20
Improving Personal Tagging Consistency through Visualization of Tag Relevancy Dr. Qin Gao*, Yusen Dai, and Kai Fu Institute of Human Factors & Ergonomics Dept. of Industrial Engineering, Tsinghua University I International 2009 24 July 09, San Diego, CA, USA

Improving Personal Tagging Consistency Through Visualization Of Tag

  • Upload
    qin-gao

  • View
    108

  • Download
    5

Embed Size (px)

DESCRIPTION

 

Citation preview

Page 1: Improving Personal Tagging Consistency Through Visualization Of Tag

Improving Personal Tagging Consistency through

Visualization of Tag Relevancy

Dr. Qin Gao*, Yusen Dai, and Kai FuInstitute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua

University

HCI International 200919-24 July 09, San Diego, CA, USA

Page 2: Improving Personal Tagging Consistency Through Visualization Of Tag

Content

Introduction

Conclusion

Research Question

Methodology

Results & Discussion

Tagging consistency is important for users to organize things effectively and to retrieve them efficiently later on.

Tag A Tag B

Page 3: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Introduction

Tagging has emerged as a new means of information organization and retrieval

Tagging is easy to use, flexible, able to harvest the intelligence of the crowdBut there are many inconsistencies in tagging systems!

Content 1

Content 4

Content 3

Content 2

Tag 1

Tag 4

Tag 3

Tag 2

Tripartite model of tagging system, from Halpin, Robu, & Sherpherd, 2007

Page 4: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Introduction

Vocabulary problems“Bad” tags: misspelt tags, badely encoded tags, mixed use of singulars and plurals, and etc.Inevitable semantic inconsistency: polysemy, synonym, and basic level variations. (Golder & Huberman 2006)

Consistency between taggersThe extent to which different users agree on selection for certain tags for specific content.

Allowing a true representation of knowledge and multiple interpretations of the same content.Trends towards stabilization (Golder & Huberman, 2005)

Consistency within individual taggersThe extent to which individual users agree on selection for certain tags for specific content at different point in time.

Page 5: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Introduction

Consistency within individual taggers is important to individual users and to the system.

Affecting efficiency of information organization and retrieval tasks for individual users

Organizing information is one of the most motivation for tagging (Ames and Naaman, 2007; Marlow, et al., 2006).Indexing research shows that reliance on consistently used indexing cues is desired for effective access of information

Impacts on users’ perceived usefulness of the system and their satisfaction.

How to improve individual tagging consistency?Providing tag suggestions based on existing tagging pattern can shape users’ tagging behavior (Sen et al, 2006; Binkowski, 2006)

How to present such suggestions?

How to select tags for suggestion?

Page 6: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Visualization of Tags

The first generation of tag clouds

Tag popularity is represented by visual cues

The second generation of tag clouds

Semantic relations among tags is revealed by visualization

Semantically clustering of tags by Montero & Solana (2006)Tag clouds from Amazon, from Bateman 2007

Nielson, 2007

Page 7: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Research Question

Goal of the study: to examine the effect of tag frequency visualization and semantically clustering on users’ tagging consistency

Hypothesis 1: visualization of occurrence frequency of tags improves personal tag consistency and reduces users’ workload.

Hypothesis 2: visualization of inter-tag relevancy improves personal tag consistency.

Page 8: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Methodology

2*2 experiment design

Page 9: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Methodology

Frequency visualization by font size

Font size level

Font size (px)

1 122 203 284 365 446 527 60

the font size was determined by the following logarithm function

6log( )1

log(120)i

i

OCurrent

Currenti is the font size level of the current tagOi is the use frequency of the current tag

The relationship between font size level and tag frequency

Definition of font size levels

Page 10: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Methodology

Visualization of tag relevancy – Semantically clustering

Clusters of relevant tags were calculated based on co-occurrence similarity with K-means algorithm developed by Montero and Solana (2006).

The approach was proved to reduce semantically density of tag clouds significantly.

ti=(d1i, d2i, d3i,

…, dni)

Definition of the vector space:ti=(d1i, d2i, … dni)cosine (t1, t2)=(t1·t2)/‖t1‖*‖t2‖

Page 11: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Methodology

Dependent variablesTagging consistency

Let Ai and Bi denote the sets of tags that assigned to the same document in two sessions, then tagging consistency with this document:

The overall tagging consistency:

Workload measured by NASA-TLX

| |( , )

| |i i

ii i

A BO A B

A B

1( , )

n

iiO A B

On

Let

and

in two sessions

Ai and Bi denote the sets of tags that assigned to the same document in two different tagging sessions

in two sessions

Page 12: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Methodology

Stimuli100 pictures selected from Flickr, tagged as “nature”, “city”, or “people”20 were stimuli, and other 80 were filler pictures

Participants40 participants, including 10 females and 30 males, aged from 20 to 31All are experienced tagging users

ProcedureTwo tagging sessions, with a disruptive interval in between.

