Institut für Health Care & Public Management

Permanent URI for this collectionhttps://hohpublica.uni-hohenheim.de/handle/123456789/25

Browse

Now showing 1 - 11 of 11

Automatische Prozessüberwachung in der Nutztierhaltung : Gestaltung eines Verfahrens zur Extraktion von Prozessindikatoren aus Bilddaten mittels Deep Learning
(2024) Riekert, Martin; Kirn, Stefan
Der Gegenstand der Arbeit sind Verfahren zur Erkennung der Prozessindikatoren zum Liegeverhaltens von Schweinen aus Videobildern. Das Liegeverhalten eines Schweines ist definiert als dessen Position im Stallbereich und ob es liegt oder nicht liegt. Hierzu zeichnen in Stallbereichen installierte Kameras kontinuierlich Videobilder auf, die hinsichtlich des Liegeverhaltens der abgebildeten Schweine analysiert werden. Die Stallposition im Tagesverlauf und die Liegehäufigkeit sind wichtige Prozessindikatoren in der Schweinehaltung. Die Problemstellung liegt in der Korrektheit der Erkennung der Prozessindikatoren zum Liegeverhalten. Bisherige Verfahren treffen zwei Annahmen: Erstens werden nur zur Tageszeit aufgenommene Videobilder verarbeitet, sodass das Liegeverhalten im Dauerbetrieb und insbesondere der Nacht und Dämmerung unerkannt bleibt. Zweitens setzen die Verfahren Homogenität der Stallbereiche voraus, sodass alle Stallbereiche ähnliche physische Eigenschaften besitzen. Diese Annahme beschränkt die Eignung der Verfahren für unbekannte Stallbereiche. Die Problemstellung wird aus der Perspektive des Deep Learning untersucht. Die vorliegende Arbeit trägt zur automatischen kontinuierlichen Erfassung der Position und der Körperhaltung von Schweinen unter Verwendung von handelsüblichen 2D-Kameras in realen Umgebungen unter verschiedenen Lichtbedingungen und Kameraausrichtungen rund um die Uhr bei. In der vorliegenden Arbeit wird untersucht, wie Deep-Learning-Verfahren angewendet und gestaltet werden können, um unter realen Bedingungen kontinuierlich eine hohen mean Average Precision zu erzielen und welche Einschränkungen bei der Anwendung von Deep-Learning-Verfahren bestehen. Das gestaltete und evaluierte Verfahren ermöglicht die Erkennung von Schweinen in einer großen Anzahl verschiedener Ställe, dies beinhaltet das Liegeverhaltens unter Verwendung realistischer Kamera- und Beleuchtungseinstellungen (inkl. Nachtaufnahmen mittels Nahinfrarot).
Beitrittsentscheidungen zu Multiagenten-Organisationen : ein Revenue Management-basierter Ansatz
(2018) Premm, Marc; Kirn, Stefan
Der Forschungsbereich Multiagentensysteme hat sich seit den späten 1970er Jahren als Teilbereich der verteilten künstlichen Intelligenz (VKI) etabliert. Gegenstand dieses Forschungsbereichs sowie dieser Arbeit sind Softwareagenten, die als Softwaresysteme zielorientiert agieren und mittels Lernverfahren eine gewisse Autonomie gegenüber ihrem Entwickler erlangen. Softwareagenten ist es möglich, sich zu Multiagentensystemen zusammenzuschließen. Multiagentensysteme sind folglich offene Systeme, aus denen Softwareagenten ein- und wieder austreten können – im Allgemeinen ohne globale Kontrolle. Hieraus ergibt sich eine gewisse Flüchtigkeit sowohl der Mitgliedermenge als auch der Interaktionsstrukturen eines Multiagentensystems und somit eine eingeschränkte Möglichkeit zur Steuerung des nach außen hin wahrnehmbaren Systemverhaltens, welcher es beim Einsatz von Softwareagenten in kommerziellen