Wojciech Świeboda

Interactive Document Indexing Method Based on Explicit Semantic Analysis

In this article we propose a general framework incorporating semantic indexing and search of texts within scientific document repositories. In our approach, a semantic interpreter, which can be seen as a tool for automatic tagging of textual data, is interactively updated based on feedback from the users, in order to improve quality of the tags that it produces. In our experiments, we index our document corpus using the Explicit Semantic Analysis (ESA) method. In this algorithm, an external knowledge base is used to measure relatedness between words and concepts, and those assessments are utilized to assign meaningful concepts to given texts. In the paper, we explain how the weights expressing relations between particular words and concepts can be improved by interaction with users or by employment of expert knowledge. We also present some results of experiments on a document corpus acquired from the PubMed Central repository to show feasibility of our approach.

Semantic analytics of pubmed content

We present an architecture aimed at semantic search and synthesis of information acquired from the document repositories. The proposed framework is expected to provide domain knowledge interfaces enabling the internally implemented algorithms to identify relationships between documents, researchers, institutions, as well as concepts extracted from various types of knowledge bases. The framework should be scalable with respect to data volumes, diversity of analytic processes, and the speed of search. In this paper, we investigate these requirements for the case of medical publications gathered in PubMed.

Extended Document Representation for Search Result Clustering

Organizing query results into clusters facilitates quick navigation through search results and helps users to specify their search intentions. Most meta-search engines group documents based on short fragments of source text called snippets. Such a model of data representation in many cases shows to be insufficient to reflect semantic correlation between documents. In this paper, we discuss a framework of document description extension which utilizes domain knowledge and semantic similarity. Our idea is based on application of Tolerance Rough Set Model, semantic information extracted from source text and domain ontology to approximate concepts associated with documents and to enrich the vector representation.

Enhancing search result clustering with semantic indexing

Semantic search results clustering is one of the most wanted functionalities of many information retrieval systems including general web search engines as well as domain specific article portals or digital libraries. It may advice the users to describe the need for information in a more precise way. In this paper, we discuss a framework of document description extension which utilizes domain knowledge and semantic similarity. Our idea is based on application of Tolerance Rough Set Model, semantic information extracted from source text and domain ontology to approximate concepts associated with documents and to enrich the vector representation. Some document representation models including document meta-data, citations and semantic information build using MeSH ontology. We compare those models in a search result clustering problem over the freely accessed biomedical research articles from Pubmed Cetral (PMC) portal. The experimental results are showing the advantages of the proposed models.

Weight Learning for Document Tolerance Rough Set Model

Creating a document model for efficient keyword search is a long studied problem in Information Retrieval. In this paper we explore the application of Tolerance Rough Set Model for Documents TRSM for this problem. We further provide an extension of TRSM with a weight learning procedure TRSM-WL and compare performance of these two algorithms in keyword search. We further provide a generalization of TRSM-WL that imposes additional constraints on the underlying model structure and compare it to a supervised variant of Explicit Semantic Analysis.

Semantic Evaluation of Search Result Clustering Methods

This paper is a continuation of the research on designing and developing a dialog-based semantic search engine for SONCA system which is a part of the SYNAT project. In previous papers we proposed some extensions of Tolerance Rough Set Model (TRSM), which can be used to improve the search result clustering algorithms. In this paper, we investigate the problem of quality analysis of presented solutions. We propose some semantic evaluation measures to estimate the quality of the proposed search clustering methods. We illustrate the proposed evaluation method on the base of the Medical Subject Headings (MeSH) thesaurus and compare different clustering techniques over the commonly accessed biomedical research articles from PubMed Central (PMC) portal. The experimental results are showing the advantages of our new clustering methods which are implemented in SONCA system.

Reduct Calculation and Discretization of Numeric Attributes in Entity Attribute Value Model

In this paper we review the problem of short reduct calculation in a sparse decision system. We also address the problem of discretization of numerical attributes in sparse decision systems. We present algorithms that provide an approximate solution to these two problems and analyze the complexity of these algorithms.

Granular Floor Plan Representation for Evacuation Modeling

In this paper we describe the architecture of a simple evacuation model which is based on a graph representation of the scene. Such graphs are typically constructed using Medial Axis Transform (MAT) or Straight Medial Axis Transform (S-MAT) transformations, the former being a part of the Voronoi diagram (Dirichlet tessellation) of the floor plan. In our work we construct such graphs for floor plans using Voronoi diagram along with the dual Delaunay triangulation of a set of points approximating the scene. Information supplied by Delaunay triangulation complements the graph in two ways: it determines capacities of some paths associated with edges, and provides a bijection between graph vertices and a set of regions forming a partition of the floor plan. We call the representation granular for this reason.

Interactive Method for Semantic Document Indexing Based on Explicit Semantic Analysis

In this article we propose a general framework incorporating semantic indexing and search of texts within scientific document repositories. In our approach, a semantic interpreter, which can be seen as a tool for automatic tagging of textual data, is interactively updated based on feedback from the users, in order to improve quality of the tags that it produces. In our experiments, we index our document corpus using the Explicit Semantic Analysis (ESA) method. In this algorithm, an external knowledge base is used to measure relatedness between words and concepts, and those assessments are utilized to assign meaningful concepts to given texts. In the paper, we explain how the weights expressing relations between particular words and concepts can be improved by interaction with users or by employment of expert knowledge. We also present some results of experiments on a document corpus acquired from the PubMed Central repository to show feasibility of our approach.

Semantic Evaluation of Text Clustering

In this paper, we investigate the problem of quality analysis of clustering results using semantic annotations given by experts. We propose a novel approach to construction of evaluation measure, called SEE (Semantic Evaluation by Exploration), which is an improvement of the existing methods such as Rand Index or Normalized Mutual Information. We illustrate the proposed evaluation method on the freely accessible biomedical research articles from Pubmed Central (PMC). Many articles from Pubmed Central are annotated by the experts using Medical Subject Headings (MeSH) thesaurus. We compare different semantic techniques for search result clustering using the proposed measure.