Udo Bub

Comparison of Four Approaches to Age and Gender Recognition for Telephone Applications

This paper presents a comparative study of four different approaches to automatic age and gender classification using seven classes on a telephony speech task and also compares the results with human performance on the same data. The automatic approaches compared are based on (1) a parallel phone recognizer, derived from an automatic language identification system; (2) a system using dynamic Bayesian networks to combine several prosodic features; (3) a system based solely on linear prediction analysis; and (4) Gaussian mixture models based on MFCCs for separate recognition of age and gender. On average, the parallel phone recognizer performs as well as Human listeners do, while loosing performance on short utterances. The system based on prosodic features however shows very little dependence on the length of the utterance.

Outline of a design science research process

Discussions about the body of knowledge of information systems, including the research domain, relevant perspectives and methods have been going on for a long time. Many researchers vote for a combination of research perspectives and their respective research methodologies; rigour and relevance as requirements in design science are generally accepted. What has been lacking is a formalisation of a detailed research process for design science that takes into account all requirements. We have developed such a research process, building on top of existing processes and findings from design research. The process combines qualitative and quantitative research and references well-known research methods. Publication possibilities and self-contained work packages are recommended. Case studies using the process are presented and discussed.

Artifact types in information systems design science – a literature review

Many information systems researchers designate their work as design science referring to the term “IT artifact” and the categorization systems that have emerged under this label Alas, there is no consensus at this point as to what the research output in design science is and what types of artifacts exist Using a widely accepted artifact typology would strengthen the scientific discussion and ease the categorization of contributions Based on a literature review of all DESRIST publications and a special MISQ issue on design science, we derived such a typology We identified eight relevant artifact types and related our typology to existing ones With this contribution, we hope to enable a discussion about what legitimate design science outputs and their main types are.

Getting closer: tailored human–computer speech dialog

This paper presents an advanced call center, which adapts presentation and interaction strategy to properties of the caller such as age, gender, and emotional state. User studies on interactive voice response (IVR) systems have shown that these properties can be used effectively to “tailor” services to users or user groups who do not maintain personal preferences, e.g., because they do not use the service on a regular basis. The adopted approach to achieve individualization of services, without being able to personalize them, is based on the analysis of a caller’s voice. This paper shows how this approach benefits service providers by being able to target entertainment and recommendation options. It also shows how this analysis at the same time benefits the customer, as it can increase accessibility of IVR systems to user segments which have particular expectations or which do not cope well with a “one size fits all” system. The paper summarizes the authors’ current work on component technologies, such as emotion detection, age and gender recognition on telephony speech, and presents results of usability and acceptability tests as well as an architecture to integrate these technologies in future multi-modal contact centers. It is envisioned that these will eventually serve customers with an avatar representation of an agent and tailored interaction strategies, matching powerful output capabilities with advanced analysis of the user’s input.

Proposal for Components of Method Design Theories

Gregor and Jones have proposed components for design theories, building on theory concepts from behavioural sciences and prior publications. Their design theory structure addresses IT artefacts in general, not specific to any type, such as constructs, models, methods or instantiations. Their work is an important contribution to the academic discussion of design theories. The authors are building on this and believe that specialised design theory structures for different types of artefacts further increases utility, usability and acceptance of the components for both academia and practise. They have analysed each of the components published by Gregor and Jones and proposed refinements specific to method design artefacts wherever applicable. For each component, they derive evaluation criteria and present examples of method publications fulfilling the criteria. They argue that by presenting method design theories according to this structure the contribution of method design artefacts to the body of knowledge will increase.

Deriving SOA Evaluation Metrics in an Enterprise Architecture Context

Service oriented architectures (SOA) are becoming reality in a corporate environment. Rather than pure technology improvements SOA intends to increase manageability and sustainability of IT systems and to better align business requirements and technology implementations. Following these intentions current SOA research is focused on the management of SOA. In this paper we present a method to identify metrics for an evaluation of SOA. Therefore we adopted the goal/question/metrics method to SOA specifics. Since SOA is not limited to questions of technology we will show where SOA might be located in an enterprise architecture (EA) context. Based on the assumption that SOA follows different goals on different levels of EA abstraction, the paper shows, how these goals can be developed to metrics which can be consolidated in a measurement program.

Benefits of SOA: Evaluation of an implemented scenario against alternative architectures

Service-oriented architecture (SOA) is based on a design principle for business software systems. The SOA movement has taken up momentum, especially because of widespread middleware vendor support, e.g. by IBM, HP, BEA, Microsoft, Sun, Software AG, TIBCO, Oracle, SAP and others. Still answers to questions of how the target architecture should look like and what the advantages over other architectures are remain open. We have implemented a scenario based on free software, which consists of a well-defined service-oriented target architecture. In this article, we compare the used service-oriented approach with other state-of-theart architectures to show that changeability and efficiency improve compared to alternative architectures.