Anwendungen im Allgemeinen bedarf. Multiagenten-Organisationen bilden einen Ansatz um die betriebswirtschaftliche Organisationstheorie zur internen Strukturierung von Multiagentensystemen zu nutzen und folglich auch deren Außenverhalten zielgerichtet zu steuern. Diese Arbeit versteht Multiagenten-Organisationen als auf Dauerhaftigkeit ausgelegte Zusammenschlüsse von mehreren unabhängigen Softwareagenten, die durch vertragliche Regelungen an der Erfüllung eines vorgegebenen Organisationsziels mitwirken. Softwareagenten benötigten Zugriff auf Ressourcen um einerseits bestimmte Dienste anbieten (z.B. Datenbank-Zugriff) aber auch um ihre eigene Ausführung sicherstellen zu können (z.B. Hardwareressourcen). Softwareagenten können die ihnen zur Verfügung stehenden Ressourcen nutzen um einer Multiagenten-Organisation im Rahmen einer Mitgliedschaft Dienste zur Verfügung zu stellen. Die Erbringung von Diensten konsumiert stets einen Teil der einem Softwareagenten zur Verfügung stehenden Ressourcen. Softwareagenten werden daher einer Multiagenten-Organisation Dienste nur gegen eine entsprechende Kompensation zur Verfügung stellen. Der Beitritt eines Softwareagenten zu einer Multiagenten-Organisation ist stets das Ergebnis von Beitrittsverhandlungen zwischen diesen beiden Akteuren, in denen neben den bereitzustellenden Diensten und der zu zahlenden Kompensation, insbesondere die Dienstgüte verhandelt wird. Softwareagenten können dabei Mitglied in mehreren Multiagenten-Organisationen sein. Jeder Softwareagent hat hierbei zu entscheiden, ob bzw. welcher Multiagenten-Organisation er beitritt und in welchem Umfang er die verfügbaren Ressourcen hierfür einsetzt. Dabei hat er bereits bestehende Mitgliedschaften in anderen Multiagenten-Organisationen, für die der Softwareagent bereits Dienste bereitstellt und somit Ressourcen auslastet, bei der Beitrittsentscheidung mit zu berücksichtigen. Hieraus lässt sich folgende Forschungsfrage ableiten: Wie sind Entscheidungsverfahren auszugestalten, die es Softwareagenten ermöglichen, nutzenmaximierende Beitrittsentscheidungen zu Multiagenten-Organisationen zu treffen? Diese Arbeit präsentiert ein Verfahren zur Optimierung von Beitrittsentscheidungen von Softwareagenten zu Multiagenten-Organisationen. Das entwickelte Verfahren basiert auf Ansätzen des Revenue Management als Teilbereich des Operations Research. Die in der Literatur vorhandenen Verfahren des Revenue Management sind dabei nicht in der Lage, die Gegebenheiten von Beitritts-entscheidungen von Softwareagenten abzubilden. Das entwickelte Revenue Management-basierte Modell der Beitrittsentscheidung versetzt Softwareagenten in die Lage Mitgliedschaften in Multiagenten-Organisationen zu bewerten, deren Dauer a-priori unbekannt ist, und potentielle Mitgliedschaften anhand verschiedener Dienstgüteklassen abzugrenzen. Das entwickelte Verfahren nutzt die auf dieser Basis vorhandenen Möglichkeiten der Optimierung von Beitritts¬entscheidungen, greift den Ansatz von Ressourcen-bezogenen Reservationspreisen (so genannten Bid-Prices) aus dem Revenue Management auf und passt diesen auf die Gegebenheit von Beitritts¬entscheidungen von Softwareagenten an. Das entwickelte Verfahren wird durch ein Simulationsexperiment auf ihre Wirksamkeit und die hierfür notwendigen Bedingungen hin evaluiert. Zur Erreichung dieser Ziele wird ein Referenz- Entscheidungsverfahren