Clustering Approach to Generalized Pattern Identification Based on Multi-instanced Objects with DARA

Rayner Alfred; Dimitar Kazakov

Clustering Approach to Generalized Pattern Identification Based on Multi-instanced Objects with DARA

Computer Science

Research output: Contribution to conference › Paper › peer-review

Abstract

Clustering is an essential data mining task with various types of
applications. Traditional clustering algorithms are based on a vector space
model representation. A relational database system often contains multirelational
information spread across multiple relations (tables). In order to
cluster such data, one would require to restrict the analysis to a single
representation, or to construct a feature space comprising all possible
representations from the data stored in multiple tables. In this paper, we present
a data summarization approach, borrowed from the Information Retrieval
theory, to clustering in multi-relational environment. We find that the data
summarization technique can be used here to capture the typical high volume of
multiple instances and numerous forms of patterns. Our experiments
demonstrate a technique to cluster data in a multi-relational environment and
show the evaluation results on the mutagenesis dataset. In addition, the effect of
varying the number of features considered in clustering on the classification
performance is also evaluated.

Original language	Undefined/Unknown
Publication status	Published - 2007

Access to Document

http://ceur-ws.org/Vol-325/paper04.pdf

Cite this

@conference{60995c32b4024c3e8ff435c3bf3e02f1,

title = "Clustering Approach to Generalized Pattern Identification Based on Multi-instanced Objects with DARA",

abstract = "Clustering is an essential data mining task with various types ofapplications. Traditional clustering algorithms are based on a vector spacemodel representation. A relational database system often contains multirelationalinformation spread across multiple relations (tables). In order tocluster such data, one would require to restrict the analysis to a singlerepresentation, or to construct a feature space comprising all possiblerepresentations from the data stored in multiple tables. In this paper, we presenta data summarization approach, borrowed from the Information Retrievaltheory, to clustering in multi-relational environment. We find that the datasummarization technique can be used here to capture the typical high volume ofmultiple instances and numerous forms of patterns. Our experimentsdemonstrate a technique to cluster data in a multi-relational environment andshow the evaluation results on the mutagenesis dataset. In addition, the effect ofvarying the number of features considered in clustering on the classificationperformance is also evaluated.",

author = "Rayner Alfred and Dimitar Kazakov",

year = "2007",

language = "Undefined/Unknown",

}

TY - CONF

T1 - Clustering Approach to Generalized Pattern Identification Based on Multi-instanced Objects with DARA

AU - Alfred, Rayner

AU - Kazakov, Dimitar

PY - 2007

Y1 - 2007

N2 - Clustering is an essential data mining task with various types ofapplications. Traditional clustering algorithms are based on a vector spacemodel representation. A relational database system often contains multirelationalinformation spread across multiple relations (tables). In order tocluster such data, one would require to restrict the analysis to a singlerepresentation, or to construct a feature space comprising all possiblerepresentations from the data stored in multiple tables. In this paper, we presenta data summarization approach, borrowed from the Information Retrievaltheory, to clustering in multi-relational environment. We find that the datasummarization technique can be used here to capture the typical high volume ofmultiple instances and numerous forms of patterns. Our experimentsdemonstrate a technique to cluster data in a multi-relational environment andshow the evaluation results on the mutagenesis dataset. In addition, the effect ofvarying the number of features considered in clustering on the classificationperformance is also evaluated.

AB - Clustering is an essential data mining task with various types ofapplications. Traditional clustering algorithms are based on a vector spacemodel representation. A relational database system often contains multirelationalinformation spread across multiple relations (tables). In order tocluster such data, one would require to restrict the analysis to a singlerepresentation, or to construct a feature space comprising all possiblerepresentations from the data stored in multiple tables. In this paper, we presenta data summarization approach, borrowed from the Information Retrievaltheory, to clustering in multi-relational environment. We find that the datasummarization technique can be used here to capture the typical high volume ofmultiple instances and numerous forms of patterns. Our experimentsdemonstrate a technique to cluster data in a multi-relational environment andshow the evaluation results on the mutagenesis dataset. In addition, the effect ofvarying the number of features considered in clustering on the classificationperformance is also evaluated.

M3 - Paper

ER -