I spent some time reading about Knowledge Representation (specifically about the Knowledge Representation part in Knowledge Representation and Reasoning) and specifically about scientific and/or engineering knowledge and my impression after cursory reading is that it’s a largely an unsolved problem. Not only that, but it seems like very few people are actually working on something useful in the field.
For example, I checked the proceeding of SCI-K and PlanetKR conferences and literally all the papers seem to be focusing on “toy problems”, as in not having even remotely practical scientific implications (other than all sorts of “search” and “data extraction”, but that’s not “representation”).
Views on the topic?
[link] [comments]