herangezogen und verschiedene Parameter der Simulation jeweils paarweise variiert. In der Mehrzahl der untersuchten Parameterkonstellationen erzielt das entwickelte Verfahren eine Steigerung der erwirtschafteten Kompensation. Softwareagenten, die in einer Domäne Dienste anbieten, in der diese Parameterkonstellationen vorzufinden sind, werden durch die Anwendung des Verfahrens in die Lage versetzt, höhere Kompensationen durch Mitgliedschaften in Multiagenten-Organisationen zu erzielen als mit dem Referenz-Entscheidungsverfahren. Sind dem Softwareagenten einzelne Parameter a-priori nicht bekannt, kann mit Hilfe dieser simulativen Evaluation bereits das Risiko des Einsatzes des Verfahrens abgeschätzt werden. Das entwickelte Verfahren konnte jedoch nicht für alle Parameterkonstellationen einen Vorteil erwirtschaften, so dass für die Anwendung zwei wesentliche Voraussetzungen zu beachten sind: (1) Differenzierungsmöglichkeiten. Grundvoraussetzung für die Anwendung des Verfahrens ist die Möglichkeit der Differenzierung und den damit verbundenen Unterschieden in der Höhe der erzielten Kompensation bei gleichem Ressourceneinsatz. Diese Differenzierung kann durch verschiedene Maßnahmen des Softwareagenten erreicht werden: (i) Für verschiedene Dienste mit gleichem Ressourcenbedarf werden unterschiedlich hohe Kompensationsforderungen gestellt, (ii) ein oder mehrere Dienste werden in unterschiedlichen Dienstgüteklassen angeboten, deren Kompensationen sich in ausreichendem Maße unterscheiden oder (iii) verschiedene Dienste nutzen eine unterschiedliche Menge an Ressourcen. (2) Nachfrage. Das in dieser Arbeit entwickelte Bid-Price-Verfahren kann nur durch Ablehnen von bestimmten Anfragen Vorteile gegenüber dem Referenz-Entscheidungsverfahren generieren. Voraussetzung ist somit eine entsprechende Nachfrage nach den angebotenen Diensten eines Softwareagenten und somit nach dessen Mitgliedschaft in Multiagenten-Organisationen. Steht diese Nachfrage nicht in ausreichendem Maße zur Verfügung, kann im Allgemeinen kein Vorteil gegenüber anderen Ansätzen erzielt werden. Abhängig von der individuellen Situation an angebotenen Diensten wirkt sich auch ein bestimmtes Verhältnis an nachfragenden Multiagenten-Organisationen positiv auf das Ergebnis des vorgestellten Verfahrens aus: Ist die überwiegende Zahl der Anfragen auf niederwertige Dienste oder Dienstgüteklassen ausgerichtet, jedoch auch eine ausreichend hohe Zahl an höherwertigen Anfragen vorhanden, erzielt das Verfahren deutliche Steigerungen gegenüber dem Referenz-Entscheidungsverfahren. Die voran genannten Voraussetzungen haben sich in der simulativen Evaluation als wesentlich für eine Vorteilhaftigkeit der Anwendung des entwickelten Verfahrens herausgestellt. Wenden Softwareagenten das entwickelte Verfahren für Beitrittsentscheidungen zu Multiagenten-Organisationen in Domänen an, die diese Voraussetzungen erfüllen, ist eine Steigerung der erwirtschafteten Kompensation gegenüber dem Referenz-Entscheidungsverfahren wahrscheinlich. Falls die Nachfrage kleiner als erwartet ausfällt, erzielte das Verfahren häufig die gleiche Kompensation wie das Referenz-Entscheidungsverfahren und erwirtschaftete nur in einzelnen Simulationsdurchläufen deutlich weniger, so dass selbst bei a-priori unbekannten oder unsicheren Parametern eine Anwendung des entwickelten Verfahrens möglich ist.
