions are most often affected, and common screening tools cannot predict in-patient falls consistently. Our objectives are to derive comprehensible fall risk classification models from a large data set of geriatric in-patients’ assessment data and to evaluate their predictive performance (aim#1), and to identify high-risk subgroups from the data (aim#2).
Methods: A data set of n = 5,176 single in-patient episodes covering 1.5 years of admissions to a geriatric hospital were extracted from the hospital’s data base and matched with fall incident reports (n = 493). A classification tree model was induced using the C4.5 algorithm as well as a logistic regression model, and their predictive performance was evaluated. Furthermore, high-risk subgroups were identified from extracted classification rules with a support of more than 100 instances.
Results: The classification tree model showed an overall classification accuracy of 66%, with a sensitivity of 55.4%, a specificity of 67.1%, positive and negative predictive values of 15% resp. 93.5%. Five high-risk groups were identified, defined by high age, low Barthel index, cognitive impairment, multi-medication and co-morbidity.
Conclusions: Our results show that a little more than half of the fallers may be identified correctly by our model, but the positive predictive value is too low to be applicable. Non-fallers, on the other hand, may be sorted out with the model quite well. The high-risk subgroups and the risk factors identified (age, low ADL score, cognitive impairment, institutionalization, polypharmacy and co-morbidity) reflect domain knowledge and may be used to screen certain subgroups of patients with a high risk of falling. Classification models derived from a large data set using data mining methods can compete with current dedicated fall risk screening tools, yet lack diagnostic precision. High-risk subgroups may be identified automatically from existing geriatric assessment data, especially when combined with domain knowledge in a hybrid classification model. Further work is necessary to validate our approach in a controlled prospective setting.
Fall events and their severe consequences represent not only a threatening problem for the affected individual, but also cause a significant burden for health care systems. Our research work aims to elucidate some of the prospects and problems of current sensor-based fall risk assessment approaches. Selected results of a questionnaire-based survey given to experts during topical workshops at international conferences are presented. The majority of domain experts confirmed that fall risk assessment could potentially be valuable for the community and that prediction is deemed possible, though limited. We conclude with a discussion of practical issues concerning adequate outcome parameters for clinical studies and data sharing within the research community. All participants agreed that sensor-based fall risk assessment is a promising and valuable approach, but that more prospective clinical studies with clearly defined outcome measures are necessary.
Complications may occur after a liver transplantation, therefore proper monitoring and care in the post-operation phase plays a very important role. Sometimes, monitoring and care for patients from abroad is difficult due to a variety of reasons, e.g., different care facilities. The objective of our research for this paper is to design, implement and evaluate a home monitoring and decision support infrastructure for international children who underwent liver transplant operation. A point-of-care device and the PedsQL questionnaire were used in patients’ home environment for measuring the blood parameters and assessing quality of life. By using a tablet PC and a specially developed software, the measured results were able to be transmitted to the health care providers via internet. So far, the developed infrastructure has been evaluated with four international patients/families transferring 38 records of blood test. The evaluation showed that the home monitoring and decision support infrastructure is technically feasible and is able to give timely alarm in case of abnormal situation as well as may increase parent’s feeling of safety for their children.
Der zielorientierte Umgang mit Wissen bildet eine zentrale Herausforderung für Unternehmen und deren Mitarbeiter. Deren Kompetenzentwicklung ist für die Unternehmen unter dem Aspekt der Wettbewerbsfähigkeit ein lohnendes Ziel. Diese Arbeit stellt ein Werkzeug zur Messung von Kompetenzen im Persönlichen Wissensmanagement vor. Auf einer Literaturstudie basierend wurde ein Kompetenzkatalog erstellt und mit Hilfe einer Befragung von Fachleuten aus dem Bereichen Informations- und Wissensmanagement validiert. Dieser Kompetenzkatalog findet Eingang in einen Referenzrahmen für Kompetenzen für Persönliches Wissensmanagement. Zur Bestimmung der Niveaustufen Experte, Könner und Kenner im Persönlichen Wissensmanagement wurde ein Messwerkzeug erarbeitet und anhand von zwei Gruppen auf Gültigkeit überprüft. Die eine Gruppe bestand aus Mitarbeitern Exzellenter Wissensorganisationen, die andere aus interessierten Mitarbeitern aus nicht-explizit wissensorientierten Unternehmen. Es konnte nachgewiesen werden, dass beide Gruppen in acht Einzelkompetenzen signifikante Unterschiede besaßen. Auch für weitere Kompetenzen konnten Messdimensionen aus den Rückmeldungen der Umfrage hergeleitet werden. In einigen Fällen allerdings konnten die Niveaustufen Könner und Kenner nicht unterschieden werden.
