No description available.

This work presents a geometry-based image retrieval approach for multi-object images. We commence with developing an effective shape matching method for closed boundaries. Then, a structured representation, called curvature tree (CT), is introduced to extend the shape matching approach to handle images containing multiple objects with possible holes. We also propose an algorithm, based on Gestalt principles, to detect and extract high-level boundaries (or envelopes), which may evolve as a result of the spatial arrangement of a group of image objects. At first, a shape retrieval method using triangle-area representation (TAR) is presented for non-rigid shapes with closed boundaries. This representation is effective in capturing both local and global characteristics of a shape, invariant to translation, rotation, scaling and shear, and robust against noise and moderate amounts of occlusion. For matching, two algorithms are introduced. The first algorithm matches concavity maxima points extracted from TAR image obtained by thresholding the TAR. In the second matching algorithm, dynamic space warping (DSW) is employed to search efficiently for the optimal (least cost) correspondence between the points of two shapes. Experimental results using the MPEG-7 CE-1 database of 1400 shapes show the superiority of our method over other recent methods. Then, a geometry-based image retrieval system is developed for multi-object images. We model both shape and topology of image objects including holes using a structured representation called curvature tree (CT). To facilitate shape-based matching, the TAR of each object and hole is stored at the corresponding node in the CT. The similarity between two CTs is measured based on the maximum similarity subtree isomorphism (MSSI) where a one-to-one correspondence is established between the nodes of the two trees. Our matching scheme agrees with many recent findings in psychology about the human perception of multi-object images. Two algorithms are introduced to solve the MSSI problem: an approximate and an exact. Both algorithms have polynomial-time computational complexity and use the DSW as the similarity measure between the attributed nodes. Experiments on a database of 13500 medical images and a database of 1580 logo images have shown the effectiveness of the proposed method. The purpose of the last part is to allow for high-level shape retrieval in multi-object images by detecting and extracting the envelope of high-level object groupings in the image. Motivated by studies in Gestalt theory, a new algorithm for the envelope extraction is proposed that works in two stages. The first stage detects the envelope (if exists) and groups its objects using hierarchical clustering. In the second stage, each grouping is merged using morphological operations and then further refined using concavity tree reconstruction to eliminate odd concavities in the extracted envelope. Experiment on a set of 110 logo images demonstrates the feasibility of our approach.

The nature of the generic chemical structures found in patents is described, with a discussion of the types of statement commonly found in them. The available representations for such structures are reviewed, with particular note being given to the suitability of the representation for searching files of such structures. Requirements for the unambiguous representation of generic structures in an "ideal" storage and retrieval system are discussed. The basic principles of the theory of formal languages are reviewed, with particular consideration being given to parsing methods for context-free languages. The Grammar and parsing of computer programming languages, as an example of artificial formal languages, is discussed. Applications of formal language theory to chemistry and information work are briefly reviewed. GENSAL, a formal language for the unambiguous description of generic structures from patents, is presented. It is designed to be intelligible to a chemist or patent agent, yet sufficiently ABSTRACT formaLised to be amenabLe to computer anaLysis. DetaiLed description is given of the facilities it provides for generic structure representation, and there is discussion of its Limitations and the principLes behind its design. A connection-tabLe-based internaL representation for generic structures, caLLed an ECTR (Extended Connection Table Representation) is presented. It is designed to represent generic structures unambiguousLy, and to be generated automatically from structures encoded in GENSAL. It is compared to other proposed representations, and its implementation using data types of the programming Language PascaL described. An interpreter program which generates an ECTR from structures encoded in a subset of the GENSAL Language is presented. The principles of its operation are described. Possible applications of GENSAL outside the area of patent documentation are discussed, and suggestions made for further work on the development of a generic structure storage and retrieval system based on GENSAL and ECTRs.

No description available.

