Search results

Feature space heterogeneity exists widely in various application fields of classification techniques, such as customs inspection decision, credit scoring and medical diagnosis. This paper aims to study the relationship between feature space heterogeneity and classification performance.

A measurement is first developed for measuring and identifying any significant heterogeneity that exists in the feature space of a data set. The main idea of this measurement is derived from a meta-analysis. For the data set with significant feature space heterogeneity, a classification algorithm based on factor analysis and clustering is proposed to learn the data patterns, which, in turn, are used for data classification.

The proposed approach has two main advantages over the previous methods. The first advantage lies in feature transform using orthogonal factor analysis, which results in new features without redundancy and irrelevance. The second advantage rests on samples partitioning to capture the feature space heterogeneity reflected by differences of factor scores. The validity and effectiveness of the proposed approach is verified on a number of benchmarking data sets.

Measurement should be used to guide the heterogeneity elimination process, which is an interesting topic in future research. In addition, to develop a classification algorithm that enables scalable and incremental learning for large data sets with significant feature space heterogeneity is also an important issue.

Measuring and eliminating the feature space heterogeneity possibly existing in the data are important for accurate classification. This study provides a systematical approach to feature space heterogeneity measurement and elimination for better classification performance, which is favorable for applications of classification techniques in real-word problems.

A measurement based on meta-analysis for measuring and identifying any significant feature space heterogeneity in a classification problem is developed, and an ensemble classification framework is proposed to deal with the feature space heterogeneity and improve the classification accuracy.

The purpose of this paper is to draw an exact and accurate picture of classification problems being faced by libraries having a reasonable amount of collections on Islam. Different classification systems have been developed during the last two centuries to organize library materials. Where these systems provided libraries with better solution for organization of materials, the systems have also some limitations. Religious materials are one of the areas where these systems could not satisfy a reasonable number of libraries. The libraries that have rich collections on Islam are also facing such problems. This study has been conducted with the aim of addressing this problem.

The study comprises a precise review of literature relevant to this problem, along with the collection and analysis of data from such libraries from all the main cities of Pakistan. The data have been collected using the interview technique.

The review of literature and empirical data collected for this study show that the libraries are facing problems regarding the classification of materials where there are rich collections on Islam. The standard classification systems have not provided proper place and enumeration to Islamic topics in their schemes. There is no standard or uniform practice among the libraries to classify materials on Islam. Different libraries are using different systems, which has resulted placement of same materials at different locations. There is no coordination among libraries having rich collections on Islam for some common plan of action to resolve the problem in hand. No work on automated classification has been done in Pakistan so far and this is the area that needs working on in the future.

The paper focusses on materials in libraries in one country only, however highlights issues relevant to other Islamic countries in organizing such materials.

The paper discusses an area of professional concern that has been discussed widely in Islamic countries, but only in a limited fashion outside of Islamic countries. Thus the paper should be of interest to researchers and practioners interested in cataloging theory.

This study aims to identify problems connected to information classification in theory and to put those problems into the context of experiences from practice.

Five themes describing problems are discussed in an empirical study, having informants represented from both a public and a private sector organization.

The reasons for problems to occur in information classification are exemplified by the informants’ experiences. The study concludes with directions for future research.

Information classification sustains the basics of security measures. The human–organizational challenges are evident in the activities but have received little attention in research.

The purpose of this paper is to develop understanding of the problems of classification, to discover the classification practices of libraries with rich collections on Islam cited in the literature, to find the gaps, and to determine the point from which to start work on further development.

Published and unpublished literature, both print and electronic, that is relevant to the problem was reviewed objectively in the compilation of this paper.

Standard classification systems lack proper space for materials on Islam for two reasons: less awareness on the part of devisers of the depth and variety of Islamic topics; and their bias and lack of interest in Islam. Different indigenous classification systems and expansions have been developed, using either the original notation or alternative notations. Some systems have been developed without following any standards or logic. This study has revealed a need for empirical study of libraries with rich collections on Islam in order to gain a better understanding of the problem and find an optimal solution.

No empirical field data are included in this study. This is a review of the literature.

The author indicates the current situation of the problem and a potential framework for its solution.

Classification systems play a fundamental role in the organization, display, retrieval and access to the knowledge materials in libraries. These systems have served the purpose adequately in most of knowledge areas; nevertheless, some grey areas lack proper place and enumeration in these systems. Islamic knowledge is among the areas that have not been properly addressed. The purpose of this paper is to examine this problem and indicate a potential solution.

This paper expands on the author's earlier research which focused on Pakistan library collections. Empirical data have been collected from 16 LIS scholars who have interest in or expertise on this issue through interviews. Scholars are from Pakistan, India, Malaysia, Iran, Saudi Arabia, Egypt, the UK, the USA and Canada. A review of the literature is also presented.

A number of approaches have been taken to work around the deficiencies of the standard classification systems when it comes to Islamic knowledge and publications, including indigenous systems and expansions. Details of some of these are presented. A range of possible improvements to existing classification systems was suggested by scholars, and an outline of what is required in a new, independent system is discussed, along with ideas about the best way for this system to be developed.

The paper discusses an area of professional concern that has been discussed widely in Islamic countries, but only in a limited fashion outside of Islamic countries. Thus, the paper should be of interest to researchers and practitioners interested in cataloguing and classification theory.

