Bayesian networks are a theoretically well-founded approach to represent large multi-variate probability distributions, and have proven useful in a broad range of applications. While several software tools for visualizing and editing Bayesian networks exist, they have important weaknesses when it comes to enabling users to clearly understand and compare conditional probability tables in the context of network topology, especially in large-scale networks. This paper describes a system for improving the ability for computers to work with people to develop intelligent systems through the construction of high-performing Bayesian networks. We describe NetEx, a tool developed as a Cytoscape plug-in, which allows a user to visually inspect and compare details concerning multiple nodes in a Bayesian network while maintaining awareness of their network context. It uses a "thought bubble line" to connect nodes in a graph representation and their internal information at the side of the graph. The tool seeks to improve the ability of experts to analyze and debug large Bayesian network models, and to help people to understand how alternative algorithms and Bayesian networks operate, providing insights into how to improve them.

Micro-task markets like Amazon MTurk enable online workers to provide human intelligence as Web-based on demand services (so called "people services"). Businesses facing large amounts of knowledge work can benefit from increased flexibility and scalability of their workforce but need to cope with reduced control of result quality. While this problem is well recognized, it has so far only rudimentarily been addressed by existing platforms and tools. In this paper, we present a flexible research toolkit which enables experiments with advanced quality management mechanisms for generic micro-task markets. The toolkit enables control of correctness and performance of task fulfillment by means of continuous sampling, dynamic majority voting and worker pooling. While we demonstrate its application and performance for an OCR scenario building on Amazon MTurk, the toolkit supports the development of advanced quality management mechanisms for a large variety of people service scenarios and platforms.

This paper introduces 'just enough' principles and 'systems engineering' approach to the practice of ontology development to provide a minimal yet complete, lightweight, agile and integrated development process, supportive of stakeholder management and implementation independence.

Traditionally routing has focused on finding shortest paths in networks with positive, static edge costs representing the distance between two nodes. Energy-optimal routing for electric vehicles creates novel algorithmic challenges, as simply understanding edge costs as energy values and applying standard algorithms does not work. First, edge costs can be negative due to recuperation, excluding Dijkstra-like algorithms. Second, edge costs may depend on parameters such as vehicle weight only known at query time, ruling out existing preprocessing techniques. Third, considering battery capacity limitations implies that the cost of a path is no longer just the sum of its edge costs. This paper shows how these challenges can be met within the framework of A* search. We show how the specific domain gives rise to a consistent heuristic function yielding an O(n 2 ) routing algorithm. Moreover, we show how battery constraints can be treated by dynamically adapting edge costs and hence can be handled in the same way as parameters given at query time, without increasing run-time complexity. Experimental results with real road networks and vehicle data demonstrate the advantages of our solution.

The design and implementation of a robotic chess agent is described. Shallow Blue, a competitor in the AAAI 2011 Small Scale Manipulation Challenge, is constructed with low-cost components including Lego NXT bricks and is programmed using Java and Lejos.

With the growth of online information, many people are challenged in finding and reading the information most important for their interests. From 2008-2010 we built an experimental personalized news system where readers can subscribe to organized channels of information that are curated by experts. AI technology was employed to radically reduce the work load of curators and to efficiently present information to readers. The system has gone through three implementation cycles and processed over 16 million news stories from about 12,000 RSS feeds on over 8000 topics organized by 160 curators for over 600 registered readers. This paper describes the approach, engineering and AI technology of the system.

Due to the rapid aging of the population, many technical solutions for the care of the elderly are being developed, often involving fall detection with accelerometers. We present a novel approach to fall detection with location sensors. In our application, a user wears up to four tags on the body whose locations are detected with radio sensors. This makes it possible to recognize the user’s activity, including falling any lying afterwards, and the context in terms of the location in the apartment. We compared fall detection using location sensors, accelerometers and accelerometers combined with the context. A scenario consisting of events difficult to recognize as falls or non-falls was used for the comparison. The accuracy of the methods that utilized the context was almost 40 percentage points higher compared to the methods without the context. The accuracy of pure location-based methods was around 10 percentage points higher than the accuracy of accelerometers combined with the context.

We report on an emerging application focused on the design of resilient long reach passive optical networks using combinatorial optimisation techniques. The objective of the application is to determine the optimal position and capacity of a set of metro nodes. We specifically consider dual parented networks whereby each customer must be associated with two metro nodes. An important property of such a placement is resilience to single node failure. Therefore excess capacity should be provided at each metro node in order to ensure that customers can be redistributed amongst the metro sites. Our application, as well as finding optimal node placements, can compute the minimum level of excess capacity on all metro nodes. In this paper we present three alternative approaches to optimal metro node placement.We present a detailed analysisof the impact of different placement approaches on the distribution of excess capacity throughout the network. We show that preferential distributions occur in practice, based on a case-study in Ireland. Finally we show that load and excess capacity provision are independent of each other.

The goal of my research is to develop and evaluate long-term declarative memory mechanisms that are effective and efficient across a variety of tasks.

The most prominent way to establish trust in online markets such as eBay are reputation systems that publish buyer feedback about a seller’s past behavior. These systems, however, critically rely on assumptions that are rarely met in realworld marketplaces: first, it is assumed that there are no reporting costs and no benefits from lying so that buyers honestly report their private experiences. Second, it is assumed that every seller is long-lived, i.e. will continue to trade on the marketplace indefinitely and, third, it is assumed that sellers cannot whitewash, i.e. create new accounts once an old one is ran down. In my thesis, I address all of these assumptions and design incentive-compatible trust mechanisms with minimal common knowledge requirements.