Digitale 3D-Modelle der Architektur – z.B. Modelle von Gebäuden, Inneneinrichtungsgegenständen und Bauteilen – haben innerhalb der letzten fünf Jahrzehnte sowohl die analogen, auf Papier basierenden Zeichnungen als auch die physischen Modelle aus ihrer planungs-, ausführungs- und dokumentationsunterstützenden Rolle verdrängt. Als Herausforderungen bei der Integration von 3D-Modellen in digitale Bibliotheken und Archive sind zunächst die meist nur rudimentäre Annotation mit Metadaten seitens der Autoren und die nur implizit in den Modellen vorhandenen
Informationen zu nennen. Aus diesen Defiziten resultiert ein aktuell starkes Interesse an inhaltsbasierter Erschließung durch vernetzte Nutzergruppen oder durch automatisierte Verfahren, die z.B. aufgrund von Form- oder Strukturmerkmalen eine automatische Kategorisierung von 3D-Modellen anhand gegebener Schemata ermöglichen. Die teilweise automatische Erkennung von objektinhärenter Semantik vergrößert die Menge an diskreten und semantisch unterscheidbaren Einheiten. Darüber hinaus sind digitale 3D-Modelle zumeist hierarchisch aufgebaut; sie enthalten weitere komplexe Modelle, die wiederum in sich geschachtelt sein können und in einzelnen Fällen einen eigenständigen Nachweis als 3D-Modell wünschenswert machen. 3D-Modelle als Content im World Wide Web können sowohl untereinander als auch mit anderen textuellen wie nichttextuellen Objekten verknüpft werden, also Teil von aggregierten Dokumenten sein. Eine weitere Notwendigkeit ist die Vernetzung mit inhaltlich relevanten Ereignissen, Orten, Begriffen, Personen oder realen Objekten sowie die explizite Beschreibung der Relationen zwischen dem Modell selbst und diesen Entitäten seines spezifischen Kontextes. Die Aggregationen bzw. der Modellkontext sowie die inhärenten Entitäten erfordern Instrumente der Organisation, um dem Benutzer bei der Suche nach Informationen einen Mehrwert zu bieten, insbesondere dann, wenn textbasiert nach Informationen zum Modell und zu dessen Kontext gesucht wird. In der vorliegenden Arbeit wird ein Metadatenmodell zur gezielten Strukturierung von Information entwickelt, welche aus 3D-Architekturmodellen gewonnen wird. Mittels dieser Strukturierung kann das Modell mit weiterer Information vernetzt werden. Die Anwendung etablierter Ontologien sowie der Einsatz von URIs machen die Informationen nicht nur explizit, sondern beinhalten auch eine semantische Information über die Relation selbst, sodass eine Interoperabilität zu anderen verfügbaren Daten im Sinne der Grundprinzipien des Linked-Data-Ansatzes gewährleistet wird. Diese Herangehensweise hat im Gegensatz zu einem Ansatz, der Metadaten als Records auffasst, das Potenzial, Relationen zu jeglichen modellrelevanten Entitäten im Suchraum herzustellen und zugleich diese Relationen für weitere wissensbildende Prozesse verfügbar zu machen.
