Search

In social media, such as blogs, since the content naturally evolves over time, it is hard or in many cases impossible to organize the content for effective navigation. Thus, one commonly has to resort to simple tools, such as tags and tag clouds, for presenting frequently used keywords to users to provide them at least some high level idea about the content of a given set of social media entries. Most visualizations of tag clouds vary the sizes of the fonts to differentiate important tags from those that are less important. We propose an alternative "contextual-layout" method, TMine, for analyzing and presenting tags that are extracted from textual content. In TMine tags are first mapped onto a latent semantic space. Then, TMine analyzes the relationships between tags relying on an extended boolean interpretation of the semantic space. The tag cloud is condensed into a hierarchy in a way that captures contextual relationships between tags: in particular, descendant terms in the hierarchy occur within the context defined by the ancestor terms. This provides a mechanism for navigation within the tag space as well as for classification of the text documents based on the contextual structure implied by the tags.

We present the tagFlake system, which supports semantically informed navigation within a tag cloud. tagFlake relies on TMine for organizing tags extracted from textual content in hierarchical organizations, suitable for navigation, visualization, classification, and tracking. TMine extracts the most significant tag/terms from text documents and maps them onto a hierarchy in such a way that descendant terms are contextually dependent on their ancestors within the given corpus of documents. This provides tagFlake with a mechanism for enabling navigation within the tag space and for classification of the text documents based on the contextual structure captured by the created hierarchy. tagFlake is language neutral, since it does not rely on any natural language processing technique and is unsupervised.

Ordering objects with respect to their various relevant properties before and during processing is a basic step in many multimedia mining problems. Examples include mining frequent patterns in sensory data and mining popularity orders in digital television. Designing multimedia and multi-modal mining techniques for complex and adaptive systems, requires the capability of dealing with rankings of diverse collection of inputs and outputs of a complex mining task, in a uniform, declarative manner. In this paper, we present a model and algebra which treat ranks of the media as first class objects to support complex mining tasks. We model each mining task as an algebraic combination of multiple subtasks, thus providing a declarative framework in which the ranked results returned by individual subtasks are combined under appropriate semantics. We also present a novel order distance function, which enables partitioning and aggregation support for mining.

A critical reality in data integration is that knowledge from different sources may often be conflicting with each other. Conflict resolutioncan be costly and, if done without proper context, can be ineffective. In this paper, we propose a novel query-driven and feedback-based approach (FICSR1) to conflict resolution when integrating data sources. In particular, instead of relying on traditional model based definition of consistency, we introduce a ranked interpretation. This not only enables FICSR to deal with the complexity of the conflict resolution process, but also helps achieve a more direct match between the users' (subjective) interpretation of the data and the system's (objective) treatment of the available alternatives. Consequently, the ranked interpretation leads to new opportunities for bi-directional (data informsover ↔ user) feedback cycle for conflict resolution: given a query, (a) a preliminary ranking of candidate results on data can inform the user regarding constraints critical to the query, while (b) user feedback regarding the ranks can be exploited to inform the system about user's relevant domain knowledge. To enable this feedback process, we develop data structures and algorithms for efficient off-line conflict/agreement analysis of the integrated data as well as for on-line query processing, candidate result enumeration, and validity analysis. The results are brought together and evaluated in the FICSR system.

We present the QUery-driven Exploration of Semistructured dataand meta-data with conflicTs and partial knowledge (QUEST) system for supporting the integration of scientific data and taxonomies in the presence of misalignments and conflicts. QUEST relies on a novel constraint-based data model that captures both value and structural conflicts and enables researchers to observe and resolve such misalignments in the integrated data by considering the context provided by the data requirements of given research questions.

Multimedia documents are collections of media objects, synchronized by means of sets of temporal and spatial constraints. Any multimedia document definition is valid as long as the referred media objects are available and the constraints are satisfiable. Document validity depends on the context in which the document has to be presented. In this paper, we introduce a framework to characterize context adaptation, in the presence of both physical and user oriented context requirements. We define semantically equivalent presentation fragments as alternative to undeliverable ones. In the absence of equivalence, undeliverable media are replaced with candidates that minimize the loss of information/quality in the presentation.

ARIA, ARchitecture for Interactive Arts, is a middleware to process, filter, and fuse sensory inputs and actuate responses in real-time. An ARIA media processing workflow describes how the data sensed through media will be processed and what audio-visual responses will be actuated. Each object streamed between ARIA processing components is subject to transformations, as described by a media workflow graph. The media capture and processing components, such as media filters and fusion operators, are programmable and adaptable; i.e, the delay, size, frequency, and quality/precision characteristics of individual operators can be controlled via a number of parameters. In this paper, we present the underlying model which captures the dynamic nature of the ARIA media processing workflows. This model enables design time verification, optimization, and runtime adaptation.

In this paper we discuss issues concerning the consistent retrieval of parts of multimedia presentations from multimedia repositories. We introduce a class of multimedia presentations made of independent and synchronized media and discuss retrieval requirements of presentation fragments. Then we discuss a retrieval model capable of reconstructing the fragments of a presentation from the atomic components returned by the execution of queries to multimedia presentation repositories. The retrieval model is based on an automaton that formally describes the presentation states entered by the events that trigger media playback. Retrieving a consistent fragment corresponds to building a new presentation with all the media related to the retrieved ones, with their original structural and synchronization relationships.

Despite the fact that thousands of applications manipulate plans, there has been no work to date on managing large databases of plans. In this paper, we first propose a formal model of plan databases. We describe important notions of consistency and coherence for such databases. We then propose a set of operators similar to the relational algebra to query such databases of plans.