Page 13: Improving Personal Tagging Consistency Through Visualization Of Tag

Results

Testing of hypothesis 1No frequency

visualization

(N=20)

With frequency

visualization (N=20)

F(1,36)p

M SD M SD

No. of tags in 1st

session

47.3 15.35 46.3 19.48 χ2 = 0.12a .73

No. of tags in 2nd

session

46.4 13.20 46.7 18.39 <0.01 .95

Consistency 0.72 0.116 0.69 0.145 0.78 .38

Workload

Mental demand 42.0 16.98 53.0 15.55 χ2 = 4.09 a .04*

Physical demand 33.9 18.97 22.4 15.84 χ2 = 4.08 a .04*

Temporal demand 32.3 17.86 44.8 20.10 χ2 = 2.93 a .08

Performance 40.4 19.85 35.87 21.89 0.49 .49

Effort 59.4 19.06 62.0 21.28 χ2 = 0.37 a .54

Frustration level 22.3 22.45 32.3 24.03 χ2 = 2.14 a .14

Global 43.0 12.73 46.0 12.35 0.52 .47

aKruskal-Wallis-test.*Significant differences at p<.05

aKruskal-Wallis-test.*Significant differences at p<.05

aKruskal-Wallis-test.*Significant differences at p<.05

aKruskal-Wallis-test.*Significant differences at p<.05

Page 14: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Results

Frequency visualization has no significant impact on tagging consistency.Frequency visualization reduces perceived physical demand significantly, but also increases mental demand.

An interaction effect on physical demand (χ2 = 6.4, p = .01)

27.8 27.8

40.0

11.7

0

5

10

15

20

25

30

35

40

45

No Yes

Physical demand

Frequency Visualization

No semantic clustering visualization

With semantic clustering visualization

Page 15: Improving Personal Tagging Consistency Through Visualization Of Tag

Results

Testing of Hypothesis 2No clustering

visualization (N=20)

With clustering

visualization (N=20)

F(1,36)p

M SD M SD

No. of tags in 1st

session

49.4 18.65 44.2 15.92 χ2 = 0.84a .36

No. of tags in 2nd

session

48.6 17.80 44.6 13.70 0.62 .43

Consistency 0.67 0.126 0.75 0.127 4.0 .05*

Workload

Mental demand 51.1 16.47 43.8 17.17 χ2 = 2.39a .12

Physical demand 27.8 18.51 28.5 18.37 χ2 = 0.04 a .82

Temporal demand 41.9 20.69 35.2 18.78 χ2 = 1.27 a .26

Performance 32.8 18.28 43.5 22.12 2.81 .10

Effort 63.3 17.32 58.2 22.50 χ2 = 0.24 a .62

Frustration level 26.8 20.66 27.8 26.58 χ2 = 0.10 a .74

Global 45.2 10.62 45.9 14.33 0.09 .76aKruskal-Wallis-test.*Significant differences at p<.05

Page 16: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Results

Semantically clustering improves personal tagging significantly. H2 was supported.

But no significant difference in workload or the number of tags given by participants.

The consistency level of participants tagging with semantically clustering is 12% higher than that of participants tagging without such visualization.

Page 17: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Discussion

Two types of tagsGeneral categorical tags, influenced by the basic level

High recall but low accuracyUsers have a strong bias to use them as first tags (Golder & Huberman, 2005).Relatively more consistent.

Descriptive/specific tags, ego-centeredHigh accuracy but low recall Major source of

inconsistencies

All participants expressed their intention to tag consistently, but often failed to do so due to limited memory.

Page 18: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Discussion

Semantically clustering of tags helps users’ tag formulation tasks and improves their consistency in identifying and deciding on specific tags

It improves the performance of specific search and increase the attention towards tags in small fonts compared to other layouts (Schrammel et al., 2009).

Frequency visualization does not provide support for search of specific tags.

When used in combination with semantically clustering, it help reduce perceived physical demand.

Page 19: Improving Personal Tagging Consistency Through Visualization Of Tag

Dr. Qin Gao, Institute of Human Factors & ErgonomicsDept. of Industrial Engineering, Tsinghua University

Conclusion

Visualizing the relevancy among tags has a significant positive effect on tagging consistency, whereas visualizing tagging frequency does not.

Empirical support for the effort of visualizing semantic relationships among tags

When the tag relevancy is visualized, highlight frequently used tags can reduce perceived physical demands; however, it increases perceived mental demands as well.

Implications for professional indexer aid design.

Page 20: Improving Personal Tagging Consistency Through Visualization Of Tag

Thank you for your attention.

Contact:[email protected]

http://trisha.snappages.com

Q & A