Electronic service allocation with private quality information
(2017) Widmer, Tobias; Kirn, Stefan
The efficient allocation of electronic services is a complex business problem. Customers demand electronic services from service providers who supply these services at a specified quality of service (QoS). Electronic marketplaces provide a platform on which multiple customers and multiple providers negotiate the allocation of electronic services. Such marketplaces might be administrated by government authorities or large corporations who aim at a socially optimal allocation. This research addresses the allocation problem for electronic services with private quality information from a mechanism design perspective. By assigning specific reservation functions, multiple customers and multiple providers enunciate their preferences for these services. Once all demands and offers for electronic services are submitted, the mechanism determines an allocation that maximizes the sum of the aggregated preferences. However, the design of such mechanisms is difficult because of the following requirements: (1) Double-sided competition: Multiple competitive customers and multiple competitive providers must be matched together appropriately to generate maximal surplus, (2) QoS-awareness: The QoS desired by customers and the QoS offered by service providers must be internalized in the allocation mechanism, (3) private information: The mechanism must facilitate the allocation of electronic services for which any QoS information is unknown, (4) incentive compatibility: The mechanism has to provide adequate incentives to strategic individuals in order to ensure truthful bidding, (5) individual rationality: The participation decision in the mechanism must be voluntarily to all bidders, (6) budget balance: The mechanism must omit any independent intermediary in order to facilitate distributed decision-making among the participants, (7) optimality: The ultimate objective of the mechanism is to achieve an outcome that is optimal from a social welfare perspective. Standard impossibility theorems from mechanism design theory assert that meeting these requirements simultaneously is not attainable. In particular, ex post optimality cannot be attained when incentive compatibility, individual rationality, and budget balance are required as well. Therefore, the mechanism designer must decide about a viable tradeoff of these requirements. One possible compromise in the presence of privately known QoS is to derive a second-best mechanism that satisfies incentive compatibility, individual rationality, and budget balance. The outcome of such second-best mechanisms can be used to estimate the efficiency loss that must be tolerated in comparison to the first-best outcome. The objectives of this research are to (1) develop a second-best mechanism for allocating electronic services with private quality information and (2) study its efficiency properties in a set of simulation experiments to demonstrate its usefulness. All experiments imply that the asymptotic efficiency of the second-best mechanism is bounded away from 100% even for large markets. This finding is related to the economic concept of informational smallness, which is defined as the incremental impact of an participants QoS on the demand of an electronic service. In the proposed model, each provider offers a service of distinct QoS, and each customer demands a service of distinct QoS. It is this feature of differentiated service quality that prevents the participants from becoming informationally small as the market becomes large. If each participant’s private information about QoS follows the uniform distribution, the mechanism must tolerate an efficiency loss of more than 31% for an increasing number of customers and providers. In contrast, if private quality information is normally distributed among participants, this research finds that the mechanisms asymptotic inefficiency can be reduced to about 7% as the market size increases on both sides. With asymmetric, beta-distributed QoS, the mechanism arrives at an asymptotic efficiency of more than 91%. These findings are crucial to social planners because in designing service allocation with double-sided competition, they can obtain an accurate estimation of potential efficiency losses that arise from asymmetric information about QoS. On the other hand, the social planner can ensure that every allocation decision is made by the participants only. Hence, the emerging mechanism implementation eludes the need for an external, independent decision maker.
