In this paper we analyse Belief Propagation over a Gaussian model in a dynamic environment. Recently, this has been proposed as a method to average local measurement values by a distributed protocol ("Consensus Propagation", Moallemi & Van Roy, 2006), where the average is available for read-out at every single node. In the case that the underlying network is constant but the values to be averaged fluctuate ("dynamic data"), convergence and accuracy are determined by the spectral properties of an associated Ruelle-Perron-Frobenius operator. For Gaussian models on Erdos-Renyi graphs, numerical computation points to a spectral gap remaining in the large-size limit, implying exceptionally good scalability. In a model where the underlying network also fluctuates ("dynamic network"), averaging is more effective than in the dynamic data case. Altogether, this implies very good performance of these methods in very large systems, and opens a new field of statistical physics of large (and dynamic) information systems.

The main problem addressed by this project is that of the creation of sustainable and meaningful long-term human robot relationships. This is a most important problem towards our ultimate goal of human-robot symbiosis, i.e. harmonious and mutually beneficial living together of the two species. In the shorter term, this is an important problem towards the successful application of robots to numerous areas: disabled and elderly assistance / companionship, supporting education, and more. So far, empirical investigations have shown that we have not advanced significantly yet towards its solution: Although existing robotic systems are interesting to interact with in the short term, it has been shown that after some weeks of quasi-regular encounters, humans gradually lose their interest, and meaningful longer-term human-robot relationships are not established. For example, in the case of Robovie [1], there was a steady and significant decrease in the total time of interaction of the robot with humans over six months - interest had worn off.

In this paper we formulate the problem of inference under incomplete information in very general terms. This includes modelling the process responsible for the incompleteness, which we call the incompleteness process. We allow the process' behaviour to be partly unknown. Then we use Walley's theory of coherent lower previsions, a generalisation of the Bayesian theory to imprecision, to derive the rule to update beliefs under incompleteness that logically follows from our assumptions, and that we call conservative inference rule. This rule has some remarkable properties: it is an abstract rule to update beliefs that can be applied in any situation or domain; it gives us the opportunity to be neither too optimistic nor too pessimistic about the incompleteness process, which is a necessary condition to draw reliable while strong enough conclusions; and it is a coherent rule, in the sense that it cannot lead to inconsistencies. We give examples to show how the new rule can be applied in expert systems, in parametric statistical inference, and in pattern classification, and discuss more generally the view of incompleteness processes defended here as well as some of its consequences.

The need for efficient monitoring of spatio-temporal dynamics in large environmental applications, such as the water quality monitoring in rivers and lakes, motivates the use of robotic sensors in order to achieve sufficient spatial coverage. Typically, these robots have bounded resources, such as limited battery or limited amounts of time to obtain measurements. Thus, careful coordination of their paths is required in order to maximize the amount of information collected, while respecting the resource constraints. In this paper, we present an efficient approach for near-optimally solving the NP-hard optimization problem of planning such informative paths. In particular, we first develop eSIP (efficient Single-robot Informative Path planning), an approximation algorithm for optimizing the path of a single robot. Hereby, we use a Gaussian Process to model the underlying phenomenon, and use the mutual information between the visited locations and remainder of the space to quantify the amount of information collected. We prove that the mutual information collected using paths obtained by using eSIP is close to the information obtained by an optimal solution. We then provide a general technique, sequential allocation, which can be used to extend any single robot planning algorithm, such as eSIP, for the multi-robot problem. This procedure approximately generalizes any guarantees for the single-robot problem to the multi-robot case. We extensively evaluate the effectiveness of our approach on several experiments performed in-field for two important environmental sensing applications, lake and river monitoring, and simulation experiments performed using several real world sensor network data sets.

This paper presents a tree-to-tree transduction method for sentence compression. Our model is based on synchronous tree substitution grammar, a formalism that allows local distortion of the tree topology and can thus naturally capture structural mismatches. We describe an algorithm for decoding in this framework and show how the model can be trained discriminatively within a large margin framework. Experimental results on sentence compression bring significant improvements over a state-of-the-art model.

