We apply hierarchical Latent Dirichlet Allocation (hLDA) to the problem of ontology annotation; automatically extending WORDNET with new concepts and annotating existing concepts with generic property fields, or attributes. The resulting annotations are evaluated along two dimensions: (1) the precision of the ranked lists of attributes at each concept, and (2) the specificity of the attribute a...

Cresceptron represents a new approach t o neural networks. It uses a hierarchical f m m e w o r k t o grow neural networks automatically, adaptively and incrementally through learning. A t every level of the hierarchy, new concepts are detected automatically and the network grows by creating new neurons and synapses which memorize the new concepts and their context. The training samples are gen...

This paper introduces a systematic approach to organizing the discipline of clinical chemistry. The approach is called a top-down systems approach because it starts at the top with the most general concepts and works down through less general concepts to the most specific details and techniques. The hypothesis is that the discipline can be organized into hierarchical levels of functional proces...

This paper proposes a framework to automatically construct taxonomies from a corpus of text documents. This framework first extracts terms from documents using a part-of-speech parser. These terms are then filtered using domain pertinence, domain consensus, lexical cohesion, and structural relevance. The remaining terms represent concepts in the taxonomy. These concepts are arranged in a hierar...

Tagging in social media system has demonstrated to be a convenient way for users to annotate objects of interest.One reason behind its success obviously because tags can be chosen by users arbitrarily without any topic and specificity constraints. Although tags are free-from keywords, there are some evidences 1 suggesting that, for a particular object type, users tend to use “similar” tag sets....

We examined the effectiveness of post-retrieval clustering that was based on the visual similarities among images to enhance the instance recall in the photo retrieval task of ImageCLEF 2008. The visual similarities are defined by the example visual concepts that were provided for the automatic photo indexing task. We tested two types of visual concepts and two kinds of clustering methods, hier...

In this paper, we take the logical approach to information retrieval in order to identify and describe new concepts required for performing hypermedia retrieval . For this purpose, we consider hypertext linking of nodes, hierarchical structure of documents and document type hierarchies. These concepts are described in Datalog, a horn logic without functions. Furthermore, we discuss terminologic...

The objective of this study is to evaluate the contribution to semantic integration of the semantic relations extracted from concept names, representing augmented knowledge. Three augmentation methods – based on linguistic phenomena – are investigated (reification, nominal modification, and prepositional attachment). The number of concepts aligned in two ontologies of anatomy before and after a...

The theory of concept lattices (i.e. hierarchical structures of concepts in the sense of Port-Royal school) is approached from the point of view of fuzzy logic. The notions of partial order, lattice order, and formal concept are generalized for fuzzy setting. Presented is a theorem characterizing the hierarchical structure of formal fuzzy concepts arising in a given formal fuzzy context. Also, ...

ION VS. REFINEMENT When modelling complex systems, one usually starts with an abstract model, which is refined in the course of gaining a better understanding of the problem domain. It is very difficult to come up with generic concepts that allow identifying relations between abstract and refined models since these concepts are heavily bound to the underlying engineering process. Thus, the conc...