Flexibilisierung synchronisationsdefekter Beschaffungsketten : Entwurf eines fachkonzeptuellen Datenmodells für Logistik-Anwendungssysteme am Beispiel mechatronischer Produkte in der Automobilwirtschaft
(2014) Weiss, Daniel; Kirn, Stefan
The thesis investigates supply chains of mechatronic products in the automotive industry. In the past supply chains often were considered as stable, intercompany relations with long-running general agreements, predefined services and service flows. Due to the differentiation of demand in the form of complex mechatronic products, those stable relations, nowadays, are increasingly fading. From a customer’s point of view “mechatronic supply chains” result in increasing information acquisition costs, as procurement decision alternatives along mechatronic supply chains are often defective (e.g. inadequately assessable). This loss of control during the procurement process occurs due to the fact that mechatronic products as well as their components and parts show a high level of variety, different technologies and divergent life cycle stages. With regard to an entire mechatronic supply chain, the thesis describes this observation as a synchronization defect. Since the specific attributes of mechatronic supply chains do not allow the use of standard synchronization methods, the thesis aims to flexibilize the supply chain by means of IT application systems. Therefore, in particular, the thesis designs a conceptual data model which reduces information acquisition costs during the procurement initiation and processing along mechatronic supply chains, opening up shorter reaction times and/or additional options for action. Conceptual data models describe business problems from an IT application systems’ point of view. The design of a new data model is required due to the fact that the investigated conceptual data models showed only a small benefit related to the problem. The investigated models often were too unspecific and/or designed for other purposes (e.g. other tasks, industries).
Interdependenzen zwischen Planungsproblemen : ein multiagentenbasierter Koordinationsansatz
(2016) Schüle, Michael; Kirn, Stefan
In this thesis, interdependencies of planning problems in a supply chain in the context of a civil engineering logistic scenario are analysed. In this construction site scenario there are - within the scheduling of activities - the resource-constrained project scheduling problem (RCPSP) and - within the spatial distribution of activities - the asymmetric traveling salesman problem (ATSP) in an interdependence. The required coordination in order to resolve the impact caused by interdependencies of several planning is address as the core problem in this work. Since the individual planning processes do not consider the optimality of the planning of the respective other planning problem, there arise potentially in terms of the overall system and its objective function from the optimum material deviated solutions, although the isolated solutions of partial problems are optimally possibly. The coordination problem is exacerbated by the asymmetric distribution of information through the supply chain. From the perspective of the supply chain, this has the consequence that the service providers must be closely integrated as autonomous actors throughout all levels of the supply chain. In this context of the contractual relationship between the service providers with potentially different interests the coordination problem is extended to the coordination effort between the potentially autonomous providers. This decentralized decision situation must therefore be considered in the coordination of planning. Thereby software agents provide a suitable approach to represent autonomous, distributed, decentralized logistics systems. Especially the field of agent-based decentralized coordination for decentralized planning of decentralized plans offers this work a perspective to develop an approach for the coordination problem. Specifically, the agent-based coordination approach of the generalized partial global planning (GPGP) is used for the planning and coordination tasks, in which both the process of planning as well as the result of planning is distributed. Based on this method, a coordination approach along the interaction protocol engineering process model is developed. The result represents a coordination protocol. The evaluation is performed using the method of simulation as part of a multi-agent simulation. The simulation model is a reduced version of the construction site scenario. The aim of the simulation is to show that the presented multi-agent system and developed coordination protocol (1) are suitable for the optimization of the interdependent planning problems and (2) showing a sound system behaviour.
Investor sentiment in blogs : design of a classifier and validation by a portfolio simulation
(2016) Klein, Achim; Kirn, Stefan
How can investment recommendations available on the web significantly improve stock selection? This dissertation shows how online investment recommendations can automatically be analyzed, aggregated, and used to achieve a return above the market’s. To this respect, it is crucial to understand how investment recommendations affect returns. Therefore, the dissertation examines the effects of direct and indirect investment recommendations from blogs in the form of investor sentiments (i.e., opinions) on the expected development of stock prices. Blogs have made it possible for everyone to publish articles on the web. The studied blog platforms Seekingalpha and Blogspot host a wealth of semi-professional stock analyses, investor opinions, company rumors, and stock recommendations. The dissertation’s study uses about 77,000 articles from Seekingalpha and about 198,000 articles from Blogspot over a five-year period (2007-2011). A novel text classification method is developed for the automatic classification of blog articles in a positive vs. negative sentiment. To achieve a high classification accuracy, experiments were carried out to configure this method. The text classification method uses machine learning techniques, which learn from manually classified articles from a novel corpus. Using behavioral finance theory, hypotheses are developed about the effects of investor sentiments on a portfolios returns. To test these hypotheses, a monthly selection of stocks of the Dow Jones Industrial Average into a portfolio was simulated (i.e., backtested). The selection is made by means of the ranking of the monthly aggregated overall sentiment of all articles regarding a specific stock. The results show that a return above the market’s can be achieved with aggregated investor sentiments from the Seekingalpha platform. In most cases, the achieved return exceeds the return of a momentum portfolio based solely on past returns. For the platform Blogspot, results are weaker. Overall, it seems advisable for investors to select a small number of stocks based on the most positive and most negative monthly investor sentiments from professional blogs.
