presentatie Reputation Management & workshop PhD community

Post on 09-Dec-2014

597 views 1 download

Tags:

description

PhD meeting 27th of November / TBM faculty Delft University of Technology. Thieme Hennis Faculty of Systems Engineering, Policy Analysis and Management Delft University of Technology +31 15 278 73 71 (work) +31 6 51855 22 0 (mobile) IM/Skype username: thiemehennis

Transcript of presentatie Reputation Management & workshop PhD community

WHY?

Thiem! Laat je snor staan

voor het kankerfonds!

Het is ook erg stoer!

Wij doen het ook!

mmm…Goed idee

“How to make people (usually guys) grow a

moustache every year in November?

don’t worry, I am only trying to be funny

but…

MOTIVATION

So… there we have it

Why people generate content? Or grow moustaches?-fun-reputation, trust-feedback, reciprocity, learning-…

But how to filter or trust content?-Trust algorithms (Google PageRank, WikiTrust)-Communities/networks (Twitter, LinkedIn, …) -…

PageRank relies on the uniquely democratic nature of the web by using its vast link structure as an indicator of an individual page's value. In essence, Google interprets a link from page A to page B as a vote, by page A, for page B. But, Google looks at more than the sheer volume of votes, or links a page receives; it also analyzes the page that casts the vote. Votes cast by pages that are themselves "important" weigh more heavily and help to make other pages important.

WikiTrust uses a content-driven reputation system: authors gain reputation when their contributions are preserved by subsequent authors, and they lose reputation when their contributions are reverted.

The trust of a portion of text is computed according to the reputation of the author, and the reputation of the people who subsequently revised the portion of text, and the text immediately surrounding it. Thus, what WikiTrust calls "text trust" is an indication of the degree with which the text has been revised. …

Another dimension to reputation - not just value, according to others

88

PHP (54)

Photography (4)

Ruby on Rails (40)

Project management (0.8)

Social skills (0.4) Writing skills (0.3)Dynamically

updated!

- but a contextualized value, assigned to1. a concept or topic2. a process or competency

Idea generation (0.3)

… but how?

Research goal Example Analysis

Action (contribution) Define the actions and contributions that (could) mean something in terms of expertise or competencies

-Answers on questions-Reviews of papers-Skype conversations

-16 answers-4 reviews-6 Skype conversations

Context of action (contribution)

Define the context relevant for this contribution: knowledge involved?

-Action itself: guiding, writing, formulating answer-Topic context: in-depth or superficial understanding of topic

-Activity focus level: providing feedback, guidance-Topic focus level: logistics, six sigma

action

Competencies (process)What kind of action?Which competencies involved?

Knowledge (topic)What kind of contribution?What kind of topic?

EX ANTE

Research goal Example Analysis

Reaction (interaction) Define the interactions with the contribution that relate to value (popularity, rating, recommending, thanking, etc.) or to context (tagging, categorizing)

-Rating of answers-Number of views-Times used in collection-Number of “thanks”

-Avg rating 4 out of 5, 12 “best answers”-452 views-33 contributions integrated in a course-12 “thank you’s”

Evaluate importance of interaction

Evaluate the value of the different interactions, and determine the relevance and importance of each interaction.

-1 thank you ~ 1 time used in collection-10 positive ratings ~ 1 thank you-10 views ~ 1 rating

For 1 contribution:- 3 times in collections (300), 4 thank you’s (400), 45 thumbs up (450), and 850 views (850) == 2000 points

Evaluate context parameters that can be extracted from the reaction

Find out how the context of each interaction can be analyzed and define context parameters.

-tags assigned to contribution-personal prodfiles of person involved in interaction-special attributed extracted from interaction (helpful, funny, etc.)

For the same contribution:-tagged SixSigma (6), logistics (7), project management (5), China (2) >> Resulting profile:- Logistics (700), SixSigma (600), project management (500), China (200).

reaction: -evaluate value use/rate/recommend-contextualize categorize/tag/embed

EX POSTERIORI based on reactions

createuse/rate/recommend/tag

profile/reputation: every value interaction contains one or more context parameters

reputation

writepaper

. reference (quality & context))

. rating (quality)

. tag (context)

expertise profile

Example: Academic literature… finding the right reviewerAcademics write papers and do reviewsA reviewer that has a good reputation can ask more moneyThe reputation of a reviewer is based on

- The quality of his own papers (expertise)- The quality of his reviews (review competency)- more… (recommendations?)

