Cognitive Computing - Part 2
Applying cognitive computing to KM

Lessons learned analytics

Among other benefits, an effective lessons learned program can help an organization:

• avoid redundancy and reinvention by reusing existing designs and building on past experiences;
• improve the quality of products and services while reducing errors, rework and cycle times; and
• standardize best practices to improve efficiency and reduce operating costs.

Despite their popularity and potential, current lessons learned systems regularly fail to deliver the intended results. Although many organizations are successful at capturing lessons, those same organizations often struggle with learning and reuse.

Now, imagine that you are a project manager in an organization with a search system that mines existing lessons learned databases to help find the most relevant lessons for your project. There is nothing particularly radical about that—most search engines provide those capabilities today.

A cognitive computing system could do more. Programmed correctly, it could analyze the lessons learned databases as well as other project logs not specifically called “lessons learned” for patterns or trends in the data that could point to new opportunities for improvement in the execution of the project.

Further, there could be a set of programs that sort those lessons and alerts by the stage and type of project. A team nearing a milestone could alert other teams approaching that same milestone (but perhaps a few weeks or months out) to changing conditions or lessons learned such as pending budget cuts, FDA or regulation changes or process glitches. Then, the alerted teams could proactively adjust to the changes or new situation.

Such capabilities have the potential to reduce the administrative burden of lesson capture, increase the rate of application and reuse and, by extension, improve organizations’ capacity for incremental collective learning.

Data visualization

Human beings typically visualize data for three reasons: to explain, explore or exhibit it. Basic data visualization approaches, such as the conversion of data tables into charts and graphs, have helped organizations digest and disseminate information for centuries. In addition to surfacing trends and patterns, visual data representations facilitate effective learning since a picture tends to be more compelling than pages of words.

With the support of more advanced technology, including cognitive computing, data visualizations have even greater potential for knowledge discovery and transfer. New visualization capabilities can help organizations codify complex knowledge and discover new relationships. Potential applications include:

• harvesting large amounts of data and information into concise visual representations to expedite the dissemination of knowledge;
• exploring data from disparate sources to identify patterns and other relationships that normally cannot be detected via flat analysis;
• visually mapping search results to help users navigate to adjacent resources (e.g., a search connectivity map that shows the interconnections between topics);
• extracting themes (e.g., common topics, sentiment analysis) from survey responses, help desk tickets, social networking threads and other social feedback mechanisms;
• developing predictive models and visualizing “what if” scenarios; and
• mapping the direct and indirect effect of changes in procedures or directives.

One benefit of cognitive systems is that they can apply sophisticated algorithms to structured and unstructured data to assist in managing, organizing and annotating large data sets. The Advanced Working Group believes that advancements in document and graph database systems will allow organizations to utilize cognitive computing more effectively while facilitating the discovery of previously unrecognized trends and connections among topics and phenomena.

The digital sidekick

Most knowledge workers would create more value for their organizations if they could delegate routine administrative and information gathering duties to others. However, it’s too costly and inefficient for firms to hire human assistants for everyone.

That’s where the concept of a digital sidekick comes in. With the introduction of the next generation of intelligent personal assistants, organizations may finally be able to minimize the time employees spend on mundane tasks, thus boosting their productivity on core work activities.

Intelligent personal assistants are characterized by an ability to interact verbally without a keyboard, to use natural language processing to interpret the meaning of requests and to be extended to new data sets and scenarios. Most people are already familiar with that concept from their personal lives, having encountered popular voice-activated digital assistants such as Apple’s Siri, Microsoft’s Cortana, and Google Now. More recently, products such as Google Home and Amazon Echo have extended the reach of digital assistants to cover an entire home, allowing anyone in a space to ask simple questions and relay verbal commands related to home-based electronics and systems.

More sophisticated versions of those technologies have the potential to transform how employees access and interact with enterprise knowledge. When used in the workplace, digital assistants could provide tailored content summaries, scheduled digests with deeper drill-downs where useful and, ultimately, insights and suggestions based on patterns in the user’s behavior and work processes. If turbo-charged by the capabilities of cognitive computing, a digital program could serve as a mentor, guide or sidekick to an employee both in the office and on the go.

Next steps

The potential for cognitive computing to transform foundational KM processes such as those for content curation, knowledge transfer, search and discovery and expertise location is both exciting and daunting. It is important to understand what’s possible, but also to remain realistic about which tasks can be automated and which will still require human intervention and oversight. The last installment of this series will describe limitations and challenges associated with applying cognitive technology to KM, lessons from previous technology shifts and predictions for how the cognitive computing market will evolve over the coming years.