Social Attention Management in Online Community using Artificial Agents
Download as PPTX, PDF2 likes421 views
This document discusses using artificial agents to manage social attention in online communities. It proposes a social translucence design approach where agents observe community activity, detect behavioral patterns, and send recommendations to users. Basic agents could notify new users or those with incomplete profiles, while advanced agents might detect bursts of collective activity on a resource and notify the owner. The goal is to emphasize the social aspects of the community and increase participation through raising awareness of behaviors and collective activity.
![Examples of Advanced Agents (2/2) Agents related to the individual-collective interactions Agent 6: Detecting burst of collective activity to a owned resource: ď¨msg to the owner of the resource: âYour profile/message was viewed by a high number of people (12 member(s)) [..]compared to the normal audience the last x daysâAgent 7: Detecting special interest of a group compared to the collective ď¨â2 members UserA, UserB viewed your profile more frequently than the othersâAgents related to the social attentionAgent 8: Detecting change in the Attraction power/charisma of a userď¨ âyouâre loosing/gaining some audience compared to the last monthâAgent 9: Detecting if the diversity of the userâattention is low: ď¨ Msg to the user: âYou donât enough diversify your interest (focus on same people or same tags)?â](https://image.slidesharecdn.com/socialtranslucencedesign-100215041727-phpapp02/85/Social-Attention-Management-in-Online-Community-using-Artificial-Agents-7-320.jpg)
Social Attention Management in Online Community using Artificial Agents
- 1. Social Attention Management in Online Community using Artificial AgentsNicolas Maisonneuve, Research Associate Centre for Advanced Learning Technologies, INSEADContext: European research project www.atgentive.com2008
- 2. Social translucence designSocial translucence design = Design knowledge management system to emphasize the social aspect (Erickson et al. (2002). âSocial translucence: designing social infrastructures that make collective activity visibleâ)Why is emphasizing the social aspect important?Change of the userâs perception: from a repository of resources to a community of people
- 3. Emergence of social mechanisms (trust , affinity)
- 4. Increase of the participation (main problem in Knowledge Management) Social translucence designObjective: To help users to better perceive/understand what happens within the community Approach 1: Multi Agent System Goal: Detecting and regulating different {collective,indivudual} behavioural scenarios (1 agent = 1 scenario)Agent behavior: 1- Observing the communityâs activity2- Detecting a given activity pattern:- Basic pattern: (e.g. the creation of a response) - More complex pattern: (a burst of activity)3-Intervention/Recommending: a message with a recommendation is sent to the given user(s)
- 5. Process1/ CommunityActivity(reading, tagging, voting, exchangingknowledge)Activity visible via RSS2/ ExternalArtificial Agents observe the activityA1A2A34/ Raising of awarenessUsersare aware of badbehavior/hidden collective activity3/ When an agent detects a given activity pattern, it recommends an action by sending a message to the given user(s)
- 6. Examples of BasicAgents (1/2) Agents related to the userâs state Agent1: Detecting if a new user is registered: ď¨ Msgsent to the user: âLook at the presentation of the community platformâ, ď¨ Msg for the community to introduce the new userAgent 2:Detecting the degree of completeness of a userâs profileď¨ Msgâcomplete yours skill in your profileâ, âadd a pictureâAgent3: Detecting Inattentive user (low connection rate) ď¨Emailing inattentive user with an activity report to attract their attentionAgents related to individual-individual interactionsAgent4: Detecting if a user A check the user B profile: ď¨Msg to the user B: âUserAviewed your profileâ Agent5:Detecting if someone responded to User Aâs messages ď¨ Msg to the user A: âUser X responded to your messageâ
- 7. Examples of Advanced Agents (2/2) Agents related to the individual-collective interactions Agent 6: Detecting burst of collective activity to a owned resource: ď¨msg to the owner of the resource: âYour profile/message was viewed by a high number of people (12 member(s)) [..]compared to the normal audience the last x daysâAgent 7: Detecting special interest of a group compared to the collective ď¨â2 members UserA, UserB viewed your profile more frequently than the othersâAgents related to the social attentionAgent 8: Detecting change in the Attraction power/charisma of a userď¨ âyouâre loosing/gaining some audience compared to the last monthâAgent 9: Detecting if the diversity of the userâattention is low: ď¨ Msg to the user: âYou donât enough diversify your interest (focus on same people or same tags)?â
- 8. Social translucence design related to resourcesDiffusion in the community / audience of a resource - is my message well diffused in the community? - Is everybody aware about these posting?Lifecycle of a resource - âis the resource dead?â - âis there a burst of activity now or these last days?â Social aspect - who were the last readers? - who was interested by this document?Approachď¨ Resource/Tag as a memory of the social activityď¨ Showing the social activity related a resource or a Tag with the resources to emphasize its social aspect using web widgetWeb widget