The value of a paper is based on the review(s), references, and ratings

The context of a paper is based on the review, keywords, journal keywords, and social tagging

- Keywords of the paper itself- Keywords of the paper referring to the paper

researcher I

evaluate quality of

review

writereview

review competency

. review (quality & context)

Research goal AnalysisAction (contribution) Write a paper about a topic -1 paper

Context of action (contribution) Topic is defined by keywords -Keywords: logistics, sixsigma, …

action

Competencies (process)Write a paper

Knowledge (topic)Logistics, six sigma

EX ANTE

Research goal Analysis

Reaction (interaction) Interactions related to value:-Review(s)-References-Rating

Review-Originality: 8/10-Methodology {Case study}: 9/10-Usefulness: 8/10-Overall: 9/10

33 references with 85 different keywords

121 positive ratings

Resulting profile:- Logistics (1785), SixSigma (655), project management (300), China (155), … etc.

Evaluate importance of interaction

1 positive review = 1001 references = 301 positive rating = 1

Evaluate context parameters that can be extracted from the reaction

-Structured review (keywords)-Keywords of journal (optional)-Keywords assigned to papers referring to paper-Social tagging

reaction: -evaluate value review/refer/rate-contextualize structured/

references/tag

EX POSTERIORI based on reactions

-Context-Quality

Challenge

I. Reputation? What is your reputation? Which elements of a reputation can you derive from interactions with contributions?

II. PhD community: what do we want to do with it?

Discussion

-Research question-A sustainable campus platform

2 ideas

Idea 1: Research questions

-Every PhD needs to formulate several research questions-That can be difficult-Some feedback could be helpful-Seeing who have similar research questions could also be interesting

Research question:

Which interactions of users with online content relate to quality?

Keywords: quality, user-generated content, Web 2.0, rating, popularity… add one yourself!

>> 7 related questions

NEEDS REFLECTION!

>> 3 comments >> 2 links/attachments

Share on Twitter

Email question

Add a comment, link, or attachment

Stay updated I like this

So, I filled in my research question, now what?-Attach a document, link, or comment-Assign label “Needs reflection”, “Needs case”, “Needs supervision/collaborators”, “Needs literature”-Others can react, give tips, link to interesting resources, suggest people, like, offer help, etc.-Be notified of others who have similar interests

-People can add simple tips and suggestions to make this campus more sustainable-Possibly including references to literature-Leads to discussion, voting, and possibly solutions-Could lead to research

Idea 2: A sustainable campus platform

Simply sustainable tip : Use tap water instead of bottled waterKeywords: water, consumptionWho can do it? EveryoneEffort: minimumHow many people do it? 29Research: 4 paper, 3 reports

>> 7 related tips

NEW!

>> 3 comments >> 2 links/attachments

Add a solution, comment, link, or attachment

Stay updated

I like this

Share on Twitter

Email tip

Summaryblablabla./… water and (Spieckse, 2007), etc.

Some things to take into account

Kollock

motivation ~ need

Need or necessity is central Heterogeneity is important (different views, people)

Bouwman et al.

• We argue that social software systems should trigger mechanisms that allow us to associate with or form social groups, whether online or in the real world.

• Such mechanisms would acknowledge human motivations, like eagerness for exploration, curiosity, inquisitiveness, civilization, valuation of belonging, achieving self-realization, enjoying one-self.

Psychological needs Approach

Recognition/reputation: audience, visibility

Critical mass/marketing: events, regular interesting news items, clear message

Leadership

Reputatie/status: “winners” are visible, professional relevance

Reciprociteit, gift economy, feedback

Quick response on contributions

Trust, familiariteit, sense of community

Coping with new people, new members

Privacy: restricted access

Events & meetings, online & offline, find a rithm

Clear objectives and role definitions, profiling.Accountability & identificatie: real names.

Identificatie with peers: every role has to be represented

Learning new things Heterogeniteit: different roles and backgrounds

Relation management (inviting people), making connections, relevant content

Support by experts

Netwerking, be more effective Communicate about value of participating

Self-efficacy Statistics: show contributions, popular resources, recognition and praise

Self management Integrate roles, authority, and responsibility

Self organization Group people and contentIntegrate communication and functionality outside community

Self regulation and categorization Order people and content (tagging, taxonomy)