Chatting About Chatbots: A review of the Chatbot Workshop

By: Claudia Mokbel and Kathleen Carson, Statistics Canada

The Data Science Network (DSN) was established more than a year ago to create a dynamic space for data scientists and enthusiasts alike to collaborate in the field of all things data science. The quote "Alone we're smart, but together we're brilliant." by educator Steven Anderson exactly describes the foundation of the DSN as it recognizes the need for collaboration within the data science community to push beyond existing limits.

With this goal in mind, the DSN organized the Chatbot Workshop and invited participants to discuss all angles of chatbot implementation, even those who were just thinking about implementing one but didn't know where to start. The workshop helped to break down the process of chatbot implementation and allowed participants to form contacts for when they're ready to begin their chatbot journey.

"Data science is a team sport and I like to think of the Data Science Network as the stadium for people to share their experiences creating business tools that solve your organization's concrete problems," said Eric Rancourt, Director General of the Modern Statistical Methods and Data Science Branch at Statistics Canada. "As trusted data science leaders, Statistics Canada is proud to facilitate these events in order to leverage our collective knowledge on emerging and innovative data science initiatives."

"This is one of the primary objectives of the Data Science Network – to create a dynamic space that allows us to network and collaborate amongst peers." added Sevgui Erman, Director of the Data Science Division at Statistics Canada. "So that together we can learn and develop knowledge and skills that help us grow as a community. The Chatbot Workshop was an excellent way to continue building toward this objective."

In case you missed it…

The Chatbot Workshop was held on December 6, 2021, and featured various projects surrounding advancements in artificial intelligence (AI) technology to help users interact with web services or applications. A total of 255 people attended the event and brought participants from the private sector, public sector and academia with some attendees tuning in from as far as Europe.

The workshop began with a welcome from Sevgui Erman and Eric Rancourt. The event's Master of Ceremony was Jeremy Solomon, Chief of the Data Science Network and Strategy at Statistics Canada, who is also responsible for the Data Science Network's Bits and Bytes monthly newsletter.
You can find the chatbot presentations from all six different federal departments in both official languages on the GCWiki Chatbot Workshop Presentations page.

For the first presentation, John Nephin from Shared Services Canada (SSC), provided a demonstration of the SSC Azure chatbot services and its architecture. He also described the framework and the services related to bot logic, user experience, cognition, data ingestion, and more.

The second presentation was given by Dr. Charith Gunasekara, PhD, P. Eng, from the Department of National Defense. The focus of this presentation was on building a question-answering AI chatbot using open-source technologies.

Mary Griffin and Fraydon Karimi from Innovation, Science and Economic Development Canada showcased their department's Virtual Assistant technology. They explained the background and purpose of it's implementation and also provided an overview of their chatbot's architecture.

After a short break, the fourth presentation by Alexandre Istrate from Statistics Canada's Data Science Division, demonstrated the progress being made on a robust chatbot solution they're in the process of developing in time for the 2024 Census Behavioural Test, which will also help Statistics Canada prepare for the 2026 Census.

Following this, Dr. Alexei Lapouchnian, manager of Natural Resources Canada's Digital Accelerator, illustrated the introduction of AI systems into enterprise ideas around intelligent assistants. This project was based on the research conducted at the University of Toronto in partnership with a large technology company. The project aimed to discover how to inject AI enterprise business processes (BPs) and augment decisions made as part of BPs with AI. The presentation describes the current problems that the adoption of AI in enterprises currently faces and provides possible solutions.

The final presentation was given by Gias Uddin from the Office of the Superintendent of Financial Institutions. The topic focused on the design of natural language processing toolkits in the industry and research domains, including lessons learned and takeaway messages.

Moving forward

The DSN has grown over the past year to become a central hub for data science practitioners, managers and anyone interested in the field – whether they're from the private sector, public sector, academia or beyond. The plan is to continue build the network based on the needs of its members. Check out The Data Science Network Turns One article for more information.

With the good feedback received about the Chatbot Workshop, the DSN is already working on bring you more events like this one. Currently, the team is working to put together a workshop on Data Hubs that is expected to be held in early March. If you would like to present at this workshop, please contact us: statcan.dsnfps-rsdfpf.statcan@statcan.gc.ca.

As a trusted data science leader, Statistics Canada will continue to contribute to the development of a strong data science ecosystem, through events and workshops to connect with other data science practitioners and managers; the DSN will continue to keep you informed on future collaborations, projects and workshops.