Building an AI-powered semantic search for historical records

The Royal Canadian Numismatic Association (RCNA) is a non-profit educational organization, founded in 1950 and incorporated in 1963. It provides a national platform for coin collectors to connect, share knowledge, and celebrate the rich historical and cultural significance encapsulated in coins. Coin collecting provides a unique lens through which to explore the evolution of societies, art, and economics. Since 1956, the RCNA has published the Canadian Numismatic Journal (CNJ), which has published seminal papers related to Canadian numismatics. Unfortunately, access to them is restricted to members due to reproduction rights, and limited to lexical search, which requires searchers not only to identify relevant keywords, but to correlate them with more than 60 years of linguistic changes. Finding detailed information within the CNJ archives is time-consuming and difficult.

In this project, a Mitacs Intern at Mohawk College will create a publicly accessible, more intuitive interface to RCNA’s knowledge base integrating Natural Language Processing (NLP) and sentiment analysis (SAn) to refine search accuracy, ensuring users receive the most pertinent information. The intern engaged in the project will apply and further develop their technical skills in programming, machine learning, and language processing within a business context.

The immediate impact of this project is expected to be a repositioning of the RCNA in the Canadian Numismatics digital landscape. The website will be positioned to become a destination for users of many skill levels to engage in the hobby, simply by leveraging the information it already holds and making it available appropriately.

Faculty Supervisor:

Stephen Adams

Student:

Partner:

Royal Canadian Numismatics Association

Discipline:

Computer science

Sector:

Other services (except public administration)

University:

Mohawk College of Applied Arts and Technology

Program: