Telligent Social Analytics Research & Tools

Post on 31-Dec-2015

16 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

DESCRIPTION

Telligent Social Analytics Research & Tools. Marc A. Smith Chief Social Scientist Telligent Systems. E-mail (and more) is from people to people. Patterns are left behind. Social Network Theory. Central tenet Social structure emerges from the aggregate of relationships (ties) - PowerPoint PPT Presentation

Transcript

Telligent Social AnalyticsResearch & Tools

Marc A. SmithChief Social Scientist

Telligent Systems

Page 2

E-mail (and more) is from people to people

2

Page 3

Patterns are left behind

3

Page 4

SOCIAL NETWORKTHEORYCentral tenet

Social structure emerges from the aggregate of relationships (ties) among members of a population

Phenomena of interestEmergence of cliques and clusters

from patterns of relationshipsCentrality (core), periphery (isolates),

betweennessMethods

Surveys, interviews, observations, log file analysis, computational analysis of matrices

(Hampton &Wellman, 1999; Paolillo,2001; Wellman, 2001)

Source: Richards, W. (1986). The NEGOPY network

analysis program. Burnaby, BC: Department of

Communication, Simon Fraser University. pp.7-16

Page 5

Social media platforms are a source of multiple Social network data sets:

“Friends”“Replies”“Follows”

“Comments”“Reads”

“Co-edits”“Co-mentions”

“Hybrids”

Page 6

Journal of Social Structure: “Visualizing the Signatures of Social Roles in Online Discussion Groups”

http://www.cmu.edu/joss/content/articles/volume8/Welser/

Distinguishing attributes:Answer person

Outward ties to local isolatesRelative absence of triangles

Few intense ties

Reply MagnetTies from local isolates

Often inward only Sparse, few triangles

Few intense ties

Page 7

Journal of Social Structure: “Visualizing the Signatures of Social Roles in Online Discussion Groups”

http://www.cmu.edu/joss/content/articles/volume8/Welser/

Distinguishing attributes:Answer person

Outward ties to local isolatesRelative absence of triangles

Few intense ties

Discussion personTies from local isolates

often inward onlyDense, many trianglesNumerous intense ties

Page 9

Use social network analysis measurements in reporting on

social media data.

Analytics calculates network metrics for all content authors.

In-degreeOut-degree

Eigenvector centralityClustering coefficient

Ingredients of User Type Scores

Page 10

Social media usage generatesSocial Networks

Page 11

Display community members sorted by network attributes using Excel Data|Sort

Page 13

User type reports in Telligent Analytics Include social network metrics to define different kinds of contributors:

Answerer: users who reply to many questions from many people. Influencer: users who are connected to other well connected users.Asker: users who raise questions that get answered by answer people.Connector: highly connected users who are replied to or linked to by many other community users.Originator: initiates new content in the site that is often linked to by others.Commenter: replies or links to content created by others.Spectator: reads but tends not to create content.Overseer: moderates content created by others.

Page 14

NodeXL: Network Overview, Discovery and Exploration for Excel

Leverage spreadsheet for storage of edge and vertex data

http://www.codeplex.com/nodexl

Page 15

The NodeXL Team

Page 16

The NodeXL project is Available via the CodePlexOpen Source Project Hosting Site:

Site:http://www.codeplex.com/nodexl

Page 17

Page 18

NodeXL:Display nodes with subgraph images sorted by network attributes using Excel Data|Sort

Page 19

Resources to supportEducational Use ofNodeXL

Free Tutorial/Manual

Data SetsAvailable

Page 20

Page 21

Page 22

NodeXL: Filtered clusters

Page 23

NodeXL: Import social networks from email

Page 24

NodeXL: Import social networks from email

Page 25

Page 26

Social Network Analysis Engine Development: NodeXL

Extend and apply social network analysis engine:

Improve layouts and visualizationsAdditional metrics and measuresTechnical architecture shift to the web and cloudScale and performanceClustering and time series analysis

Page 27

NodeXL Partnerships and community

University of MarylandOhio University

Stanford UniversityUniversity of Pennsylvania

YOU?

10,000 + downloads on Codeplex

Page 28

Telligent Analytics

Provides a source of network edge lists and integrates social network metrics in User Type Scores

Further possible social network analysis applicationsRecommendations: my friend’s edit what documents?

Search optimization: show documents from “answer people”Role discovery: who are the topic starters? The answer people?

Page 29

Telligent Social AnalyticsResearch & Tools

Marc A. SmithChief Social Scientist

Telligent Systems

Marc.Smith@Telligent.comhttp://www.telligent.com

http://www.connectedaction.net

top related