C hatti ng w i th I n C hat C hatbots f or Resear ch - PowerPoint PPT Presentation

C hatti ng w i th I n C hat C hatbots f or Resear ch Purposes

I n C hat H ands O n C ase studi es D esi gni ng di al ogues Rapi d testi ng w i th hum an chatbot depl oym ent Refl ect & shar e pr oposed di al ogue system s G r oup di scussi on

Ethnobot (ethnogr aphi c chatbot) Ethnogr aphy Ethnos – ‘ f ol k, peopl e, or nati on’ G r ahpo - ‘ I w ri te’ C hatbot “a com puter pr ogr am desi gned to si m ul ate conversati on w i th hum an users”

D um b A I , sm art desi gn

Royal H i ghl and S how RHS bot in Action Participants Chat Log Participants Selfie

E di nbur gh C i ty C ounci l Publ i c A rt Pr oj ect Stakeholder in Conversation with Ethnobot

E di nbur gh Li vi ng Labs: Park Li f e Pr oj ect Rapid Human-Robot Testing ELL Alfred the Owl Deployment

Benefi ts of ethnobot m ethods ● S cal abl e, l ow cost engagem ent ● Ri ch m ul ti m edi a narr ati ve ● Fri cti onl ess, understandabl e chat app f orm at ● D ecr eased observer eff ect ● Real ti m e & context ri ch ● G ener ati ve & open ended ● Ethi cal r esponsi bi l ty tow ar ds user engagem ent & perspecti ves

[ Ethnobot V i deo D em onstrati on ] Pl ease get i n touch i f you’ d l i ke to see the dem o vi deo. hector@ i nchat. desi gn / rory@ i nchat. desi gn

C hat entri es ar e saved w i th ti m estam ps, al ong w i th m edi a, and geol ocati on data

Later w e vi ew thi s data usi ng a r eader app…

H ands on Adventur e ti m e

C hal l enge I deati on  C an you thi nk of servi ces, pr oducts, behavi ours you w oul d l i ke to di srupt, i nnovate or unpack?  W ho ar e ther e users you w oul d l i ke to pr obe?  D efi ne r esear ch questi ons, goal s, i nterventi ons  Thi nk thr ough context, i denti fy dri vers, r esour ces, barri ers  U nderstand the outputs needed, di al ogue m ust l ead ther e

D esi gni ng di al ogues & i nter acti ons • D eci de chal l enge • Thi nk thr ough context, r esour ces, barri ers, dri vers • Pr ototype di al ogue  I ntr o  M ai n C onversati on  O utr o • Rapi d hum an-bot testi ng • I ter ate on di al ogue, noti ng i m pr ovem ents • Refl exi ve gr oup di scussi on

C ontr ol l i ng fl ow & i nput  Fr ee text i nput  Buttons  N um eri c i nput  O pti onal vs m andated pi ctur e

Rapi d hum an-bot testi ng

D esi gni ng di al ogues & i nter acti ons • D eci de chal l enge • Thi nk thr ough context, r esour ces, barri ers, dri vers • Pr ototype di al ogue  I ntr o  M ai n C onversati on  O utr o • Rapi d hum an-bot testi ng • I ter ate on di al ogue, noti ng i m pr ovem ents • Refl exi ve gr oup di scussi on

C hatti ng w i th I n C hat C hatbots f or Resear ch Purposes http://inchat.design C hat to Event Bot about your experi ence today: https://bots.inchat.design/eventbot/