Carla Delgado

Content recommendation and service costs in swarming systems

Recommendation systems and the performance of computer network systems have fundamental implications over each other. While recommendation systems impact system performance, the latter can be used to guide the former. In this paper, we study the interconnections between recommendation systems and the performance of the network. Focusing on swarming systems à la Bittorrent, we propose an analytical model to capture the revenue and the cost to a content provider as a function of the quality of its recommendations and the cost to serve the content. The model is then used to suggest heuristics on how to recommend content accounting for service costs and user preferences.

On Modalities for Vague Notions

We examine modal logical systems, with generalized operators, for the precise treatment of vague notions such as ‘often’, ‘a meaningful subset of a whole’, ‘most’, ‘generally’ etc. The intuition of ‘most’ as “all but for a ‘negligible’ set of exceptions” is made precise by means of filters. We examine a modal logic, with a new modality for a local version of ‘most’ and present a sound and complete axiom system. We also discuss some variants of this modal logic.

A recommendation approach for consuming linked open data

Abstract Most of linked open data (LOD) applications focus on the search and visualization of information, not efficiently using the links among objects in different data sources and the semantics of their relations. This work aims to create a LOD-consuming approach that uses recommendation techniques based on items’ description, their relations, users’ interests and social network. The proposed approach was instantiated by an application that uses movie related LOD. The results obtained in our experiments were promising: accuracy of the recommendations generated was equal or better, compared to other recommender algorithms used in conventional (not LOD) scenario.

Content Recommendation and Service Cost in P2P Systems

Recommendation systems and the performance of computer network systems have fundamental implications over each other. While recommendation systems impact system performance, the latter can be used to guide the former. In this paper, we study the interconnections between recommendation systems and the performance of the network. We propose an analytical model to capture the revenue and the cost to a content provider as a function of the quality of its recommendations and the cost to serve the content. The model is then used to suggest heuristics on how to recommend content accounting for service costs and user preferences.

Ontology Alignment with Weightless Neural Networks

In this paper, we present an ontology matching process based on the usage of Weightless Neural Networks (WNN). The alignment of ontologies for specific domains provides several benefits, such as interoperability among different systems and the improvement of the domain knowledge derived from the insights inferred from the combined information contained in the various ontologies. A WiSARD classifier is built to estimate a distribution-based similarity measure among the concepts of the several ontologies being matched. To validate our approach, we apply the proposed matching process to the knowledge domain of algorithms, software and computational problems, having some promising results.

A Trust and Reputation Framework for Game Agents: Providing a Social Bias to Computer Players

This work presents the application of trust and reputation models in the context of interactive games. This type of application aims to avoid that other players (humans or computers) can easily predict the behaviour of non-human players, and consequently loose interest in the game. In our approach, trust and reputation are mechanisms used to bring a social bias to non-human players, with the intention to emulate different types of social profiles. The main idea is to combine the social profile with the rationale the agent has regarding the game rules and game state, so that the agent uses both these traces (social profile and intelligent reasoning) in order to decide the next action to take. In this paper we present the conceptual model and the architecture of the proposed framework, and also report the results of a case study based on a game played by non-human players with different social profiles.

Proving Epistemic and Temporal Properties from Knowledge Based Programs

In this work we investigate two approaches for representing and reasoning about knowledge evolution in Multi-Agent System MAS. We use the language of the Logic of Time and Knowledge TKL [5] as a specification language for the whole system, and for each agent, we use Knowledge Based Programs KBP,[4]. We propose a method to translate a a KBP system into a set of TKL formulas. The translation method presented provides an strategy to model and prove properties of MAS without being attached to local or global representations, giving the alternative to switch from one representation to another. We also present a formal model of computation to our KBP system and prove the correctness our translation function w.r.t. this model.In order to illustrate usefulness of the translation method two examples are presented: the Muddy Children Puzzle and the Bit Exchange Protocol.

A Compositional Automata-based Approach for Model Checking Multi-Agent Systems

This paper addresses the issue of model checking knowledge in concurrent systems. The work benefits from many recent results on model checking and combined logics for time and knowledge, and focus on the way knowledge relations can be captured from automata-based system specifications. We present a formal language with compositional semantics and the corresponding Model Checking algorithms to model and verify Multi-Agent Systems (MAS) at the knowledge level, and a process for obtaining the global automaton for the concurrent system and the knowledge relations for each agent from a set of local automata that represents the behavior of each agent. Our aim is to describe a model suitable for model checking knowledge in a pre-defined way, but with the advantage that the knowledge relations for this would be extracted directly from the automata-based model.