Showing papers by "Payam Barnaghi published in 2007"

Semantic support for medical image search and retrieval

Abstract: The need for annotating digital image data is recognised in a variety of different medical information systems, covering both professional and educational usage of medical imaging Due to the high recall and low precision attribute of keyword-based search, multimedia information search and retrieval based on textual descriptions is not always an efficient and sufficient solution, particularly for specific applications such as the medical diagnosis information systems On the other hand, using image processing techniques to provide search on the content specific data for multimedia information is not a trivial task In this paper we use the semantic web technologies in medical image search and retrieval process for a medical imaging information system We employ an ontology-based knowledge representation and semantic annotation for medical image data The proposed system defines data representation structures which are given well-defined meanings The meanings are machine-accessible contents which could be interpreted by the software agents to find and retrieve the information based on the standard vocabularies and meaningful relationships between the data items

Semantic-Enhanced Information Search and Retrieval

Abstract: Information Retrieval (IR) techniques have been extensively studied since late 1940s and achieved great success evidenced particularly by popular online search engines. However, various classical information retrieval models also have witnessed criticism for emphasizing computation with occurrence of words while ignoring semantics (i.e. meaning of words, search context and etc). Research of the Semantic Web in recent years has provided an opportunity to migrate from mere word-computing to semantic-enhanced information search and retrieval. In this paper, we describe a methodology by combing the Semantic Web, information retrieval, information extraction and social network analysis techniques to elicit semantics from available metadata and ontology in order to develop a semantic-enhanced information search and retrieval system.

Relation Robustness Evaluation for the Semantic Associations

Abstract: The search tools and information retrieval systems on the contemporary Web use keywords, lexical analysis, popularity, and statistical methods to find and prioritise relevant data to a specific query. In recent years, Semantic web has introduced new approaches to specify Web data using machine-interpretable structures. This has led to the establishment of new frameworks for search engines and information systems based on discovering complex and meaningful relationships between information resources. In this paper we discuss a semantic supported information search and retrieval system to answer users’ information queries. The paper focuses on knowledge discovery aspects of the system and in particular analysis of semantic associations. The information resources are multimedia data, which could be retrieved from heterogeneous resources. The main goal is to provide a hypermedia presentation, which narratively conveys relevant information to the queried term. The structure describes the related entities to the queried topic and a ranking mechanism assigns weights to the entities. The assigned weights express the degree of relevancy of each related entity in the presentation structure.

A Context-Aware Ranking Method for the Complex Relationships on the Semantic Web

Abstract: In this paper we propose a method to analyse contextual perspectives in information search and retrieval process in the Semantic Web framework. We start with definitions of context in different computer science applications. We discuss our specific purpose of context-aware information representation and then describe utilising the context confidence factor for ranking the relationships. A confidence factor specifies relevancy degree of an entity to a context in a particular domain. We demonstrate how the uncertainty and Bayesian rule are taken in to account to compute the confidence factor for the results of a query. The main goal is providing a context-aware ranking method which defines degree of relevancy between resources in a set of related information to a particular query.

Semantic Media in Medical Information Systems

Abstract: The content-based image retrieval (CBIR) methods have been applied to medical field to aid diagnosis and other medical processes for a relatively long time. This type of systems, with the growth of digital medical image databases, sometimes does not provide information tailored to the end user requirements. Typically there is a semantic gap between the low-level features extracted form the images and high-level concepts required by the user. This paper proposes an ontology-based search and retrieval as a supplementary method to associate high- level concepts and semantics to medical image data and exploit these semantic media for information search and retrieval. The main goal is to perform a search on a medical image repository and to retrieve relevant medical images and information to a particular case in order to assist diagnosis process. We demonstrate the functionality of the system in case of a mammography imaging database. The system uses an ontology-based search and retrieval for the medical data as a complementary solution to provide more efficient and insight access to the stored data.

A survey of automated multimedia presentation generation frameworks

Abstract: The widespread usage of multimedia data in recent years has led to an enormous interest in making these data available through the Web. Due to the complex requirements of multimedia data, the design and implementation of web-based information systems to fulfil the search, retrieval, and presentation necessities for such systems is not a trivial task. This paper surveys systems and methods for creating meaningful multimedia presentations to answer user information queries. Initial work in this area is carried out by introducing a standard reference model for the intelligent multimedia presentation systems (SRM-IMMPS). The standard reference model introduces a high-level architecture and a plan-based approach to generate automated multimedia presentation. The paper focuses on annotated multimedia data processing and describes different frameworks for automated multimedia presentation generation.