In this paper we present a fuzzy approach for image databases. We exploit the concept of NF2 relational model as a foundation for building image catalogues containing the semantic description of a given image database. New algebraic operators are defined in order to capture the fuzziness related to the semantic descriptors of an image. We compare our model to the First Normal Form annotated relation model, and show that in a number of interesting cases they can be considered equivalent, from the operational point of view, but in general NF2 relational model is more powerful, and provides a more suitable framework for dealing with uncertainties in image databases.

In this paper we present a powerful authorization mechanism which provides support for: (1) periodic authorizations (both positive and negative), that is, authorizations that hold only in specific periods of time; (2) user-defined deductive temporal rules, by which new authorizations can be derived from those explicitly specified; (3) a hierarchical organization of subjects and objects, supporting a more adequate representation of their semantics. From the authorizations explicitly specified, additional authorizations are automatically derived by the system based on those hierarchies. The resulting model is therefore very flexible in terms of the kinds of protection requirements that it can represent. The flexibility provided to the users requires a non trivial underlying formal model where temporal constraints, derivation rules and object and subject hierarchies can be represented. In particular, when inheritance and derivation rules are used simultaneously, there is need for conditions ensuring that the authorization base is free from ambiguities. In this paper, we introduce a notion of safeness , and prove that it guarantees the absence of ambiguities and inconsistencies in the specification. Moreover, we define an efficient algorithm for computing authorizations from safe specifications. Finally, we provide a methodology for supporting temporal authorizations in heterogeneous, distributed systems.

We present a powerful authorization mechanism which supports: (1) positive and negative temporal authorizations; (2) user-defined temporal rules, by which authorizations can be derived; (3) the hierarchical organization of subjects and objects, with an associated mechanism for authorization inheritance.

The problem of integrating multiple heterogeneous legacy databases is an important problem. Many papers [7, 9, 3] to date on this topic have assumed that all the databases comprising a mediated/federated system share the same security ordering. This assumption is often not true as the databases may have been developed independently by different agencies at different points in time. In this paper, we present techniques by which we may merge multiple security orderings into a single unified ordering that preserves the security relationships between orderings. We present a logic programming based approach, as well as a graph theoretical approach to this problem.

The paper presents an extension of the abduction framework proposed by Eshghi and Kowalski to give the semantics of negation as failure in logic programming. The extension allows to cope with the presence of integrity constraints, not only for pruning inconsistent solutions, but for computing new solutions. This is achieved by adding constraints to the program and using them as logical formulae. Furthermore a new semantics based on defaults, equivalent to the previous one, is given for logic programs with negation as failure and constraints. Finally a transformation is presented which allows constraints to be eliminated, so that proof procedures for general clauses without constraints will be applicable.

Educational discussion boards are growing in use as they help students share knowledge and doubts in a working/studying environment. Although these interactive tools are highly effective in learning environments, their usability by blind students is very poor. In this paper we develop techniques to improve accessibility of educational content for students who are blind. Threads of messages in discussion boards evolve with new postings, thus just by investigating the subject headings or contents of earlier postings in a message thread, students may not be able to guess the contents of the postings deeper in the hierarchy. In order to overcome the navigation obstacle for users, it is essential to develop techniques that help identify how the content of a discussion board grows. We develop a technique to organise messages in a message board, by automatically classifying and annotating postings to guide users through discussion segments relevant to their navigational goals.

We introduce a logic programming language which supports hypothetical and counterfactual reasoning. The language is based on a conditional logic which enables to formalize conditional updates of the knowledge base. Due to the presence of integrity constraints, alternative revisions of the knowledge base may result from an update. We develop an abductive semantics which captures different evolutions of the knowledge base. Furthermore, we provide a goal-directed abductive proof procedure to compute the alternative solutions for a goal. We finally analyze our conditional programming language in the context of belief revision theory, and we establish a connection with Nebel's prioritized base revision.

In this paper, we propose a semantics for logic programs with negation as failure, the Finite Failure Stable Model semantics (FF-SM semantics), which is a three-valued extension of Gelfond and Lifschitz' Stable Model semantics. FF-SM semantics is defined in the style of Gelfond and Lifschitz Stable Model semantics, but it builds on an underlying Kripke/Kleene semantics, in which loops causing nonterminating computations are modeled by means of the truth-value undefined . It is different from the eXtended Stable Model (XSM) semantics defined by Przymusinski, since it does not capture infinite failure. We also introduce an abductive proof procedure which is an abductive extension of SLDNF-resolution based on the ideas underlying Eshghi and Kowalski's abductive procedure. We prove that our procedure is sound and complete with respect to FF-SM semantics. We compare the FF-SM semantics with the XSM semantics, and provide a reconstruction for it within the bilattice-based framework proposed by Fitting. In the paper, we deal with the propositional case.

A common challenge in multimedia data understanding is the unsupervised discovery of recurring patterns, or motifs, in time series data. The discovery of motifs in uni-variate time series is a well studied problem and, while being a relatively new area of research, there are also several proposals for multi-variate motif discovery. Unfortunately, motif search among multiple variates is an expensive process, as the potential number of sub-spaces in which a pattern can occur increases exponentially with the number of variates. Consequently, many multi-variate motif search algorithms make simplifying assumptions, such as searching for motifs across all variates individually, assuming that the motifs are of the same length, or that they occur on a fixed subset of variates. In this paper, we are interested in addressing a relatively broad form of multi-variate motif detection, which seeks frequently occurring patterns (of possibly differing lengths) in sub-spaces of a multi-variate time series. In particular, we aim to leverage contextual information to help select contextually salient patterns and identify the most frequent patterns among all. Based on these goals, we first introduce the contextually salient multi-variate motif (CS-motif) discovery problem and then propose a salient multi-variate motif (SMM) algorithm that, unlike existing methods, is able to seek a broad range of patterns in multi-variate time series.