Elsevier

Information Sciences

Volume 178, Issue 1, 2 January 2008, Pages 1-20
Information Sciences

A multiview approach for intelligent data analysis based on data operators

https://doi.org/10.1016/j.ins.2007.08.011Get rights and content

Abstract

Multiview intelligent data analysis explores data from different perspectives to reveal various types of structures and knowledge embedded in the data. Each view may capture a specific aspect of the data and hence satisfy the needs of a particular group of users. Collectively, multiple views provide a comprehensive description and understanding of the data. In this paper, we propose a multiview framework of intelligent data analysis based on modal-style data operators. The classes of the data operators include basic set assignment, sufficiency, dual sufficiency, necessity and possibility operators. They demonstrate various types of data relationships and characterize various features and granulated views of the data. It is shown that different structures of the data can also be constructed based on the different data operators.

Introduction

Huge data sets and various data types lead to new types of problems and require the development of new types of techniques for modern intelligent data analysis [21]. An important objective of intelligent data analysis is to reveal and indicate diverse non-trivial features or views of a large amount of data. Many techniques and models in data mining, machine learning, pattern recognition, statistics, and other fields have been proposed. Each technique or model focuses on one particular view of the data and discovers a specific type of knowledge embedded in data.

In this paper, we propose a multiview approach that provides a unified framework for integrating multiple views of intelligent data analysis. It is easy to know that an integrated and unified framework that allows a multiple view approach on the understanding, computation, and interpretation of data provides not only a platform to explore different aspects of the data, but also a tool to integrate many techniques of intelligent data analysis.

Modal-style data operators can be used to define, represent and analyze various types of data relations and structures [6], [52]. Therefore, these modal-style data operators can provide a unified way to examine, characterize and construct different types of knowledge. In this paper, a multiview approach is introduced based on modal-style data operators in a formal context.

The rest of the paper is organized as follows. In the next section, discussions about motivations and some related works are provided. Section 3 introduces formal contexts and modal-style data operators. In Section 4, we show that these data operators can be used to define different relations and granular structures of a universe. Section 5 examines several types of hierarchical structures defined based on the modal-style data operators. These granular and hierarchical structures are always be viewed as different types of knowledge embedded in the data set [6], [11], [38], [52]. The conclusion of this study is given in Section 6.

Section snippets

Motivations and related works

In this section, we give motivations of the present study and discuss some related works.

Formal contexts and modal-style data operators

In this section, definitions for formal contexts and modal-style data operators are given. Some connections between modal-style data operators are investigated.

Granular structures of the universe

In this section, we investigate various relations between objects and granular structures induced by the relations based on modal-style data operators.

Hierarchical structures embedded in data

In the classical view, a concept is jointly determined by its extension and intention. The extension consists of all objects belonging to the concept. The intension includes all attributes that are used to characterize the objects belonging to the concept. A pair of extension and intention is considered as a representation of a concept. Lattice theory provides a vocabulary for hierarchical structures and can be applied to construct concept lattices [17], [49], [53], [54].

In this section, the

Conclusion

In this paper, we propose an approach to investigate multiple views of intelligent data analysis based on modal-style data operators. We provides a basic framework to explore the significant importance of multiview for the research on intelligent data analysis. The integration of these multiple views may bring more powerful and useful data analysis tools.

By defining various types of data operators, we present various data relationships and construct different types of hierarchical structures,

Acknowledgements

The authors thank anonymous referees of this paper for their constructive comments and partial support for NSERC Canada.

References (65)

  • A. Bargiela et al.

    Granular Computing: An Introduction

    (2002)
  • B. Belkhouche, C. Lemus-Olalde, Multiple veiw analysis of designs, Foundations of Software Engineering, Joint...
  • B. Belkhouche et al.

    Multiple views analysis of software designs

    International Journal of Software Engineering and Knowledge Engineering

    (2000)
  • L. Breiman et al.

    Classification and Reggression Trees

    (1984)
  • W. Buszkowski, Approximation spaces and definability for incomplete information systems, Rough Sets and Current Trends...
  • Y.H. Chen, Y.Y. Yao, Multiview Intelligent Data Analysis based on Granular Computing, in: Proceedings of 2006 IEEE...
  • P.M. Cohn

    Universal Algebra

    (1965)
  • T.M. Cover et al.

    Nearest neighbor pattern classification

    IEEE Transactions on Information Theory

    (1967)
  • I. Düntsch et al.

    Approximation operators in qualitative data analysis

  • M. Ester, H.P. Kriegel, J. Sander, X. Xu, A density-based algorithm for discovering clusters in large spatial data sets...
  • A.A. Freitas

    Understanding the crucial differences between classification and discovery of association rules – A position paper

    SIGKDD Explorations

    (2000)
  • B. Ganter et al.

    Formal Concept Analysis: Mathematical Foundations

    (1999)
  • G. Gediga, I. Düntsch, Modal-style operators in qualitative data analysis, in: Proceedings of the 2002 IEEE...
  • U. Grenander

    General Pattern Theory

    (1993)
  • H. Han, E. Manavoglu, G.C. Lee, H.Y. Zha, Rule-based word clustering for text classification, in: Proceedings of the...
  • A.K. Jain et al.

    Data clustering: a review

    ACM Computing Surveys

    (1999)
  • A.K. Jain et al.

    Statistical pattern recognition: a review

    IEEE Transactions on Pattern Analysis and Machine Intelligence

    (2000)
  • V. Jeffries et al.

    Social Stratification: A Multiple Hierarchy Approach

    (1980)
  • B. Jeon et al.

    Partially supervised classification using weighted unsupervised clustering

    IEEE Transactions on Geoscience and Remote Sensing

    (1999)
  • B. King

    Step-wise clustering procedures

    Journal of the American Statistical Association

    (1967)
  • S. Kotsiantis et al.

    Recent advances in clustering: a brief survey, WSEAS

    Transactions on Information Science and Applications

    (2004)
  • S.R. Kulkarni et al.

    Learning pattern classification – a survey

    IEEE Transactions on Information Theory

    (1998)
  • Cited by (104)

    • Textural dependency and concept lattices

      2021, International Journal of Approximate Reasoning
    • Dependence space models to construct concept lattices

      2020, International Journal of Approximate Reasoning
    View all citing articles on Scopus
    View full text