The purpose of the paper is to develop a method of automatic classification of the components of the assembly units. The method is crucial for developing an automatic ship assembly planning tools. The proposed method takes into account the assumptions specific for shipbuilding technology processes: high complexity of structures, difficult expert-based classification of components, fixed priority relations between connections resulting from geometrical constraints and demands of welding processes.

The set of ex post determined liaisons and assembly sequences constitutes the database of structures which have been made-up earlier. The components classification problem is solved using matrix coding of graphs. Information in such form is stored in the database. The minimization of number of cycles in the graph of classes sequence and minimization of diversity of classes within all constructions has been proposed as criteria of optimization. The genetic algorithm has been proposed as a solution method.

The proposed method solves the problem of components’ classifications. It allows setting the pattern of priorities between classes of various connections. This gives a chance to determine the relationship constraints between the connections of new structures for which assembly sequences are not established.

Mathematical formulation of the database is quite laborious. The possibility of partial automation of this process should be considered. Owing to the complexity of the problem, a relatively simple objective function has been proposed. During a ship hull assembly, additional criteria should be taken into account, what will be the direction of further research.

Automatic classification of components is dedicated for implementation in shipyards and similar assembly systems. Tests performed by the authors confirm efficiency of presented method in supporting management of the database and assembly of new structures planning. Suggested activity-oriented approach allows for easy conversion of any assembly unit structure to the form of a matrix.

The new approach for components classification according to its assembly features distinguishes the proposed method from others. The use of nilpotent matrix theory in an acyclicity of graphs analysis is also a unique achievement. Original crossover and mutation operators for assembly sequence were proposed in the article.

The group has continued to meet regularly since the publication of the last bulletin and has welcomed a number of new members and visitors from both home and overseas. Many members who joined at the beginning or very early on in the Group's history still attend regularly, but several long‐standing members have also left, or ceased active participation, in the period under review. Towards the end of 1972 Mr Wells relinquished the chairmanship of the Group, due to pressure of work, and his place was taken by Mr Mills. Another departure, and one that robbed the Group of one of its most active and forceful members, was that of Jason Farradane. He left the country in 1974, and the Group presented him with a book as a memento of many enjoyable and provocative discussions stimulated by his presence at the meetings which he unfailingly attended. It was with great pleasure that he was welcomed back to a meeting while he was visiting this country in January 1976.

The proper identification of a decision problem is paramount to finding a course of action or solution. This paper attempts to capture the general laws or principles that define decision problems. These principles are then used to establish a decision classification system called the decision‐order taxonomy.

The decision‐order taxonomy is developed by performing a content analysis on the seminal literature in the natural, social, and applied sciences. By identifying the semantic descriptors used to partition various domains, an implicit taxonomy for proper identification of decision problems is hypothesized.

The multidimensional taxonomic classification system and defined nomenclature, together with the identification process, comprise the complete decision‐order taxonomy developed in this paper. While applying the decision‐order taxonomy to an actual decision problem, insights are exposed which will guide the decision maker toward appropriate solution methodologies.

The theoretical foundation developed can be used to promote future research in decision classification. By providing a theoretically derived model, rich opportunities to test the taxonomy empirically are offered. Researchers are also given a foundation upon which they can build interdisciplinary decision models.

For practitioners, the decision‐order taxonomy provides a new paradigm for communicating decision problems across disciplinary boundaries. The taxonomy also provides guidance to the practitioner as they search for appropriate solution methodologies in unfamiliar disciplines.

The establishment of a useful decision‐problem taxonomy is a significant contribution to understanding the multidimensional interdisciplinary nature of real world decision problems. The original classifications will promote cross disciplinary communication, a central element in business success.

I gather that this paper was given its somewhat pretentious title in an endeavour to emphasize the practical approach which it was hoped I should bring to this subject this evening. How far I shall be able to live up to this hope I hesitate to say. I realize that the arrangement of information in your libraries presents a very practical problem and that it is very much of a practical issue with you what form of arrangement you shall choose for your catalogues and materials. Some of you, I believe, have not solved the problem at all and others of you are dis‐satisfied with the solution which you have found. Only a very few enviable geniuses have found the complete solution to their own particular problems.

Rapid developments in social networks and their usage in everyday life have caused an explosion in the amount of short electronic documents. Thus, the need to classify this type of document based on their content has a significant implication in many applications. The need to classify these documents in relevant classes according to their text contents should be interested in many practical reasons. Short-text classification is an essential step in many applications, such as spam filtering, sentiment analysis, Twitter personalization, customer review and many other applications related to social networks. Reviews on short text and its application are limited. Thus, this paper aims to discuss the characteristics of short text, its challenges and difficulties in classification. The paper attempt to introduce all stages in principle classification, the technique used in each stage and the possible development trend in each stage.

The paper as a review of the main aspect of short-text classification. The paper is structured based on the classification task stage.

This paper discusses related issues and approaches to these problems. Further research could be conducted to address the challenges in short texts and avoid poor accuracy in classification. Problems in low performance can be solved by using optimized solutions, such as genetic algorithms that are powerful in enhancing the quality of selected features. Soft computing solution has a fuzzy logic that makes short-text problems a promising area of research.

Using a powerful short-text classification method significantly affects many applications in terms of efficiency enhancement. Current solutions still have low performance, implying the need for improvement. This paper discusses related issues and approaches to these problems.