+ All Categories
Home > Documents > Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at...

Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at...

Date post: 14-Apr-2018
Category:
Upload: panayiotis-zaphiris
View: 217 times
Download: 0 times
Share this document with a friend

of 26

Transcript
  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    1/26

    SMAP 2010

    Describing and Modeling User Behaviour in Social Media

    Dr. Panayiotis Zaphiris > Associate [email protected] | http://www.zaphiris.org/

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Online Communities

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    2/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Web 2.0 Social Media

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    3/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Entertainment (social) Systems

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Virtual Worlds

    http://www.youtube.com/watch?v=pmTXGQ2BhUA

    http://www.youtube.com/watch?v=pmTXGQ2BhUA
  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    4/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    5/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    In my view in 3 steps:

    1. Analysing existing behaviour

    2. Analysing social networks formed around socialmedia

    3. Modeling/Simulating interactions

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Describing and Modeling User Behaviour inSocial Media

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    6/26

    Query based Techniques (interviews,questionnaires, focus groups)

    Usability and Accessibility Evaluations

    Content Analysis/Ethnography

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    1. Analysing existing behaviour

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Case study: Patterns of empathy in onlinecommunication for older people

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    7/26

    SeniorNet (http://www.seniornet.org/ )

    aims to educate older people about computer and internetusage

    hosts a large number of discussion groups on its website our study concentrates on the discussion board about

    depression

    400 messages (6 Aug 2000 14 Feb 2002) from 47members

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Data source

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Qualitative Content Analysis

    Read all messages

    Define the unit ofanalysis

    Read

    Extract/refinekeywords and

    themes

    Order/group thethemes

    Saturation Final codescheme

    Draft ofcode

    scheme

    Meaning

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    8/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Code Scheme

    Case Study: Trends, Similarities and Differences

    in Newsgroup Usage by Younger and OlderInternet Users

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    2. Analysing social networks formed aroundsocial media

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    9/26

    There is a correlation between social interaction

    and quality of life for seniors (Czaja, Nair et al.,1993).

    Older people who used computers thought theyhad more social interaction, memory enhancementand mental stimulation (Eilers, 1989).

    Getting a better understanding about how seniorcitizens interact online through CMC could give theresearch community insights as to how this

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Older People and Technology

    This study focuses on Computer MediatedCommunication (CMC), and investigates thesimilarities and differences in CMC usage betweenteenagers and seniors.

    Concentrate on two newsgroups (alt.teens andsoc.senior.issues) and analyse the online socialcommunities that emerged within these groups.

    Qualitative and quantitative data analysis,ethnographic techniques and SNA

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Key Objectives

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    10/26

    1. What are the properties of the networks and actors

    within the two newsgroups under investigation?2. Who is central and powerful within the social

    structure of these two newsgroups?

    3. Have any subgroups (cliques) formed within eachof the newsgroups under investigation?

    4. What are the network/group positions and social

    roles within each of the newsgroups?5. What are the significant patterns, relations and

    structures within each of the newsgroups underinvestigation?

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Social Network Analysis Questions

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    11/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    (t=11.320, p

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    12/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    (t=-5.565, p

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    13/26

    Popular topics: alt.teens (alt.abortion, alt.bible,

    politics, parents, drugs and music). soc.senior.issues (retirement, politics, military, culture,engineering housing, and health)

    Abbreviations: teens newsgroup members usemore abbreviations (t=2.868, p0.168).

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Virtual Ethnography

    200 messages were observed from eachnewsgroup and the message sender and receiver(s) recorded. The teens newsgroup had 70 uniqueauthors and the seniors newsgroup 90.

    From this data, a 70*70 adjacency matrix wasproduced for the alt.teens newsgroup data set.

    Similarly, a 90*90 adjacency matrix was producedfor the soc.seniors.issues data set.

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Social Network Analysis (SNA)

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    14/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    alt.teensSociogram

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    soc.senior.issues Sociogram

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    15/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    In Degree Centrality

    alt.teens soc.senior.issues

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Out Degree Centrality

    alt.teens soc.senior.issues

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    16/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Clique Analysis

    alt.teens soc.senior.issues

    Case Study: Modeling/Simulating interactions in

    Massively multiplayer online role-playing games(MMORPGs)

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    3. Modeling/Simulating interactions

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    17/26

    games played over networked computers, where

    potentially thousands of players interact with eachother in a real-time perpetual environment

    expected to grow from $31.63 billion in 2006 to$48.88 billion in 2011

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    MMORPGs

    Percentage of teenagers in MMORPG?

    mean age?

    Percentage of males players?

    Average hours playing a week?

    Percentage of players whose most positiveexperience over the past 7 days had been fromthe game?

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    MMORPG quiz

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    18/26

    identify social roles that emerge from

    individuals interaction in thecommunity

    examine the interaction styles for thesocial roles

    identify the relationship betweendifferent social roles

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Social Roles of WoW

    Most previous work in this area assumed theconventional social science analyticalperspective. They studied who the users are(gender, age, etc) and what the users do

    SNA: what the users do with whom? A role is relational to the group with whom one

    interacts.

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    SNA perspectives

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    19/26

    Dwelt in the game for weeks to familiarise

    ourselves with the general practice and culturearound the game

    Then we joined a relatively large guild

    Using the in-game chat-log function, we managedto keep a record of guild members chat activities

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Methods

    A total of 1944 guild messages were collected in30 hours of observation spanning across 2months.

    We categorised the messages

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Methods

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    20/26

    Group Management

    Coordination

    Ask for Help

    Give Help

    Friendly Remarks

    Game Chats

    Real Life Chats

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    More than just killing

    Tabulated the messages into socio-matrices forSNA by identifying who talked to whom

    UCiNET: CONCOR block model

    Blocking is based on similar patterns of interaction

    the players have with each other within the sameblock and with players from other blocks. (basedon structural equivalence)

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Methods

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    21/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Results

    Block ABlock B

    Block C(R2 = 0.409)

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Interaction Types of each Block

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    22/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    EI-index for each block

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Interaction between blocks

    A

    B

    C

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    23/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Social Network Simulation

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Validation

    Answer/question network Social chat network

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    24/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Validation

    we carried out virtual experiments with the simulation toidentify the relationship between the parameters and thesocial network characteristics.

    The value of the five parameters was varied at 11 levelsindependently and 30 social networks were generated for

    each level. The average of the SNA measures was calculated for each

    set of the 30 social networks.

    A total of 30 x 11 x 5 (parameters) = 1650 social networkswere produced for the main analysis.

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Virtual experiments

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    25/26

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Example of results

    Understanding the mechanisms and the structure of socialnetworks around social media.

    Answering research questions about the structure, evolution orcollapse of online social networks.

    Describing and modeling/predicting user behaviour and providinguser specific adaptations to the interface. E.g. provide userinterface elements and mechanisms to encourage non active usersto participate in the social network.

    Studying success or failure of user interface changes without riskydeployment

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Semantic Media Adaptation and Personalization

  • 7/30/2019 Zaphiris - 2010 - Describing and Modeling User Behaviour in Social Media - Keynote Presentation at SMAP 2010 - 5

    26/26

    Need for triangulation of methods

    Study of user behaviour in Social Media needsmultidisciplinary approaches

    Social simulation can provide predictive answersto research and design questions

    Dr. Panayiotis Zaphiris | [email protected]

    SMAP 2010

    Discussion/Conclusion


Recommended