Background: After kidney transplantation, immunosuppressive therapy causes impaired cellular immune defense leading to an increased risk of viral complications. Trough level monitoring of immunosuppressants is insufficient to estimate the individual intensity of immunosuppression. We have already shown that virus-specific T cells (Tvis) correlate with control of virus replication as well as with the intensity of immunosuppression. The multicentre IVIST01-trial should prove that additional steering of immunosuppressive and antiviral therapy by Tvis levels leads to better graft function by avoidance of over-immunosuppression (for example, viral infections) and drug toxicity (for example, nephrotoxicity).
Methods/design: The IVIST-trial starts 4 weeks after transplantation. Sixty-four pediatric kidney recipients are randomized either to a non-intervention group that is only treated conservatively or to an intervention group with additional monitoring by Tvis. The randomization is stratified by centre and cytomegalovirus (CMV) prophylaxis. In both groups the immunosuppressive medication (cyclosporine A and everolimus) is adopted in the same target range of trough levels. In the non-intervention group the immunosuppressive therapy (cyclosporine A and everolimus) is only steered by classical trough level monitoring and the antiviral therapy of a CMV infection is performed according to a standard protocol. In contrast, in the intervention group the dose of immunosuppressants is individually adopted according to Tvis levels as a direct measure of the intensity of immunosuppression in addition to classical trough level monitoring. In case of CMV infection or reactivation the antiviral management is based on the individual CMV-specific immune defense assessed by the CMV-Tvis level. Primary endpoint of the study is the glomerular filtration rate 2 years after transplantation; secondary endpoints are the number and severity of viral infections and the incidence of side effects of immunosuppressive and antiviral drugs.
Discussion: This IVIST01-trial will answer the question whether the new concept of steering immunosuppressive and antiviral therapy by Tvis levels leads to better future graft function. In terms of an effect-related drug monitoring, the study design aims to realize a personalization of immunosuppressive and antiviral management after transplantation. Based on the IVIST01-trial, immunomonitoring by Tvis might be incorporated into routine care after kidney transplantation.
Research information, i.e., data about research projects, organisations, researchers or research outputs such as publications or patents, is spread across the web, usually residing in institutional and personal web pages or in semi-open databases and information systems. While there exists a wealth of unstructured information, structured data is limited and often exposed following proprietary or less-established schemas and interfaces. Therefore, a holistic and consistent view on research information across organisational and national boundaries is not feasible. On the other hand, web crawling and information extraction techniques have matured throughout the last decade, allowing for automated approaches of harvesting, extracting and consolidating research information into a more coherent knowledge graph. In this work, we give an overview of the current state of the art in research information sharing on the web and present initial ideas towards a more holistic approach for boot-strapping research information from available web sources.
Clinical scores and motion-capturing gait analysis are today’s gold standard for outcome measurement after knee arthroplasty, although they are criticized for bias and their ability to reflect patients’ actual quality of life has been questioned. In this context, mobile gait analysis systems have been introduced to overcome some of these limitations. This study used a previously developed mobile gait analysis system comprising three inertial sensor units to evaluate daily activities and sports. The sensors were taped to the lumbosacral junction and the thigh and shank of the affected limb. The annotated raw data was evaluated using our validated proprietary software. Six patients undergoing knee arthroplasty were examined the day before and 12 months after surgery. All patients reported a satisfactory outcome, although four patients still had limitations in their desired activities. In this context, feasible running speed demonstrated a good correlation with reported impairments in sports-related activities. Notably, knee flexion angle while descending stairs and the ability to stop abruptly when running exhibited good correlation with the clinical stability and proprioception of the knee. Moreover, fatigue effects were displayed in some patients. The introduced system appears to be suitable for outcome measurement after knee arthroplasty and has the potential to overcome some of the limitations of stationary gait labs while gathering additional meaningful parameters regarding the force limits of the knee.
Background: Maintenance of metal homeostasis is crucial in bacterial pathogenicity as metal starvation is the most important mechanism in the nutritional immunity strategy of host cells. Thus, pathogenic bacteria have evolved sensitive metal scavenging systems to overcome this particular host defence mechanism. The ruminant pathogen Mycobacterium avium ssp. paratuberculosis (MAP) displays a unique gut tropism and causes a chronic progressive intestinal inflammation. MAP possesses eight conserved lineage specific large sequence polymorphisms (LSP), which distinguish MAP from its ancestral M. avium ssp. hominissuis or other M. avium subspecies. LSP14 and LSP15 harbour many genes proposed to be involved in metal homeostasis and have been suggested to substitute for a MAP specific, impaired mycobactin synthesis.
