Ways of Seeing and Being Seen: People in the Algorithmic Knowledge Base

Machine learning enables knowledge management and knowledge bases to move beyond putting things in the right folders to the (re-)shaping and surfacing of knowledge, not only from published content but also from the ongoing work that people do. With this, a knowledge base is no longer just a repository but becomes an active part of people’s work. In this position paper, we discuss how knowledge bases can scaffold knowledge as a process rather than as knowledge items. We propose using interaction patterns to reframe the focus of knowledge base design to focus on active knowing. In particular, we discuss the challenges of associating specific individuals with knowledge or content.

This position paper was accepted for inclusion at the workshop on CSCW and Algorithmic Systems (opens in new tab) at ECSCW 2022.