Designing Fuzzy Relations in Orthogonal Persistence Object-Oriented Database Engines

García‐Barriocanal

González-Sotos

2006

Int. J. Intell. Syst.

Imperfection in information can be considered a crosscutting concern that manifests itself in diverse kinds of imprecision, uncertainty, or inconsistency in the data models of a software system. The extension of existing programming and querying interfaces for the different aspects of information imperfection requires a proper modularization of the different concerns of numerical imprecision handling, so that the extensions do not interfere with existing programming practices and do not obscure the original design. Aspect-oriented design~AOD! enables such a form of nonintrusive extensions to be added to existing software libraries, clearly separating fuzziness or other imperfections in data as a differentiated concern that can be considered from the early phases of development. In this article, a general framework for aspect-based extension of data models and fuzzy databases is described, and some design and implementation issues of such AOD-based extensions on OJB database libraries are described as a case study.

Section: Describing Imperfect-data Elementsmentioning

confidence: 99%

Introducing fuzziness in object models and database interfaces through aspects

García‐Barriocanal

González-Sotos

2006

Int. J. Intell. Syst.

“…-Closeness between users is modelled as a fuzzy relation, represented as a <<fuzzy>> association (see [19]). This relationship is constructed from the history of recommendations, that serves as an estimator due to the fact that deeper knowledge of each other is assumed to be correlated to the quantity of interactions.…”

Section: Conceptual Modelmentioning

confidence: 99%

Fuzzy Group Models for Adaptation in Cooperative Information Retrieval Contexts

SOFSEM 2004: Theory and Practice of Computer Science

García

2004

Self Cite

Abstract. Cooperation in information retrieval contexts can be used to share query results inside groups of individuals with common objectives, provided that all of them are aware of each other. The strength of the social relationships between group members is in most cases a matter of comparative degree, and thus relationships can be modelled through fuzzy conceptual associations. These associations can then be used to implement personalized features, aimed at improving the interaction of the user with the query tool. In this paper, an approach to modelling imprecise relationships between users in the context of information retrieval is described, along with a concrete case study implemented as a wrapper of a conventional search engine, using a fuzzy database to store the model of the group members.

“…These preferences may be stated in a general way as assertions in the form (u, c) : likesGranularity = 0.8 where u is a user of the system u : U ser = 1 , and c may be a content, node (or any other form of UI element description), which can be expressed as the assertion u : Content N ode U IElement = 1 in a generic way. Once again, storing assertions for the whole cartesian product of users and elements may be unpractical in many situations, specially for large relations stored in databases (see [20]). We have devised an straightforward approach to granulate preferences, consisting on the representation of user preferences as related to prototypical instances representing vague categories.…”

Section: Preference Modelingmentioning

confidence: 99%

“…This form of computing relevances calls for a very specific storage format for preferences, to avoid retrieving all the database of objects. One possible approach may be that of storing the likesX predicates as fuzzy relations in compact α-cut format [20] thus obtaining a O(m+m·log m) complexity, being m the number of relations likeX from u i having significant degrees, and provided that the relations are sorted in m · log m by object to enable the computation of the final degrees in a single pass.…”

Section: Information Filteringmentioning

confidence: 99%

The Role of Vague Categories in Semantic and Adaptive Web Interfaces

On the Move to Meaningful Internet Systems 2003: OTM 2003 Workshops

2003

Abstract. Current Semantic Web technologies provide a logic-based framework for the development of advanced, adaptive applications based on ontologies. But the experience in using them has shown that, in some cases, it would be convenient to extend its logic support to handle vagueness and imprecision in some way. In this paper, the role of vagueness in the description of Web user interface characteristics is addressed, from the viewpoint of the design of adaptive behaviors that are connected to such descriptions. Concretely, vague descriptions combined with quantified fuzzy rules and flexible connectors are described, and their usefulness is illustrated through preference modeling, filtering and adaptive linking scenarios.