This paper describes a method for the automatic inference of structural transfer rules to be used in a shallow-transfer machine translation (MT) system from small parallel corpora. The structural transfer rules are based on alignment templates, like those used in statistical MT. Alignment templates are extracted from sentence-aligned parallel corpora and extended with a set of restrictions which are derived from the bilingual dictionary of the MT system and control their application as transfer rules. The experiments conducted using three different language pairs in the free/open-source MT platform Apertium show that translation quality is improved as compared to word-for-word translation (when no transfer rules are used), and that the resulting translation quality is close to that obtained using hand-coded transfer rules. The method we present is entirely unsupervised and benefits from information in the rest of modules of the MT system in which the inferred rules are applied.

This paper presents a new method for inferring the semantic properties of documents by leveraging free-text keyphrase annotations. Such annotations are becoming increasingly abundant due to the recent dramatic growth in semi-structured, user-generated online content. One especially relevant domain is product reviews, which are often annotated by their authors with pros/cons keyphrases such as ``a real bargain'' or ``good value.'' These annotations are representative of the underlying semantic properties; however, unlike expert annotations, they are noisy: lay authors may use different labels to denote the same property, and some labels may be missing. To learn using such noisy annotations, we find a hidden paraphrase structure which clusters the keyphrases. The paraphrase structure is linked with a latent topic model of the review texts, enabling the system to predict the properties of unannotated documents and to effectively aggregate the semantic properties of multiple reviews. Our approach is implemented as a hierarchical Bayesian model with joint inference. We find that joint inference increases the robustness of the keyphrase clustering and encourages the latent topics to correlate with semantically meaningful properties. Multiple evaluations demonstrate that our model substantially outperforms alternative approaches for summarizing single and multiple documents into a set of semantically salient keyphrases.

Since its birth, the web provided many ways of interacting between us [6], revealing huge social network structures [17], a phenomenon amplified by web 2.0 applications [11]. Researchers extracted social networks from emails, mailinglist archives, hyperlink structure of homepages, cooccurrence of names in documents and from the digital traces created by web 2.0 application usages [9]. Facebook, LinkedIn or Myspace provide huge amounts of structured network data. The emergence of the semantic web approaches led researchers to build models of such online interactions using ontologies like FOAF, SIOC or SCOT. This paper starts with a brief state of the art on these enhanced RDF-based representations. We will see that the graphs built using these ontologies have a great potential that is not fully exploited so far. Then, we present a new framework for applying SNA to RDF representations of social data. In particular, the use of graph models underlying RDF and SPARQL extensions enables us to extract efficiently and to parameterize the classic SNA features directly from these representations.

Inspired by Hofstadter's Coffee-House Conversation (1982) and by the science fiction short story SAM by Schattschneider (1988), we propose and discuss criteria for non-mechanical intelligence. Firstly, we emphasize the practical need for such tests in view of massively multiuser online role-playing games (MMORPGs) and virtual reality systems like Second Life. Secondly, we demonstrate Second Life as a useful framework for implementing (some iterations of) that test.

Real world datasets are sparse, dirty and contain hundreds of items. In such situations, discovering interesting rules (results) using traditional frequent itemset mining approach by specifying a user defined input support threshold is not appropriate. Since without any domain knowledge, setting support threshold small or large can output nothing or a large number of redundant uninteresting results. Recently a novel approach of mining only N-most/Top-K interesting frequent itemsets has been proposed, which discovers the top N interesting results without specifying any user defined support threshold. However, mining interesting frequent itemsets without minimum support threshold are more costly in terms of itemset search space exploration and processing cost. Thereby, the efficiency of their mining highly depends upon three main factors (1) Database representation approach used for itemset frequency counting, (2) Projection of relevant transactions to lower level nodes of search space and (3) Algorithm implementation technique. Therefore, to improve the efficiency of mining process, in this paper we present two novel algorithms called (N-MostMiner and Top-K-Miner) using the bit-vector representation approach which is very efficient in terms of itemset frequency counting and transactions projection. In addition to this, several efficient implementation techniques of N-MostMiner and Top-K-Miner are also present which we experienced in our implementation. Our experimental results on benchmark datasets suggest that the NMostMiner and Top-K-Miner are very efficient in terms of processing time as compared to current best algorithms BOMO and TFP.