Results: In the present study, we found that a LSP14 located putative IrtAB-like iron transporter encoded by mptABC was induced by zinc but not by iron starvation. Heterologous reporter gene assays with the lacZ gene under control of the mptABC promoter in M. smegmatis (MSMEG) and in a MSMEGΔfurB deletion mutant revealed a zinc dependent, metalloregulator FurB mediated expression of mptABC via a conserved mycobacterial FurB recognition site. Deep sequencing of RNA from MAP cultures treated with the zinc chelator TPEN revealed that 70 genes responded to zinc limitation. Remarkably, 45 of these genes were located on a large genomic island of approximately 90 kb which harboured LSP14 and LSP15. Thirty-five of these genes were predicted to be controlled by FurB, due to the presence of putative binding sites. This clustering of zinc responsive genes was exclusively found in MAP and not in other mycobacteria.
Conclusions: Our data revealed a particular genomic signature for MAP given by a unique zinc specific locus, thereby suggesting an exceptional relevance of zinc for the metabolism of MAP. MAP seems to be well adapted to maintain zinc homeostasis which might contribute to the peculiarity of MAP pathogenicity.
Background: Epidemiological and experimental studies suggest that exposure to ultrafine particles (UFP) might aggravate the allergic inflammation of the lung in asthmatics.
Methods: We exposed 12 allergic asthmatics in two subgroups in a double-blinded randomized cross-over design, first to freshly generated ultrafine carbon particles (64 μg/m3; 6.1 ± 0.4 × 105 particles/cm3 for 2 h) and then to filtered air or vice versa with a 28-day recovery period in-between. Eighteen hours after each exposure, grass pollen was instilled into a lung lobe via bronchoscopy. Another 24 hours later, inflammatory cells were collected by means of bronchoalveolar lavage (BAL). (Trial registration: NCT00527462)
Results: For the entire study group, inhalation of UFP by itself had no significant effect on the allergen induced
inflammatory response measured with total cell count as compared to exposure with filtered air (p = 0.188). However, the subgroup of subjects, which inhaled UFP during the first exposure, exhibited a significant increase in total BAL cells (p = 0.021), eosinophils (p = 0.031) and monocytes (p = 0.013) after filtered air exposure and subsequent allergen challenge 28 days later. Additionally, the potential of BAL cells to generate oxidant radicals was
significantly elevated at that time point. The subgroup that was exposed first to filtered air and 28 days later to UFP did not reveal differences between sessions.
Conclusions: Our data demonstrate that pre-allergen exposure to UFP had no acute effect on the allergic inflammation. However, the subgroup analysis lead to the speculation that inhaled UFP particles might have a long-term effect on the inflammatory course in asthmatic patients. This should be reconfirmed in further studies with an appropriate study design and sufficient number of subjects.
The objective was to establish and standardise a broth microdilution susceptibility testing method for porcine Bordetella (B.) bronchiseptica. B. bronchiseptica isolates from different geographical regions and farms were genotyped by macrorestriction analysis and subsequent pulsed-field gel electrophoresis. One reference and one type strain plus two field isolates of B. bronchiseptica were chosen to analyse growth curves in four different media: cation-adjusted Mueller-Hinton broth (CAMHB) with and without 2% lysed horse blood, Brain-Heart-Infusion (BHI), and Caso broth. The growth rate of each test strain in each medium was determined by culture enumeration and the suitability of CAMHB was confirmed by comparative statistical analysis. Thereafter, reference and type strain and eight epidemiologically unrelated field isolates of B. bronchiseptica were used to test the suitability of a broth microdilution susceptibility testing method following CLSI-approved performance standards given in document VET01-A4. Susceptibility tests, using 20 antimicrobial agents, were performed in five replicates, and data were collected after 20 and 24 hours incubation and statistically analysed. Due to the low growth rate of B. bronchiseptica, an incubation time of 24 hours resulted in significantly more homogeneous minimum inhibitory concentrations after five replications compared to a 20-hour incubation. An interlaboratory comparison trial including susceptibility testing of 24 antimicrobial agents revealed a high mean level of reproducibility (97.9%) of the modified method. Hence, in a harmonization for broth microdilution susceptibility testing of B. bronchiseptica, an incubation time of 24 hours in CAMHB medium with an incubation temperature of 35°C and an inoculum concentration of approximately 5 x 105 cfu/ml was proposed.
