Adaptive Learning and Personalization Within the Enterprise
Enterprise users of KM systems are at a particularly auspicious moment. With the relatively recent advancements in language models, intelligent agents infused with such models, and what’s popularly termed “AI,” the potential to create personalized user experiences supplemented by adaptive learning has never been greater.
The magnitude of this fact cannot be overstated. The capacity to tailor what information is disseminated to someone based on individual preferences, as well as to surface relevant information based on personal intentions, has untold possibilities for boosting productivity, efficiency, and organizational effectiveness.
Realizing this promise, however, requires an artful implementation of numerous technologies, including knowledge graphs, vector databases, document repositories, content management systems, and access controls. Regardless of which technological approach an organization adopts, success ultimately lies in mastering an indispensable element of personalization which, on its own, has little to do with contemporary cognitive computing models—it’s KM’s emphasis on people.
According to Franz CEO Jans Aasman, the more pressing matter is, “Do you understand the user, and why that person really needs this content, especially if you have a massive overflow of information? And, how do you represent what a user wants?” Such questions may well provide the impetus for adaptive learning and personalization, while also illustrating the basic requirements for these objectives at enterprise scale. Organizations that discern how and when to personalize their information retrieval, as well as what the finer points of doing so are for different users and goals, will ultimately reap the benefits of adaptive learning.
Such questions may well provide the impetus for adaptive learning and personalization, while also illustrating the basic requirements for these objectives at enterprise scale. Organizations that discern how and when to personalize their information retrieval, as well as what the finer points of doing so are for different users and goals, will ultimately reap the benefits of adaptive learning.
Contextual Knowledge
The overall worth of language models, particularly when deployed to complete autonomous or semi-autonomous actions from agents, can refine how content is created, delivered, and acted upon for individuals. Anyone can have a model change the diction or tone of voice for their applications. According to Tony Grout, M-Files’ chief product officer, “People see AI as being able to surface personalization quickly. We can all generate slide decks. We can all learn quickly with content that’s being created for us through both basic chat experiences and agentic experiences.”
What’s more difficult for these constructs is to identify a user’s characteristics that are most influential for curating their content needs, then to personalize outputs based upon them. According to Aasman, employing these tools and their supporting infrastructure is “only 50% of the story. The other half is living in the heads of the users and the developers.
This knowledge of what easily spans any number of aspects, including what content someone wants, why, how it will be used, what the optimal means of expressing it is, and how those factors change for a given application at a particular time, comprises the context required for personalization. Granted, basic keyword mechanisms can provide some of this information, particularly that pertaining to business units or specific applications. However, these and more modern information retrieval methods “can’t necessarily know without more context that, for example, I’m a kinesthetic learner and need to learn by doing something, not reading about it or watching a video,” Grout commented. “I’ve got to be able to tell the system that I’m a kinesthetic learner.”