The Future of Telecommunications

For two decades the telecommunications industry has been facing severe changes caused by the growing dominance of the Internet architecture as general purpose communications infrastructure, mass media, and also by new generations for mobile communications. This development will continue: new network technologies (e.g. Next Generation Networks, Long Term Evolution) will change cost and market structures, innovation cycles for products and technologies will become shorter, value chains will change. This is illustrated by vendors entering online service markets or offering their own network communication through so-called managed services. The change of operations system platforms and the emergence of application market places, especially in the mobile domain with the so-called apps, influence the future development considerably. Established network operators, on the other hand, start to specify individual consumer devices; classic computer manufacturers break into the telecommunications market by way of consumer devices as well as new services. The acquisition of Motorola Mobility by Google is only one further indicator that this ecosystem is subject to fundamental change. As sign of emancipation from services and infrastructure, numerous new enterprises offer carrier agnostic telecommunications and data services without owning the infrastructure, so-called over the top providers. Along with this a gradual shift of shares of the value chain with neighboring industries such as the media industry can be observed. Information technology (IT) and the next generation of telecommunications – based on the Internet protocol IP supporting services as well as the infrastructure – will get more and more similar. This is why many IT paradigms will become transferable to the telecommunications infrastructure. The growing convergence is coined by the term information and communications technology (ICT). Operators are often faced with the decision if they should build their future on pure network operations or seek premiums through new types of network specific services (using so-called enablers) or network agnostic services, respectively. The development of the telecommunications industry and the impulses and innovations caused by it in business and society depend considerably from the institutional guidelines, especially from regulations. The controversially discussed topic of network neutrality, i.e. the equal treatment of all Internet traffic allows for different views on the same thing: network neutrality can be interpreted as productive or counterproductive to innovation. In any case the internal operations of the involved players have to be designed very effectively and efficiently to reach a higher degree of automation. This is where reference models such as the Enhanced Telecom Operations Map (eTOM) help to reach a standardized course of action when needed. The harmonization between business processes and enterprise information systems along with a well tuned enterprise architecture plays a crucial role here. Further new challenges offer opportunities, but require a thorough techno-economical investigation. The current network transformation to an IP-based infrastructure of the Next Generation Networks offers potential for increased efficiency, but also requires high investments of billions of Euros with various decision alternatives. The started rollout of fiber optics connection right into homes (Fiber To The Home) will not only offer bandwidth of well over 100 Mbit/s to end customers, but will also give operators new opportunities to shape the network architecture, as well as to further automate the internal business processes up to a zero touch (i.e. fully automated) provisioning of services. This enormous build-out will not be done by former monopolies (so-called incumbents) alone, but will be shared. The business models of infrastructure suppliers will require a mutual ability for wholesale. This implies that in the future, incumbents might also lease some access capacity from other operators, which was not the case before. In addition to that it turns out that in some geographical regions public spending might be necessary (e.g. by means of Public-Private-Partnerships) in order to achieve full coverage of high speed broadband access. It all sums up to the fact that innovative products and applications have to be found that are convincing to the customer both in functionality and price. As a matter of

Platform-as-a-Service for business customers

Telecommunication providers are still searching for business models beyond the bit pipe. While added-value services like Software-as-a-Service (SaaS) become more and more common for private customers as well as small and medium enterprises, in this paper we demonstrate the opportunities of Platform-as-a-Service offerings for specific business domains. By offering such platforms as cloud services, telecommunication providers can leverage their existing communications infrastructure to provide added-value services. Customer-specific SaaS can be created rapidly on such platforms, enabling customized business applications on a rental pricing scheme.

Vorschlag für Komponenten von Methodendesigntheorien Steigerung der Nutzbarkeit von Methodendesignartefakten

ZusammenfassungGregor und Jones haben, aufbauend auf Theoriekonzepten der Verhaltensforschung und früheren Publikationen, Komponenten von Designtheorien vorgeschlagen. Ihre Designtheoriestruktur behandelt IT-Artefakte im Allgemein und ist nicht auf einen bestimmten Typ wie z. B. Konstrukte, Modelle, Methoden oder Instanziierungen zugeschnitten. Ihre Arbeit ist ein wichtiger Beitrag zur wissenschaftlichen Diskussion über Designtheorien. Die Autoren bauen darauf auf und nehmen an, dass spezialisierte Designtheoriestrukturen für unterschiedliche Artefakttypen die Nützlichkeit, Nutzbarkeit und Akzeptanz der Komponenten für Wissenschaft und Praxis weiter steigern. Sie haben alle von Gregor und Jones publizierten Komponenten analysiert und schlagen spezifische Verfeinerungen für Methodendesignartefakte vor, wo es passend erscheint. Für jede Komponente leiten sie Evaluationskriterien ab und präsentieren Beispiele publizierter Methoden, die die Kriterien erfüllen. Ihrer Ansicht nach wird der Beitrag von Methodendesignartefakten zum Wissensbasis durch die Formulierung von Methodendesigntheorien nach der vorgeschlagenen Struktur gesteigert.AbstractGregor and Jones have proposed components for design theories, building on theory concepts from behavioural sciences and prior publications. Their design theory structure addresses IT artefacts in general, not specific to any type, such as constructs, models, methods or instantiations. Their work is an important contribution to the academic discussion of design theories. The authors are building on this and believe that specialised design theory structures for different types of artefacts further increases utility, usability and acceptance of the components for both academia and practice. They have analysed each of the components published by Gregor and Jones and proposed refinements specific to method design artefacts wherever applicable. For each component, they derive evaluation criteria and present examples of method publications fulfilling the criteria. They argue that by presenting method design theories according to this structure the contribution of method design artefacts to the body of knowledge will increase.