Background: Often preventive measures are not accessed by the people who were intended to be reached. Programs for older adults may target men and women, older adults, advanced old age groups and/or chronically ill patients with specific indications. The defined target groups rarely participate in the conception of programs or in the design of information materials, although this would increase accessibility and participation. In the German “Reaching the Elderly” study (2008–2011), an approach to motivating older adults to participate in a preventive home visit (PHV) program was modified with the participatory involvement of the target groups. The study examines how older men and women would prefer to be addressed for health and prevention programs.
Methods: Four focus groups (N = 42 participants) and 12 personal interviews were conducted (women and men in 2 age groups: 65–75 years and ≥ 76 years). Participants from two districts of a major German city were selected from a stratified random sample (N = 200) based on routine data from a local health insurance fund. The study focused on the participants’ knowledge about health and disease prevention and how they preferred to be approached and addressed. Videos of the focus groups were recorded and analysed using mind mapping techniques. Interviews were digitally recorded, transcribed verbatim and subjected to qualitative content analysis.
Results: A gender-specific approach profile was observed. Men were more likely to favor competitive and exerciseoriented activities, and they associated healthy aging with mobility and physical activity. Women, on the other hand, displayed a broader understanding of healthy aging, which included physical activity as only one aspect as well as a healthy diet, relaxation/wellness, memory training and independent living; they preferred holistic and socially oriented services that were not performance-oriented. The “older seniors” (76+) were ambivalent towards
certain wordings referring to aging.
Conclusions: Our results suggest that gender-specific needs must be considered in order to motivate older adults to participate in preventive services. Age-specific characteristics seem to be less relevant. It is more important to pay attention to factors that vary according to the individual state of health and life situation of
the potential participants.
The CogALex-V Shared Task provides two datasets that consists of pairs of words along with a classification of their semantic relation. The dataset for the first task distinguishes only between related and unrelated, while the second data set distinguishes several types of semantic relations. A number of recent papers propose to construct a feature vector that represents a pair of words by applying a pairwise simple operation to all elements of the feature vector. Subsequently, the pairs can be classified by training any classification algorithm on these vectors. In the present paper we apply this method to the provided datasets. We see that the results are not better than from the given simple baseline. We conclude that the results of the investigated method are strongly depended on the type of data to which it is applied.
Integrating distributional and lexical information for semantic classification of words using MRMF
(2016)
Semantic classification of words using distributional features is usually based on the semantic similarity of words. We show on two different datasets that a trained classifier using the distributional features directly gives better results. We use Support Vector Machines (SVM) and Multirelational Matrix Factorization (MRMF) to train classifiers. Both give similar results. However, MRMF, that was not used for semantic classification with distributional features before, can easily be extended with more matrices containing more information from different sources on the same problem. We demonstrate the effectiveness of the novel approach by including information from WordNet. Thus we show, that MRMF provides an interesting approach for building semantic classifiers that (1) gives better results than unsupervised approaches based on vector similarity, (2) gives similar results as other supervised methods and (3) can naturally be extended with other sources of information in order to improve the results.