Multiagent resource allocation in service networks
(2014) Karänke, Paul; Kirn, Stefan
The term service network (SN) denotes a network of software services in which complex software applications are provided to customers by aggregating multiple elementary services. These networks are based on the service-oriented computing (SOC) paradigm, which defines the fundamental technical concepts for software services over electronic networks, e.g., Web services and, most recently, Cloud services. For the provision of software services to customers, software service providers (SPs) have to allocate their scarce computational resources (i.e., hardware and software) of a certain quality to customer requests. The SOC paradigm facilitates interoperability over organizational boundaries by representing business relationships on the software system level. Composite software services aggregate multiple software services into software applications. This aggregation is denoted as service composition. The loose coupling of services leads to SNs as dynamic entities with changing interdependencies between services. For composite software services, these dependencies exist across SN tiers; they result from the procurement of services, which are themselves utilized to produce additional services, and constitute a major problem for resource allocation in SNs. If these dependencies are not considered, the fulfillment of agreements may become unaccomplishable (overcommitment). Hence, the consideration of service dependencies is crucial for the allocation of service providers resources to fulfill customer requests in SNs. However, existing resource allocation methods, which could consider these dependencies -- such as combinatorial auctions with a central auctioneer for the whole SN -- are not applicable, since there are no central coordinating entities in SNs. The application of an allocation mechanism that does not consider these dependencies might negatively affect the actual service delivery; results are penalty payments as well as a damage to the reputation of the providers. This research is conducted in accordance to the design science paradigm in information system research. It is a problem-solving paradigm, which targets the construction and evaluation of IT artifacts. The objectives of this research are to develop and evaluate an allocation protocol, which can consider multi-tier service dependencies without the existence of central coordinating entities. Therefore, an interaction protocol engineering (IPE) perspective is applied to solve the problem of multi-tier dependencies in resource allocation. This approach provides a procedure model for designing interaction protocols for multiagent systems, and is closely related to the well-established area of communication protocol engineering. Automated resource allocation in SNs is analyzed in this research by representing the actors as autonomous software agents in the software system. The actors delegate their objectives to their software agents, which conduct the negotiations for service provision on their behalf. Thus, these agents communicate concerning the resource allocation; in this process, the sequence of communication interactions is crucial to the problem addressed. Interaction protocols define a structured exchange of defined messages between agents; they facilitate agent conversations. When multiple agents have to reach agreements by negotiation and bargaining, such as in case with allocating scarce resources, game theory provides means to formalize and analyze the most rational choice of actions for the interacting agents. Based on a formal framework for resource allocation in SNs, this research first performs a game-theoretic problem analysis; it is concerned with the existence, as well as the complexity of computing optimal allocations. In addition, Nash equilibria are analyzed for optimal allocations. Second, a distributed, auction-based allocation protocol, which prevents overcommitments and guarantees socially optimal allocations for single customer requests under certain assumptions, is proposed. Therefore, a game-theoretic model and an operationizable specification of the protocol are presented. Third, it is formally verified that the protocol enables multi-tier resource allocation and avoids overcommitments by proofs for the game-theoretic model and by model checking for the interaction protocol specification; using the model checker Spin, safety properties like the absence of deadlock are as well formally verified as the protocol enabling multi-tier resource allocation. Fourth, the efficacy and the benefits of the proposed protocol are demonstrated by multiagent simulation for concurrent customers. The experimental evaluation provides evidence of the protocols efficiency compared to the socially optimal allocation as a centralized benchmark in different settings, e.g., network topologies and different bidding policies.
