I am an experienced Conversational Artificial Intelligence (CAI) specialist, working in the Information Technology sector. I have extensive experience in analysis, architecture, design, and implementation of Natural Language Interaction (NLI) applications, a form of Artificial Intelligence (AI) that allows people to converse with applications and electronic devices in free-format, natural language.
I completed my PhD in Artificial Intelligence in Education (AIED) in 2008, with particular focus on the use of conversational agents in e-learning. During 2010 and 2011 I studied with the OU Institute of Educational Technology and completed a Postgraduate Certificate in Online and Distance Education.
I am active in the UK AI and international CAI communities, and am particularly interested in the broadening of CAI and educational technology research from academia to real-world and commercial platforms.
Supervisors: Susan Bull (PhD supervisor)
I completed my PhD in Artificial Intelligence in Education (AIED) in 2008, with particular focus on the use of conversational agents in e-learning. During 2010 and 2011 I studied with the OU Institute of Educational Technology and completed a Postgraduate Certificate in Online and Distance Education.
I am active in the UK AI and international CAI communities, and am particularly interested in the broadening of CAI and educational technology research from academia to real-world and commercial platforms.
Supervisors: Susan Bull (PhD supervisor)
less
InterestsView All (13)
Uploads
Papers by Alice Kerly
The thesis makes contributions to three main areas.
Effects of Inspectable and Negotiated Learner Models on Self-Assessment.
Inspectable learner models can lead to significant improvements in self-assessment accuracy for children aged 10 or more. For adult and child learners, negotiating the learner model is shown to lead to significantly greater improvements than those attained by inspection alone.
Use of Inspectable and Negotiated Learner Models by Primary School Age Children.
Young children aged 8-9 make some improvements to self-assessment accuracy using an inspectable learner model; children aged 10-11 make significant improvements. Children aged 8-11 using negotiated learner models make highly significant improvements in self-assessment accuracy.
Use of a Conversational Agent for Negotiating the Learner Model.
Adult and child users respond positively to a conversational agent for negotiation. Learners actively use it to negotiate their learner model. A conversational agent can provide the facility to negotiate the learner model contents.
The thesis makes contributions to three main areas.
Effects of Inspectable and Negotiated Learner Models on Self-Assessment.
Inspectable learner models can lead to significant improvements in self-assessment accuracy for children aged 10 or more. For adult and child learners, negotiating the learner model is shown to lead to significantly greater improvements than those attained by inspection alone.
Use of Inspectable and Negotiated Learner Models by Primary School Age Children.
Young children aged 8-9 make some improvements to self-assessment accuracy using an inspectable learner model; children aged 10-11 make significant improvements. Children aged 8-11 using negotiated learner models make highly significant improvements in self-assessment accuracy.
Use of a Conversational Agent for Negotiating the Learner Model.
Adult and child users respond positively to a conversational agent for negotiation. Learners actively use it to negotiate their learner model. A conversational agent can provide the facility to negotiate the learner model contents.