عنوان

Big data integration /

(Number (ISBN

1627052232

(Number (ISBN

9781627052238

Number

dltt

Title Proper

Big data integration /

General Material Designation

[Book]

First Statement of Responsibility

Xin Luna Dong, Google Inc., Divesh Srivastava, AT&T Labs-Research

Edition Statement

First edition

Specific Material Designation and Extent of Item

xx, 178 pages :

Other Physical Details

color illustrations ; 24 cm

Series Title

Synthesis lectures on data management,

Volume Designation

#40

ISSN of Series

2153-5418 ;

Text of Note

Includes bibliographical references (pages 165-173) and index

Text of Note

1. Motivation: challenges and opportunities for BDI -- 1.1 Traditional data integration -- 1.1.1 The flights example: data sources -- 1.1.2 The flights example: data integration -- 1.1.3 Data integration: architecture & three major steps -- 1.2 BDI: challenges -- 1.2.1 The "V" dimensions -- 1.2.2 Case study: quantity of deep web data -- 1.2.3 Case study: extracted domain-specific data -- 1.2.4 Case study: quality of deep web data -- 1.2.5 Case study: surface web structured data -- 1.2.6 Case study: extracted knowledge triples -- 1.3 BDI: opportunities -- 1.3.1 Data redundancy -- 1.3.2 Long data -- 1.3.3 Big data platforms -- 1.4 Outline of book --

Text of Note

2. Schema alignment -- 2.1 Traditional schema alignment: a quick tour -- 2.1.1 Mediated schema -- 2.1.2 Attribute matching -- 2.1.3 Schema mapping -- 2.1.4 Query answering -- 2.2 Addressing the variety and velocity challenges -- 2.2.1 Probabilistic schema alignment -- 2.2.2 Pay-as-you-go user feedback -- 2.3 Addressing the variety and volume challenges -- 2.3.1 Integrating deep web data -- 2.3.2 Integrating web tables --

Text of Note

3. Record linkage -- 3.1 Traditional record linkage: a quick tour -- 3.1.1 Pairwise matching -- 3.1.2 Clustering -- 3.1.3 Blocking -- 3.2 Addressing the volume challenge -- 3.2.1 Using MapReduce to parallelize blocking -- 3.2.2 Meta-blocking: pruning pairwise matchings -- 3.3 Addressing the velocity challenge -- 3.3.1 Incremental record linkage -- 3.4 Addressing the variety challenge -- 3.4.1 Linking text snippets to structured data -- 3.5 Addressing the veracity challenge -- 3.5.1 Temporal record linkage -- 3.5.2 Record linkage with uniqueness constraints --

Text of Note

4. BDI: data fusion -- 4.1 Traditional data fusion: a quick tour -- 4.2 Addressing the veracity challenge -- 4.2.1 Accuracy of a source -- 4.2.2 Probability of a value being true -- 4.2.3 Copying between sources -- 4.2.4 The end-to-end solution -- 4.2.5 Extensions and alternatives -- 4.3 Addressing the volume challenge -- 4.3.1 A MapReduce-based framework for offline fusion -- 4.3.2 Online data fusion -- 4.4 Addressing the velocity challenge -- 4.5 Addressing the variety challenge --

Text of Note

5. BDI: emerging topics -- 5.1 Role of crowdsourcing -- 5.1.1 Leveraging transitive relations -- 5.1.2 Crowdsourcing the end-to-end workflow -- 5.1.3 Future work -- 5.2 Source selection -- 5.2.1 Static sources -- 5.2.2 Dynamic sources -- 5.2.3 Future work -- 5.3 Source profiling -- 5.3.1 The Bellman system -- 5.3.2 Summarizing sources -- 5.3.3 Future work --

Text of Note

6. Conclusions -- Bibliography -- Authors' biographies -- Index

0

8

8

8

8

8

Text of Note

The big data era is upon us: data are being generated, analyzed, and used at an unprecedented scale, and data-driven decision making is sweeping through all aspects of society. Since the value of data explodes when it can be linked and fused with other data, addressing the big data integration (BDI) challenge is critical to realizing the promise of big data. BDI differs from traditional data integration along the dimensions of volume, velocity, variety, and veracity. First, not only can data sources contain a huge volume of data, but also the number of data sources is now in the millions. Second, because of the rate at which newly collected data are made available, many of the data sources are very dynamic, and the number of data sources is also rapidly exploding. Third, data sources are extremely heterogeneous in their structure and content, exhibiting considerable variety even for substantially similar entities. Fourth, the data sources are of widely differing qualities, with significant differences in the coverage, accuracy and timeliness of data provided. This book explores the progress that has been made by the data integration community on the topics of schema alignment, record linkage and data fusion in addressing these novel challenges faced by big data integration. Each of these topics is covered in a systematic way: first starting with a quick tour of the topic in the context of traditional data integration, followed by a detailed, example-driven exposition of recent innovative techniques that have been proposed to address the BDI challenges of volume, velocity, variety, and veracity. Finally, it presents emerging topics and opportunities that are specific to BDI, identifying promising directions for the data integration community

Big data

Data integration (Computer science)

Number

Edition

Class number

Book number

Dong, Xin Luna,1975-

Srivastava, Divesh

Date of Transaction

20160129110409.0

Cataloguing Rules (Descriptive Conventions))

rda

Electronic name

[Book]

Y