How can chatbots help our students?

by | 4 Apr, 2019 | 0 comments

Chatbots, or conversational agents, are an emerging technology that might be a game-changer for higher education. Read on to find out more about our project to build a chatbot for first-year accounting students.

As a subject coordinator, I’m often inundated with emails from students about information contained elsewhere on our LMS. Do I need to attend lectures? Are lectures recorded? When is the assignment due? Where do I find information on how to reference my work? The time spent by academics and teaching staff responding to these queries could be better spent on helping students with learning, especially in large subjects of 400+ students.

Our project

My team (Dr James Wakefield, Dr Nelson Ma and Atieh Fallahi) and I applied for a First and Further Year Experience Grant to create a conversational agent – a chatbot – to assist coordinators in first year accounting by automatically handle these administrative queries. The goal is to support students in transitioning to university and its complex systems of learning and education, provide support to these students when they need it and relieve some of their administrative workload. Our initial scoping indicated that building a chatbot should be relatively easy – focused around developing a set of question and answers for the system to understand – like a big automated FAQ.

Learning more about chatbots

The universe must have been looking out for our team because the LX.lab announced a new workshop. Jonathan Grudin (a former computer science academic, now a researcher at Microsoft) was coming to talk about conversational agents. The workshop was an eye-opener – and timely for our project – because we discovered a whole host of factors we need to consider when building our chatbot for first-year accounting. These include:

  • What sort of agent do we want to build? An intelligent assistant like Siri and Deakin Genie, a task-focused chatbot or a complex virtual companion?
  • Keeping the scope of the chatbot very clear and not letting it creep,
  • Do we give our chatbot a personality?
  • How do we break down complicated real-life questions into information the bot can understand?

We plan to build our bot and have it reside within Microsoft Teams. Subject-specific bots haven’t been built before at UTS, but we are excited to give it a try. It also raises questions like whether students will be able to access the bot from the Teams app, and will they even know what Microsoft Teams is? We’ve got a long way to go, but thinking about these questions before we start, will hopefully increase our chances of success and improve student engagement and support.

Feature image by Drew Beamer.

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *

 
Share This