Learning from multiple social networks, Liqiang Nie, Xuemeng Song, and Tat-Seng Chua

Resource Information

The instance Learning from multiple social networks, Liqiang Nie, Xuemeng Song, and Tat-Seng Chua represents a material embodiment of a distinct intellectual or artistic creation found in Sydney Jones Library, University of Liverpool. This resource is a combination of several types including: Instance, Electronic.

Image

http://covers.openlibrary.org/b/isbn/9781627059862-M.jpg?default=false

Label

Learning from multiple social networks, Liqiang Nie, Xuemeng Song, and Tat-Seng Chua

Statement of responsibility

Liqiang Nie, Xuemeng Song, and Tat-Seng Chua

Instantiates

• Learning from multiple social networks

Publication

• San Rafael, California 1537 Fourth Street, San Rafael, CA 94901 USA, Morgan & Claypool, 2016

Bibliography note

Includes bibliographical references (pages 87-100)

Carrier category

online resource

Carrier MARC source

rdacarrier

Color

multicolored

Content category

text

Content type MARC source

rdacontent

Contents

• 1. Introduction -- 1.1 Background -- 1.2 Motivation -- 1.3 Challenges -- 1.4 Our solutions and applications -- 1.5 Outline of this book --
• 2. Data gathering and completion -- 2.1 User accounts alignment -- 2.2 Missing data problems -- 2.3 Matrix factorization for data completion -- 2.4 Multiple social networks data completion -- 2.5 Summary --
• 3. Multi-source mono-task learning -- 3.1 Application: volunteerism tendency prediction -- 3.2 Related work -- 3.2.1 Volunteerism and personality analysis -- 3.2.2 Multi-view learning with missing data -- 3.3 Multiple social network learning -- 3.3.1 Notation -- 3.3.2 Problem formulations -- 3.3.3 Optimization -- 3.4 Experimentation -- 3.4.1 Experimental settings -- 3.4.2 Feature extraction -- 3.4.3 Model comparison -- 3.4.4 Data completion comparison -- 3.4.5 Feature comparison -- 3.4.6 Source comparison -- 3.4.7 Size varying of positive samples -- 3.4.8 Complexity discussion -- 3.5 Summary --
• 4. Mono-source multi-task learning -- 4.1 Application: user interest inference from mono-source -- 4.2 Related work -- 4.2.1 Clustered multi-task learning -- 4.2.2 User interest mining -- 4.3 Efficient clustered multi-task learning -- 4.3.1 Notation -- 4.3.2 Problem formulation -- 4.3.3 Grouping structure learning -- 4.3.4 Efficient clustered multi-task learning -- 4.4 Experimentation -- 4.4.1 Experimental settings -- 4.4.2 Feature extraction -- 4.4.3 Evaluation metric -- 4.4.4 Parameter tuning -- 4.4.5 Model comparison -- 4.4.6 Necessity of structure learning -- 4.5 Summary --
• 5. Multi-source multi-task learning -- 5.1 Application: user interest inference from multi-source -- 5.2 Related work -- 5.3 Multi-source multi-task learning -- 5.3.1 Notation -- 5.3.2 Problem formulations -- 5.3.3 Optimization -- 5.3.4 Construction of interest tree structure -- 5.4 Experiments -- 5.4.1 Experimental settings -- 5.4.2 Model comparison -- 5.4.3 Source comparison -- 5.4.4 Complexity discussion -- 5.5 Summary --
• 6. Multi-source multi-task learning with feature selection -- 6.1 Application: user attribute learning from multimedia data -- 6.2 Related work -- 6.3 Data construction -- 6.3.1 Data crawling strategy -- 6.3.2 Ground truth construction -- 6.4 Multi-source multi-task learning with Fused Lasso -- 6.5 Optimization -- 6.6 Experiments -- 6.6.1 Experimental settings -- 6.6.2 Feature extraction -- 6.6.3 Overall model evaluation -- 6.6.4 Component-wise analysis -- 6.6.5 Source integration -- 6.6.6 Parameter tuning -- 6.6.7 Computational analysis -- 6.7 Other application -- 6.8 Summary --
• 7. Research frontiers -- Bibliography -- Authors' biographies

Control code

201603ICR048

Dimensions

unknown

Extent

1 PDF (xv, 102 pages)

File format

multiple file formats

Form of item

online

Isbn

9781627059862

Media category

electronic

Media MARC source

isbdmedia

Other control number

10.2200/S00714ED1V01Y201603ICR048

Other physical details

illustrations.

Record ID

b4048147

Reformatting quality

access

Specific material designation

remote

System details

System requirements: Adobe Acrobat Reader