MultiRelationalDataMiningTheCurrentFrontiers资源-CSDN文库

数据挖掘

需积分: 9 51 浏览量 2008-04-16 21:53:44 上传评论收藏 40KB PDF 举报

资源推荐

资源详情

资源评论

Multi-Relational Data Mining: The Current Frontiers

so D

zeroski

Jozef Stefan Institute

Jamova 39

SI-1000 Ljubljana, Slovenia

saso.dzeroski@ijs.si

Luc De Raedt

Institut f

ur Informatik, Albert-Ludwigs-University,

Georges-Koehler-Allee, Building 079, D-79110

Freiburg, Germany

deraedt@informatik.uni-freiburg.de

Data mining algorithms look for patterns in data. Most

existing data mining approaches are propositional and look

for patterns in a single data table. Most real-world databases,

however, store information in multiple tables.

Relational data mining (RDM) approaches (Dˇzeroski and

Lavraˇc 2001), look for patterns that involve multiple tables

(relations) from a relational database. To emphasize this

fact, RDM is often referre d to as multi-relational data min-

ing (MRDM) (Dˇzeroski et al. 2002). We will adopt this

term in the present special issue.

When we are looking for patterns in multi-relational

data, it is natural that the patterns involve multiple rela-

tions. They are typically stated in a more expressive lan-

guage than patterns deﬁned on a single data table. The

major types of multi-relational patterns extend the types of

propositional patterns considered in single table data min-

ing. We can thus have multi-relational classiﬁcation rules,

multi-relational regression trees, and multi-relational asso-

ciation rules, among others.

Just as many data mining algorithms come from the

ﬁeld of machine learning, many MRDM algorithms come

form the ﬁeld of inductive logic programming (ILP, Muggle-

ton 1992; Lavraˇc and Dˇzeroski 1994). Situated at the in-

tersection of machine learning and logic programming, ILP

has been concerned with ﬁnding patterns e xpres sed as logic

programs. Initially, ILP focusse d on automated program

synthesis from examples, formulated as a binary classiﬁca-

tion task. In recent years, however, the scope of ILP has

broadened to cover the whole spectrum of data mining tasks

(classiﬁcation, regre ssion, clustering, association analysis).

The most common types of patterns have been extended to

their multi-relational versions and so have the major data

mining algorithms (decision tree induction, distance-based

clustering and prediction, etc.).

There is also a growing interest in the development of

data mining algorithms for various types of structured data.

These include, for example, graph-based data mining. There

is also an increasing body of work on mining tree-structured

and XML documents. Mining data which consists of com-

plex/structured objects also falls within the scope of MRDM,

as the normalized repres entation of such objects in a rela-

tional database requires multiple tables.

The rise of several KDD application areas that are in-

trinsically relational has provided and continues to provide a

strong motivation for the development of MRDM approaches.

These include in the ﬁrst place bioinformatics and more

broadly computational biology. The World Wide Web and

data mining tasks related to it, such as information extrac-

tion follow closely. Related tasks, such as social network

analysis also hold signiﬁcant importance.

In this special issue, we ﬁrst provide an introduction to

multi-relational data mining (Dˇzeroski 2003). An extensive

overview of the ﬁeld is given by Dˇzeroski and Lavraˇc (2001).

The overview in this issue focusses on the most important

approaches, giving a brief introduction to inductive logic

programming, then covering several multi-relational data

mining tasks and techniques, such as multi-relational as-

sociation rules and multi-relational decision trees.

Most of the full-length contributions in this special issue

cover important recent advances at the frontiers of multi-

relational data mining. These include probabilistic logic

learning (De Raedt and Kersting 2003), kernel-based learn-

ing for structured data (G¨artner 2003), and graph-based

data mining (Washio and Motoda 2003). Blockeel and Se-

bag (2003) survey scalability and eﬃciency issues in MRDM

and approaches taken to resolving these. Finally, Page and

Craven (2003) survey several applications of MRDM in the

area of bioinformatics, the area where MRDM has had most

successes so far and which still holds many challenges for

MRDM.

In addition to the six full-length contributions, this spe-

cial issue contains three shorter articles/position statements.

While Washio and Motoda (2003) mainly survey methods

for frequent subgraph discovery, Holder and Cook (2003) dis-

cuss current and future directions in graph-based relational

learning, which also addresses other data mining tasks, such

as classiﬁcation. Getoor (2003) introduces and discusses sev-

eral tasks of link mining, where links/relations among ob-

jects are of central importance. Finally, Domingos (2003)

discusses several applications areas for MRDM, as well as

challenges that MRDM must address to be successful in

these.

A summer school on relational data mining, covering

both basic and advanced topics in this area, was organized

in Helsinki in August 2002, preceding ECML /PKDD 2002

(The 13th European Conference on Machine Learning and

The 6th European Conference on Principles and Practice

of Knowledge Discovery in Databases). The slides from

this summer school are available online (Dˇzeroski and

Zenko

2002). A report on this event by Dˇzeroski and

Zenko (2003)

is included at the end of this special issue.

本内容试读结束，登录后可阅读更多

下载后可阅读完整内容，剩余1页未读，立即下载

评论收藏

内容反馈

morre

粉丝: 187
资源: 2337

MultiRelational Data Mining The Current Frontiers

最新资源

MultiRelational Data Mining The Current Frontiers

MultiRelational Data Mining

MultiRelational Data Mining 2005 Workshop Report

MultiRelational Data Mining An Introduction

Machine Learning and Data Mining in Pattern Recognition.

Frontiers in massive Data Analysis

Frontiers in Data Science

UNIX Internals: The New Frontiers (Paperback)

Frontiers in Massive Data Analysis

Unix_Internals_The_New_Frontiers

Frontiers_in_Massive_Data_Analysis

UNIX Internals The New Frontiers 深入理解UNIX系统内核 中文完整扫描版 PART2

大数据分析前沿

big data - Frontiers in Massive Data Analysis

[UNIX Internals][The New Frontiers][Uresh Vahalia][En]

Machine Learning for Big Data Analyis

Big.Data.and.Business.Analytics.1466565780

UNIX Internals : The New Frontiers PDF Uresh Vahalia

Linear Algebra Foundations to Frontiers

Advanced Topics in Analysis of Economic and Financial Data Using R and SAS

New Frontiers in High Performance Computing and Big Data 无水印原版pdf

frontiers2.0

第一讲：python网络爬虫基础知识

Python基于机器学习实现的股票价格预测、股票预测源码+数据集，机器学习大作业

身份证前6位对应的省市区代码（超详细）

【Python实战】-Python+Opencv是实现车牌自动识别（源码+数据+字符匹配模板）

抖音用户浏览行为数据集

两阶段鲁棒优化/综合能源系统/需求响应/微电网/多目标优化/优化调度matlab-yalmip-cplex/gurobi文章复现

regress函数实例代码

SPSS中介效应分析插件（Process和mediate插件）

最新资源

UNIX Internals The New Frontiers 深入理解UNIX系统内核中文完整扫描版 PART2