Knowledge management and multidisciplinary collaboration at KMWorld 2022

Tacking down the exact processes of knowledge management, and if those processes should interconnect deeply with other, “more empirical” disciplines, is a heated discourse within KM communities.

In the opening keynote on the second day of sessions at KMWorld 2022, a panel led by Patrick Lambe, partner at Straits Knowledge and author of Principles of Knowledge Auditing, discussed the connections between the conference streams of knowledge management, taxonomy work, text analytics, and search, and what their collaboration ideally looks like.

To begin, speakers Susann Roth, chief of knowledge management at Asian Development Bank (ADB); Irena Zadonsky, director at Data & Analytics Architecture, Amtrak and former data governance manager for the Federal Reserve Board; and Dave Clarke, founder of Synaptica, answered Lambe’s query, “How does science play in knowledge management?”

“Knowledge management is definitely a science. We run experiments, we craft hypotheses, and we study them,” said Roth. “We need to have a scientific approach to managing the data; KM is needed because we have to make sense of the world and take action appropriately.”

Zadonksy added, “we’re heading into a world that’s more complicated and more complex. Everyone has knowledge on their particular subject matter, and all of that has to be managed and shared.”

Lambe was not so convinced, sharing that, “KM is a set of practices which are not consistent, not consistently described, and is the landscape for people from multiple disciplines, perhaps some from scientific backgrounds. As a practice, it does not have a single, coherent, scientific framework.”

These arguments transitioned into a conversation about if KM is not technically a science, how does it interact with disciplines that are? Specifically, Lambe questioned, “Does knowledge management, information management, and information science have their own, isolated territories? Or are they fated to overlap?”

“Balance is where we need to be,” Clarke explained. “Everyone competes for resources—but if a human put all of their effort into acquiring water, they would starve and die. If they put all their effort into food, they’d thirst and die. All of these information, data, and knowledge groups are needed in balance.”

“Even if your organization depends on data science, you need KM to facilitate their work. You need the sense-making and learning aspect of knowledge management,” said Roth.

The panel concluded with investigations into what “good” collaboration looks like within an enterprise.

Clarke offered an analogy: “If one thinks about data as ingredients, then information becomes a meal, prepared in a way that is consumable for humans. If we think of knowledge as nutrition, then we’re thinking of how those meals fuel and power human thoughts and feats. You need to be in the restaurant and the kitchen, managing it all simultaneously with knowledge management.”

“It depends on what the functions of an organization are, and what they focus on,” said Zadonsky. “People become defensive and protective of their territory; the more multi-discipline your teams are, the better the outcome will be. I think it would help to expose other disciplines during the education and training process.”

Using unstructured and language-based data to invite responsible AI strategies

Christophe Aubry, global head of value creation at expert.ai, highlighted the innovative ways language assets can be transformed into data for enhanced analytics purposes.

When today, nearly 80% of enterprise data is unstructured and language-based, accessibility is extensively dampened. expert.ai’s approach helps teams turn language into data so they can make better, data-driven decisions amidst unstructured data.

Focused on innovation, agility, resilience, and designing products to support AI strategies across industries, expert.ai employs a “green glass approach” to responsible AI when implementing it in unstructured and language-based data. By building sustainable, highly performant, light language models, the approach revolves around practicality, transparency, and human-centered efforts.

“We need to have all this processing power in order to make good analyses of textual information,” said Aubry.

expert.ai allows you to solve specific business problems with dedicated AI tools, understand how AI arrives at results via knowledge apps to keep track of the original source of information, and enable scalability and data transparency, and adapting to human needs.