From the shoebox to computer-assisted conceptual mining

In the late 1960s, as part of his doctoral thesis on materialism in the work of Karl Marx, Jean‑Guy Meunier used the reliable old system of handwritten notecards filed in shoeboxes. Some 40 years later, this department of philosophy professor from the Université du Québec à Montréal (UQAM) is helping pave the way for digital humanities by innovating in the field of text mining with an automated system for computer‑assisted reading and conceptual analysis of text (CARCAT).

“It is important that we provide the new generation of students in the social sciences and humanities with a strong foundation and training, and to foster interdisciplinary collaboration in order that they can participate, both critically and constructively, in the development of the digital economy,” emphasizes Meunier.

Following his “shoebox experience,” Meunier was convinced of the need to find a more efficient way to search through a large amount of text with the use of a computer. “At the time, computer scientists did not believe in it,” he recalls. “The computer was used mostly for making calculations in accounting or physics.” Initial funding from the SSHRC allowed him to further his idea, which eventually resulted in the creation of CARCAT.

This computer system, which uses different algorithms in the reading and conceptual analysis of digitized texts, also makes it possible to process an impressive number of documents. For example, within the scope of work done by the Bouchard‑Taylor Commission, a master’s student used CARCAT to examine the concept of reasonable accommodation in three large Quebec daily newspapers. He searched a corpus of more than 3,380 documents produced by 679 different journalists over a 22‑year period. The study revealed that the concept of reasonable accommodation was used long before today, especially in contexts dealing with religion, identity, language and employment.

In addition to being popular in the university community, this prototype technology is gaining more and more momentum among professionals. Lawyers, journalists, students and researchers have a greater need for sophisticated tools that allow them to carry out detailed textual data analysis while abiding by their complex analytic methodology. Meunier welcomes this growing interest, as he sees in it a confirmation that the cutting‑edge tool he designed can be adapted to various fields and sectors.

Over the years, Jean‑Guy Meunier also mentored a number of students, many of whom have gone on to become researchers, university professors and high‑level professionals in service industries including, among others, insurance and marketing.