Automatic detection of relevant terms in medical reports is useful for educational purposes and for clinical research. Natural language processing techniques can be applied in order to identify them. The main goal of this research is to develop a method to identify whether medical reports of imaging studies (usually called radiology reports) written in Spanish are important (in the sense that they have non-negated pathological findings) or not. We also try to identify which finding is present and if possible its relationship with anatomical entities.

This paper presents two frameworks that generalize Conditional Preference networks (CP-nets). The first generalization is the LCP-theory, first order logic theory that provides a rich framework to express preferences. The the second generalization, the PCP-networks, is a probabilistic generalization of CP-nets that models conditional preferences with uncertainty.

The creation of novel recommendation algorithms for social networks is currently struggling with the volume of available data originating in such environments. Given that social networks can be modeled as graphs, a distributed graph-oriented support to exploit the computing capabilities of clusters arises as a necessity. In this thesis, a platform for graph storage and processing named Graphly is proposed along with GraphRec, an API for easy specification of recommendation algorithms. Graphly and GraphRec hide distributed programming concerns from the user while still allowing fine-tuning of the remote execution. For example, users may customize an algorithm execution using job distribution strategies, without modifying the original code. GraphRec also simplifies the design of graph-based recommender systems by implementing well-known algorithms as “primitives” that can be reused.

My thesis is largely focused on the parallelisation of UCT (and other Best-First Search techniques) and the ramifications of doing so. I have identified issues with chunking in UCT, created by some forms of parallelisation, and developed a solution to this involving buffering of simulations that appear “out of order” and reevaluation of propagation data. I have developed a technique for scalable distribution of both tree data and computation across a large scale compute cluster. The context of most of my work is General Game Playing, but the techniques themselves are largely agnostic to domain.

Reinforcement learning algorithms typically require too many `trial-and-error' experiences before reaching a desirable behaviour. A considerable amount of ongoing research is focused on speeding up this learning process by using external knowledge. We contribute in several ways, proposing novel approaches to transfer learning and learning from demonstration, as well as an ensemble approach to combine knowledge from various sources.

Semi-Supervised Learning (SSL) techniques have become very relevant since they require a small set of labeled data. In this scenario, graph-based SSL algorithms provide a powerful framework for modeling manifold structures in high-dimensional spaces and are effective for the propagation of the few initial labels present in training data through the graph. An important step in graph-based SSL methods is the conversion of tabular data into a weighted graph. The graph construction has a key role in the quality of the classification in graph-based methods. Nevertheless, most of the SSL literature focuses on developing label inference algorithms without studying graph construction methods and its effect on the base algorithm performance. This PhD project aims to study this issue and proposes new methods for graph construction from flat data and improves the performance of the graph-based algorithms.

When an optimal solution is not required, satisficing search methods such as greedy best-first search are often used to find solutions quickly. In work on satisficing search, there has been substantial attention devoted to how to solve problems associated with local minima or plateaus in the heuristic function. One technique that has been shown to be quite promising is using an alternative heuristic function that does not estimate cost-to-go, but rather estimates distance-to-go. There is currently little beyond intuition to explain its superiority. We begin by empirically showing that the success of the distance-to-go heuristic appears related to its having smaller local minima. We then discuss a reasonable theoretical model of heuristics and show that, under this model, the expected size of local minima is higher for a cost-to-go heuristic than a distance-to-go heuristic, offering a possible explanation as to why distance-to-go heuristics tend to outperform cost-to-go heuristics.

Stream reasoning is the task of continuously deriving conclusions on streaming data. As a research theme, it is targeted by different communities which emphasize different aspects, e.g., throughput vs. expressiveness. This thesis aims to advance the theoretical foundations underlying diverse stream reasoning approaches and to convert obtained insights into a prototypical expressive rule-based reasoning system that is lacking to date.

Most work in heuristic search considers problems where a low cost solution is preferred (MIN problems). In this paper, we investigate the complementary setting where a solution of high reward is preferred (MAX problems). Example MAX problems include finding a longest simple path in a graph, maximal coverage, and various constraint optimization problems. We examine several popular search algorithms for MIN problems and discover the curious ways in which they misbehave on MAX problems. We propose modifications that preserve the original intentions behind the algorithms but allow them to solve MAX problems, and compare them theoretically and empirically. Interesting results include the failure of bidirectional search and close relationships between Dijkstra's algorithm, weighted A*, and depth-first search.

Hunger is a major problem even in developed countries like Australia. We are working with a social startup, Foodbank Local, and local charities at distributing donated food more efficiently. This food must first be allocated to these charities and then delivered to the end customers. In this abstract, we give a formulation of this real-world online fair division problem that the food banks face every day. The products arrive during the day and are indivisible. As a very first step, we focus in here on designing simple mechanisms allocating the food more efficiently. In future, we also plan on investigating more closely the frontier between the allocation and the transportation frameworks within this mixed setting. For instance, shall we dispatch the items as soon as they arrive or shall we apply a given waiting strategy?