How Could End-Users Identify Interesting Resources?

Preview:

DESCRIPTION

Slides of my talk at the HSWI workshop about a concept allowing users to identify interesting resources on objective facets.

Citation preview

Madrid, 14.06.2013 Martin Voigt

How Could End-Users Identify Interesting Resources?

Attract them visually!

Titel der Präsentation Folie Nr. 2 von XYZ How to find interesting resources? | Martin Voigt 2

Structure

Motivation and problems

Iterative development of a user interface

Demo

Concept and implementation of a back-end

Conclusion

14.06.2013

Titel der Präsentation Folie Nr. 3 von XYZ How to find interesting resources? | Martin Voigt 3

Growing amount of data

Motivation

14.06.2013

[Key2012]

Titel der Präsentation Folie Nr. 4 von XYZ How to find interesting resources? | Martin Voigt 4

How to get insights from the data?

Motivation

14.06.2013

Titel der Präsentation Folie Nr. 5 von XYZ How to find interesting resources? | Martin Voigt 5

How to get insights from the data?

Information Visualization!

Motivation

14.06.2013

Titel der Präsentation Folie Nr. 6 von XYZ How to find interesting resources? | Martin Voigt 6

Motivation

BUT

14.06.2013

InfoVis for Semantic Web data?

Titel der Präsentation Folie Nr. 7 von XYZ How to find interesting resources? | Martin Voigt 7

Problem

End-users have no idea about creating effective InfoVis and have no programming skills.

14.06.2013

Titel der Präsentation Folie Nr. 8 von XYZ How to find interesting resources? | Martin Voigt 8

Overall concept of VizBoard

Semantic-assisted InfoVis-Process [Voigt2013]

How to find interesting resources?

objective vs. subjective interestingness [Silberschatz1996]

14.06.2013

Data UploadData Pre-Selection

Data & VisSelection

Visualization Configuration

Perception & Internalization

Data Augmentation

Visualization Recommendation

Visualization Integration

Knowledge Externalization

Data Reduction

Syst

em

Use

r

Human Interaction

System Interaction

31 5 7 9

2 4 6 8 10

Data UploadData Pre-Selection

Data & VisSelection

Visualization Configuration

Perception & InternalizationU

ser

Human Interaction

31 5 7 9

Titel der Präsentation Folie Nr. 9 von XYZ How to find interesting resources? | Martin Voigt 9

Main problems to solve

1st: Interactive, understandable user interface

2nd: Backend techniques to identify relevant resources

14.06.2013

Titel der Präsentation Folie Nr. 10 von XYZ How to find interesting resources? | Martin Voigt 10

Structure

Motivation and problems

Iterative development of a user interface

Demo

Concept and implementation of a back-end

Conclusion

14.06.2013

Titel der Präsentation Folie Nr. 11 von XYZ How to find interesting resources? | Martin Voigt 11

User Interface Concept

Prerequisites - Visualization

node link diagrams vs. nested views

insufficient screen estate interaction required

14.06.2013

Titel der Präsentation Folie Nr. 12 von XYZ How to find interesting resources? | Martin Voigt 12

User Interface Concept

Prerequisites - Interaction

use widely adapted techniques

14.06.2013

Titel der Präsentation Folie Nr. 13 von XYZ How to find interesting resources? | Martin Voigt 13

User Interface Concept

14.06.2013

Titel der Präsentation Folie Nr. 14 von XYZ How to find interesting resources? | Martin Voigt 14

User Interface Concept

User Study - Setup

Task Load Index to track efforts, e.g., mental or temporal

10 tasks based on “Visual Information-seeking Mantra” [Shneiderman1996]

Questionnaire for additional feedback

10 students with less/no SemWeb knowledge

14.06.2013

Titel der Präsentation Folie Nr. 15 von XYZ How to find interesting resources? | Martin Voigt 15

User Interface Concept

User Study – Results

+ node link to get to know the data structure

+ navigation history and bookmarking helpful

- too much information shown

- too many options

- SW knowledge required

functionality vs. user satisfaction

14.06.2013

Titel der Präsentation Folie Nr. 16 von XYZ How to find interesting resources? | Martin Voigt 16

User Interface Concept

14.06.2013

Titel der Präsentation Folie Nr. 17 von XYZ How to find interesting resources? | Martin Voigt 17

Structure

Motivation and problems

Iterative development of a user interface

Demo

Concept and implementation of a back-end

Conclusion

14.06.2013

Titel der Präsentation Folie Nr. 18 von XYZ How to find interesting resources? | Martin Voigt 18

Structure

Motivation and problems

Iterative development of a user interface

Demo

Concept and implementation of a back-end

Conclusion

14.06.2013

Titel der Präsentation Folie Nr. 19 von XYZ How to find interesting resources? | Martin Voigt 19

Assisting Backend

Implementation

• Java Jersey REST WS

• Jena TDB

• RMonto & RapidMiner

• JGraphT

14.06.2013

Data Repository

Storage, Filtering, Recommendation

Augmentation

Data Access

Analyzer

Storage

Annotator

RESTful API

Pivoting

Filter

upload get | filter

1 2

3

Titel der Präsentation Folie Nr. 20 von XYZ How to find interesting resources? | Martin Voigt 20

Structure

Motivation and problems

Iterative development of a user interface

Demo

Concept and implementation of a back-end

Conclusion

14.06.2013

Titel der Präsentation Folie Nr. 21 von XYZ How to find interesting resources? | Martin Voigt 21

Sum it up!

End-user need assistances to identify interesting SemWeb resources (in big datasets)

Approach using objective measures, understandable visualizations, and easy interaction

14.06.2013

Titel der Präsentation Folie Nr. 22 von XYZ How to find interesting resources? | Martin Voigt 22

Sum it up!

SWIs for end-users should be task-specific as possible

less is more: interaction, information, SW intrinsics

Visualizing huge SW datasets in browsers is cumbersome

Current tasks

Subjective interestingness

Planning an advanced user study

14.06.2013

Titel der Präsentation Folie Nr. 23 von XYZ How to find interesting resources? | Martin Voigt 23

Thanks for listening. Questions?

14.06.2013

Titel der Präsentation Folie Nr. 24 von XYZ How to find interesting resources? | Martin Voigt 24

References

[Cyganiak2011] Cyganiak, R. & Jentzsch, A.: Linking Open Data cloud diagram. http://lod-cloud.net/, 2011

[Key2012] Key, A.; Howe, B.; Perry, D. & Aragon, C.: VizDeck: self-organizing dashboards for visual analytics. Proceedings of the 2012 ACM SIGMOD International Conference on Management of Data, ACM, 2012, 681-684

[Shneiderman1996] Shneiderman, B.: The eyes have it: a task by data type taxonomy for information visualizations. Visual Languages, 1996. Proceedings., IEEE Symposium on, 1996, 336-343

[Silberschatz1996] A. Silberschatz, A. & Tuzhilin, A.: What makes patterns interesting in knowledge discovery systems. Knowledge and Data Engineering, IEEE Transactions on, 8(6):970 -974, dec 1996.

[Voigt2013] M. Voigt, S. Pietschmann, K. Meißner: A Semantics-Based, End-User-Centered Information Visualization Process for Semantic Web Data. Semantic Models for Adaptive Interactive Systems, Springer, 2013

14.06.2013

Recommended