Older adults’ internet use, outdoor activity, and the urban environment : empirical analysis
(2021) Schehl, Barbara; Kirn, Stefan
In an ageing society, information technology (IT) can be a powerful tool to support older adults in staying physically, mentally, and socially active. A growing number of studies show positive effects of Internet use on active ageing. Specifically, understanding the relationship between older adults’ Internet use and their participation in outdoor activity has gained attention. However, little is known about the role of specific online activities, as previous research mainly considered general Internet use, e.g., writing emails and browsing the Web, which does not provide a rationale for increased outdoor activity. Another important factor affecting outdoor activity is the urban environment in which older adults live, especially as the environment can either motivate or hinder participation in outdoor activity. Against this backdrop, the aim of this thesis is to get a better understanding of the association between Internet use, the urban environment, and participation in outdoor activity. By adopting an ecological perspective, this dissertation is first in comprehensively assessing this association. Three studies provide empirical insights into patterns of factors and contribute an enhanced understanding of the association between older adults’ Internet use, the urban environment, and outdoor activity. In summary, this thesis contributes to a relevant growth of knowledge on older adults’ Internet use and outdoor activity. The approach in this thesis uncovered important differences in the roles of individual factors as well as environmental factors in explaining older adults’ outdoor activity. Based on this contributions, important implications for research can be identified. For instance, future research can use the conceptual framework as a basis to investigate further types of outdoor activities and further types of online activities, such as social networks and messaging services which attain increasing relevance for the group of older adults. Further, this dissertation also has some practical implications, for example, as all three studies showed that socio-demographic predictors play important different roles, the findings are specifically relevant to identify specific groups of older adults. For instance, developers and providers can tailor their services to the needs of older adults, e.g., by designing responsive and barrier-free interfaces that adapt to individual capabilities in cognition, vision, and motor function. Concluding, older adults should be encouraged to go online because the Internet increasingly provides information about opportunities and offerings in the city. The findings of this dissertation provide a rationale for the development of online services that offer neighborhood information and ultimately support older adults to be active outdoors.
Qualitatives, räumliches Schließen zur Kollisionserkennung und Kollisionsvermeidung autonomer BDI-Agenten
(2011) Kirn, Stefan; Schüle, Michael
Die Trends und Veränderungen in der Logistik führen zu einem dezentralen Ansatz der Steuerungssysteme, um die Komplexität logistischer Systeme zu reduzieren. Softwareagenten, insbesondere BDI-Agenten (Belief-Desire-Intention), als Gegenstand dieser Arbeit, bieten aufgrund ihrer Eigenschaften geeignete Konzepte diesen Ansatz umzusetzen. Im räumlichen Kontext ist das Wissen der Agenten über ihre Umwelt häufig unsicher. Dieser Beitrag adressiert das Problem der autonomen, kollisionsfreien Bewegung von mehreren, interagierenden Agenten auf Basis von unsicherem Wissen im Kontext der Transportlogistik. Zu diesem Zweck bietet die Perspektive des qualitativ räumlichen Schließens geeignete Konzepte. Der Ansatz wird durch eine Multiagentensimulation in einem transportlogistischen Szenario, im Speziellen mit verschiedenen organisationstheoretischen Konzepten bezüglich des Verhaltens der Agenten ausgewertet.