Digitized by Kansas Correctional Industries

Digitized by Kansas Correctional Industries

by Tang Ting Kap. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2001. / Includes bibliographical references (leaves 88-94). / Abstracts in English and Chinese. / List of Figures --- p.vi / List of Tables --- p.vii / Chapter 1. --- INTRODUCTION --- p.1 / Chapter 1.1 --- Information Overloading and Low Utilization --- p.2 / Chapter 1.2 --- Problem Needs To Solve --- p.3 / Chapter 1.3 --- Research Contributions --- p.4 / Chapter 1.3.1 --- Using Concept Space in Summarization --- p.5 / Chapter 1.3.2 --- New Extraction Method --- p.5 / Chapter 1.3.3 --- Experiments on New System --- p.6 / Chapter 1.4 --- Organization of This Thesis --- p.7 / Chapter 2. --- LITERATURE REVIEW --- p.8 / Chapter 2.1 --- Classical Approach --- p.8 / Chapter 2.1.1 --- Luhn's Algorithm --- p.9 / Chapter 2.1.2 --- Edumundson's Algorithm --- p.11 / Chapter 2.2 --- Statistical Approach --- p.15 / Chapter 2.3 --- Natural Language Processing Approach --- p.15 / Chapter 3. --- PROPOSED SUMMARIZATION APPROACH --- p.18 / Chapter 3.1 --- Direction of Summarization --- p.19 / Chapter 3.2 --- Overview of Summarization Algorithm --- p.20 / Chapter 3.2.1 --- Document Pre-processing --- p.21 / Chapter 3.2.2 --- Vector Space Model --- p.23 / Chapter 3.2.3 --- Sentence Extraction --- p.24 / Chapter 3.3 --- Evaluation Method --- p.25 / Chapter 3.3.1 --- "Recall, Precision and F-measure" --- p.25 / Chapter 3.4 --- Advantage of Concept Space Approach --- p.26 / Chapter 4. --- SYSTEM ARCHITECTURE --- p.27 / Chapter 4.1 --- Converge Process --- p.28 / Chapter 4.2 --- Diverge Process --- p.30 / Chapter 4.3 --- Backward Search --- p.31 / Chapter 5. --- CONVERGE PROCESS --- p.32 / Chapter 5.1 --- Document Merging --- p.32 / Chapter 5.2 --- Word Phrase Extraction --- p.34 / Chapter 5.3 --- Automatic Indexing --- p.34 / Chapter 5.4 --- Cluster Analysis --- p.35 / Chapter 5.5 --- Hopfield Net Classification --- p.37 / Chapter 6. --- DIVERGE PROCESS --- p.42 / Chapter 6.1 --- Concept Terms Refinement --- p.42 / Chapter 6.2 --- Sentence Selection --- p.43 / Chapter 6.3 --- Backward Searching --- p.46 / Chapter 7. --- EXPERIMENT AND RESEARCH FINDINGS --- p.48 / Chapter 7.1 --- System-generated Summary v.s. Source Documents --- p.52 / Chapter 7.1.1 --- Compression Ratio --- p.52 / Chapter 7.1.2 --- Information Loss --- p.54 / Chapter 7.2 --- System-generated Summary v.s. Human-generated Summary --- p.58 / Chapter 7.2.1 --- Background of EXTRACTOR --- p.59 / Chapter 7.2.2 --- Evaluation Method --- p.61 / Chapter 7.3 --- Evaluation of different System-generated Summaries by Human Experts --- p.63 / Chapter 8. --- CONCLUSIONS AND FUTURE RESEARCH --- p.68 / Chapter 8.1 --- Conclusions --- p.68 / Chapter 8.2 --- Future Work --- p.69 / Chapter A. --- EXTRACTOR SYSTEM FLOW AND TEN-STEP PROCEDURE --- p.71 / Chapter B. --- SUMMARY GENERATED BY MS WORD2000 --- p.75 / Chapter C. --- SUMMARY GENERATED BY EXTRACTOR SOFTWARE --- p.76 / Chapter D. --- SUMMARY GENERATED BY OUR SYSTEM --- p.77 / Chapter E. --- SYSTEM-GENERATED WORD PHRASES FROM TEST SAMPLE --- p.78 / Chapter F. --- WORD PHRASES IDENTIFIED BY SUBJECTS --- p.79 / Chapter G. --- SAMPLE OF QUESTIONNAIRE --- p.84 / Chapter H. --- RESULT OF QUESTIONNAIRE --- p.85 / Chapter I. --- EVALUATION FOR DIVERGE PROCESS --- p.86 / BIBLIOGRAPHY --- p.88

Thesis (M.S.)--Oregon State University, 1984. / Typescript (photocopy). Includes bibliographical references (leaf 57). Also available on the World Wide Web.

Chinese is very different from alphabetical languages such as English, as there are no delimiters between Chinese words. So Chinese segmentation is an important step for most Chinese natural language processing (NLP) tasks. We propose a tightness continuum for Chinese semantic units. The construction of the continuum is based on statistical informations. Based on this continuum, sequences can be dynamically segmented, and then that information can be exploited in a number of information retrieval tasks. In order to show that our tightness continuum is useful for NLP tasks, we propose two methods to exploit the tightness continuum within IR systems. The first method refines the result of a general Chinese word segmenter. The second method embeds the tightness value into IR score functions. Experimental results show that our tightness measure is reasonable and does improve the performance of IR systems.

Electrodermal activity (EDA), an indicator of arousal of the sympathetic nervous system, was investigated as a potential correlate of feeling-of-knowing (FOK) and tip-of-the-tongue (TOT) states. In Experiment 1, skin conductance was measured while participants answered general knowledge questions and made binary FOK and TOT judgments. Significant correlations were found between frequency of skin conductance responses (SCRs) and presence of both FOK and TOT states. In Experiment 2, warmth ratings were used and a follow-up clue session was added to offer participants the opportunity to resolve initially unanswered questions. SCR frequency during TOT states was significantly predictive of resolution during the clue period, although not as predictive as participants' warmth ratings. The potential of EDA as an on-line, non-intrusive measure of metamemory and memory retrieval is discussed.