Editorial for the 15th European Networked Knowledge Organization Systems Workshop (NKOS 2016)
(2016)
Knowledge Organization Systems (KOS), in the form of classification systems, thesauri, lexical databases, ontologies, and taxonomies, play a crucial role in digital information management and applications generally. Carrying semantics in a well-controlled and documented way, Knowledge Organisation Systems serve a variety of important functions: tools for representation and indexing of information and documents, knowledge-based support to information searchers, semantic road maps to domains and disciplines, communication tool by providing conceptual framework, and conceptual basis for knowledge based systems, e.g. automated classification systems. New networked KOS (NKOS) services and applications are emerging, and we have reached a stage where many KOS standards exist and the integration of linked services is no longer just a future scenario. This editorial describes the workshop outline and overview of presented papers at the 15th European Networked Knowledge Organization Systems Workshop (NKOS 2016) in Hannover, Germany.
We compare the effect of different segmentation strategies for passage retrieval of user generated internet video. We consider retrieval of passages for rather abstract and complex queries that go beyond finding a certain object or constellation of objects in the visual channel. Hence the retrieval methods have to rely heavily on the recognized speech. Passage retrieval has mainly been studied to improve document retrieval and to enable question answering. In these domains best results were obtained using passages defined by the paragraph structure of the source documents or by using arbitrary overlapping passages. For the retrieval of relevant passages in a video no author defined paragraph structure is available. We compare retrieval results from 5 different types of segments: segments defined by shot boundaries, prosodic segments, fixed length segments, a sliding window and semantically coherent segments based on speech transcripts. We evaluated the methods on the corpus of the MediaEval 2011 Rich Speech Retrieval task. Our main conclusions are (1) that fixed length and coherent segments are clearly superior to segments based on speaker turns or shot boundaries; (2) that the retrieval results highly depend on the right choice for the segment length; and (3) that results using the segmentation into semantically coherent parts depend much less on the segment length. Especially, the quality of fixed length and sliding window segmentation drops fast when the segment length increases, while quality of the semantically coherent segments is much more stable. Thus, if coherent segments are defined, longer segments can be used and consequently fewer segments have to be considered at retrieval time.
Das vorliegende Shortpaper befasst sich mit dem Thema Datenschnittstellen und Datenmapping im Bibliothekswesen für den Austausch von bibliografischen Daten. Ausgangspunkt der Arbeit ist die Schnittstelle Z39.50, welche den Austausch von Metadaten zwischen Bibliotheken als derzeitigen Standard bildet. Neben diesem Austauschformat sind die Darstellungsformate MAB, MARC 21, MARC-XML und als Weiterentwicklung von Z39.50 das SRU (Search/Retrieve via URL) genannt. Weiterhin wird auf das Linked Open Data Netzwerk Bezug genommen. Dieses Netzwerk, auch Semantic Web genannt, wird im Allgemeinen auch mit den Begriffen Web 3.0, Linked Open Data und Web of Data in Verbindung gebracht. Die technische Seite wird anhand von RDF (Resource Description Framework), XML (Extensible Markup Language) erläutert und im Weiteren wird auf die Wichtigkeit der CC0-Lizenzen (Creative Commons Lizenzen) eingegangen. Abschließend wird auf den Begriff Mapping im Kontext der Bibliothekslandschaft eingegangen und die Zusammenarbeit zwischen den Institutionen hervorgehoben.
In this article, we present the software architecture of a new generation of advisory systems using Intelligent Agent and Semantic Web technologies. Multi-agent systems provide a well-suited paradigm to implement negotiation processes in a consultancy situation. Software agents act as clients and advisors, using their knowledge to assist human users. In the presented architecture, the domain knowledge is modeled semantically by means of XML-based ontology languages such as OWL. Using an inference engine, the agents reason, based on their knowledge to make decisions or proposals. The agent knowledge consists of different types of data: on the one hand, private data, which has to be protected against unauthorized access; and on the other hand, publicly accessible knowledge spread over different Web sites. As in a real consultancy, an agent only reveals sensitive private data, if they are indispensable for finding a solution. In addition, depending on the actual consultancy situation, each agent dynamically expands its knowledge base by accessing OWL knowledge sources from the Internet. Due to the standardization of OWL, knowledge models easily can be shared and accessed via the Internet. The usefulness of our approach is proved by the implementation of an advisory system in the Semantic E-learning Agent (SEA) project, whose objective is to develop virtual student advisers that render support to university students in order to successfully organize and perform their studies.