Marco Viviani

A Survey on User Modeling in Multi-application Environments

Sharing heterogeneous data among distributed environments in a user-centric way represents today the main challenge for personalization. In recent years several techniques have been proposed to support user modeling for cross-system personalization, following two main approaches. A first group of researchers base their solutions on the use of a top-down approach, involving standard ontologies or unified user models (standardization-based user modeling); in a second research direction, a bottom-up approach based on mappings between different user model representations is the envisaged solution (mediation-based user modeling). The aim of this paper is to discuss general problems connected to user modeling in multiapplication environments and to provide a short survey on current research in this field. Based on this, we briefly draw out some further research.

Bottom-Up Extraction and Trust-Based Refinement of Ontology Metadata

We present a way of building ontologies that proceeds in a bottom-up fashion, defining concepts as clusters of concrete XML objects. Our rough bottom-up ontologies are based on simple relations like association and inheritance, as well as on value restrictions, and can be used to enrich and update existing upper ontologies. Then, we show how automatically generated assertions based on our bottom-up ontologies can be associated with a flexible degree of trust by nonintrusively collecting user feedback in the form of implicit and explicit votes. Dynamic trust-based views on assertions automatically filter out imprecisions and substantially improve metadata quality in the long run

Knowledge Extraction from Semi-structured Data Based on Fuzzy Techniques

In this work we propose a fuzzy technique to compare XML documents belonging to a semi-structured flow and sharing a common vocabulary of tags. Our approach is based on the idea of representing documents as fuzzy bags and, using a measure of comparison, evaluating structural similarities be-tween them. Then we suggest how to organize the extracted knowledge in a class hierarchy, choosing a technique related to the domain of interest, later to be converted into a user ontology.

Relating RSS News/Items

Merging related RSS news (coming from one or different sources) is beneficial for end-users with different backgrounds (journalists, economists, etc.), particularly those accessing similar information. In this paper, we provide a practical approach to both: measure the relatedness, and identify relationships between RSS elements. Our approach is based on the concepts of semantic neighborhood and vector space model, and considers the content and structure of RSS news items.

Semantic-based Merging of RSS Items

Merging XML documents can be of key importance in several applications. For instance, merging the RSS news from same or different sources and providers can be beneficial for end-users in various scenarios. In this paper, we address this issue and explore the relatedness measure between RSS elements. We show here how to define and compute exclusive relations between any two elements and provide several predefined merging operators that can be extended and adapted to human needs. We also provide a set of experiments conducted to validate our approach.

Chapter 13 Bottom-up extraction and maintenance of ontology-based metadata

Abstract In this chapter, several flexible techniques aimed at extracting, maintaining and enriching semantic-web style metadata are discussed. Such techniques were designed for being applied in the framework of dynamic Communities of Practice (CoP) interactions. Namely, we present a way of building ontologies that proceeds in a bottom-up fashion, defining concepts as clusters of concrete objects. Unlike huge, “supply-side” normative ontologies, our bottom-up ontologies are based on use of implicit and, therefore, parsimonious part-whole and is-a relations. This makes them suitable for the ad-hoc style of conceptualization used within communities of practice and peer-to-peer (P2P) communities. Also we discuss how metadata based on bottom-up ontologies can be associated with a flexible degree of trust by collecting user feedback. Our bottom-up extraction method complements current practice, where, as a rule, ontologies are built top-down. It is not claimed that bottom-up construction is a generally valid recipe; rather, the approach is intended to enrich the ontology developers palette when designing and implementing Semantic Web applications.

Adding a Trust Layer to Semantic Web Metadata

We outline the architecture of a modular Trust Layer that can be superimposed to generic semantic Web-style metadata generation facilities. Also, we propose an experimental setting to generate and validate trust assertions on classification metadata generated by different tools (including our ClassBuilder) after a process of metadata standardization. Our experimentation is aimed at validating the role of our Trust Layer as a non-intrusive, user-centered quality improver for automatically generated metadata.

Measuring Inference Exposure in Outsourced Encrypted Databases

Database outsourcing is becoming increasingly popular introducing a new paradigm, called database-as-a-service, where an encrypted clients database is stored at an external service provider. Existing proposals for querying encrypted databases are based on the association, with each encrypted tuple, of additional indexing information obtained from the plaintext values of attributes that can be used in the queries. However, the relationship between indexes and data should not open the door to inference and linking attacks that can compromise the protection granted by encryption.

Adding a peer-to-peer trust layer to metadata generators

In this paper we outline the architecture of a peer-to-peer Trust Layer that can be superimposed to metadata generators producing classifications, like our ClassBuilder and BTExact’s iPHI tools. Different techniques for aggregating trust values are also discussed. Our ongoing experimentation is aimed at validating the role of a Trust Layer as a non-intrusive, peer-to-peer technique for improving quality of automatically generated metadata.

Emergent Semantics in Distributed Knowledge Management

Organizations and enterprises have developed complex data and information exchange systems that are now vital for their daily operations. Currently available systems, however, face a major challenge. On todays global information infrastructure, data semantics is more and more context- and time-dependent, and cannot be fixed once and for all at design time. Identifying emerging relationships among previously unrelated information items (e.g., during data interchange) may dramatically increase their business value. This chapter introduce and discuss the notion of Emergent Semantics (ES), where both the representation of semantics and the discovery of the proper interpretation of symbols are seen as the result of a selforganizing process performed by distributed agents, exchanging symbols and adaptively developing the proper interpretation via multi-party cooperation and conflict resolution. Emergent data semantics is dynamically dependent on the collective behaviour of large communities of agents, which may have different and even conflicting interests and agendas. This is a research paradigm interpreting semantics from a pragmatic prospective. The chapter introduce this notion providing a discussion on the principles, research area and current state of the art.