Smarte Städtebauliche Objekte für eine adaptive Stadt : ein Verfahren der Künstlichen Intelligenz zur Erhöhung der Wohlfahrt
(2021) Hubl, Marvin; Kirn, Stefan
The objective of the thesis is the advancement of urban areas with intelligent information technology to enhance urban life. In doing so, social inclusion of people with motor impairments is of particular interest. At this, an important group is the set of older adults. The aspired advancement aims at enabling self-determined participation in urban life, and hence in social life, up until old age: Participation in social life essentially depends on the opportunities for self-determined exertion of outdoor activities. Motor impairments in old age lead to a perceived significant reduction of safety in urban areas and hence concerned people are worried of using the urban area. To counteract the resulting avoidance of outdoor activities the thesis pursues the approach to transform urban objects by means of technologies of the Internet of Things into novel so-called Smart Urban Objects that actively provide support in outdoor activities. Smart Urban Objects are equipped with sensors, actuators and information processing capabilities and can adapt to individual requirements of pedestrians. Due to age-correlated motor impairments, there are for example special requirements for seating. Besides the information technological transformation of single urban objects, there are furthermore important requirements for networked Smart Urban Objects. By means of intelligently coordinated, goal-oriented availability of the supportive functionalities of single Smart Urban Objects, the urban area as overall system is enabled for adaptivity with respect to pedestrians requirements. The thesis studies the conception for an adaptive city with a safety-engineering approach at the example of smart seating and develops a method for an intelligent coordination of the networked Smart Urban Objects. For an accurate allocation of single smart seats as public objects with respect to individual requirements a welfare criterion is applied which shall avoid unfairness. Using methods of Artificial Intelligence the thesis develops a heuristic procedure for finding a solution according to the welfare criterion. This implements aspects of the theory of justice by John Rawls that underlie the welfare criterion for the example of the urban area. A scenario-based simulation substantiates that the developed solution approach can effectively enhance the safety-oriented welfare in the urban area.
Theory-based knowledge acquisition for ontology development
(2016) Scheuermann, Andreas; Kirn, Stefan
This thesis concerns the problem of knowledge acquisition in ontology development. Knowledge acquisition is essential for developing useful ontologies but it is a complex and error-prone task. When capturing specific knowledge about a particular domain of interest, the problem of knowledge acquisition occurs due to linguistic, cognitive, modelling, and methodical difficulties. For overcoming these four difficulties, this research proposes a theory-based knowledge acquisition method. By studying the knowledge base, basic terms and concepts in the areas of ontology, ontology development, and knowledge acquisition are defined. A theoretical analysis of knowledge acquisition identifies linguistic, cognitive, modelling, and methodical difficulties, for which a survey of 15 domain ontologies provides further empirical evidence. A review of existing knowledge acquisition approaches shows their insufficiencies for reducing the problem of knowledge acquisition. As the underpinning example, a description of the domain of transport chains is provided. Correspondingly, a theory in business economics, i.e. the Contingency Approach, is selected. This theory provides the key constructs, relationships, and dependencies that can guide knowledge acquisition in the business domain and, thus, theoretically substantiate knowledge acquisition. Method construction uses an approach from the field of Method Engineering, which defines how to develop a tailored method with respect to specific requirements on method design, functionality, components, and the underlying assumptions. The development of the method for theory-based knowledge acquisition covers the specification of the (method and outcome) metamodel, activity model, outcomes, roles, and techniques. The evaluation comprises two descriptive approaches to demonstrate the proposed method’s utility. First, a criteria-based approach evaluates the method with respect to design-related, functional, and component-related requirements. Second, a scenario-based evaluation applies the method within a scenario from the domain of intermodal transport chains for acquiring knowledge to build a domain ontology. The contribution of this research is a theory-based knowledge acquisition method for ontology development. The application and usefulness of this method is demonstrated for a particular domain (transport chains) and uses a particular theory of business economics (the Contingency Approach).

Browse

Browsing Institut für Health Care & Public Management by Person "Kirn, Stefan"