Why Kingland is Focused on Extracting Important Data from Unstructured Content

Posted by Matt Good on 6/5/18 7:32 AM

Where is Kingland focused within the overall Cognitive / AI / Machine Learning space? Our passion and investment are in two primary areas: 

  • Extracting data / important information out of unstructured content
  • Learning from vast amounts of data to make discoveries and decisions
Data mining, analytics and visualization

Examples of Enterprise Data Management Requests

Regulatory documents, government registries, corporate sites, news and industry trade journals, product documents, you name it – they all represent important sources that contain an incredible amount of valuable data, insight, sentiment and analysis that is ripe for processing. Clients approach Kingland because of a great opportunity for efficiency – they typically already staff teams of data / business subject matter expert analysts to manually extract data and information out of these types of unstructured sources, yet they continue to pursue higher quality, better timeliness, and more productivity from these teams when managing data from such sources.

Historical data from the enterprise data management programs within our clients offer a level of quality management and predictive capabilities, if harnessed appropriately. In one example, legal entity data that has been classified with NAICS codes can offer machine learning opportunities to build models that can verify the classification accuracy of millions of entities, including additionally predicting the NAICS classifications of newly onboarded entities. In another example, retailers benefit from recommendations learned from past consumer behavior, for the benefit of timely product and complementary offerings.

How Cognitive Personas Fit?

Having seen these examples many times over led us to create the Cognitive Computing suite in our Kingland Platform. From the Cognitive / AI / ML terminology perspective, we aligned with the Cognitive terminology due to its ability to describe and represent technology that includes AI and assists us with a variety of tasks and decisions.

The Collector, Scholar, Inventor and CEO personas within our Cognitive suite represent important groupings of functionality within our Cognitive architecture and allow us to humanize the capabilities to a level of better understanding by our clients. The AI terminology will still show up in our product marketing, as a subset of Cognitive (and to harness the general marketing “buzz” around this terminology), but our Cognitive personas lie at the heart of our capabilities and are what our internal development teams talk about and work on every day.

Our clients often will hear us talk about how we are “teaching our software to read.”  To expand upon this, let’s apply our Cognitive personas to the business problem of extracting data and information from unstructured content:

  • The Collector will collect all unstructured content and ready it for processing.
  • The Scholar will read through the content and will “take notes” by tagging, annotating, indexing, and otherwise organizing the content for additional downstream processing. Natural language processing and named entity recognition is executed as part of the Scholar’s reading process.
  • The Inventor will surface interesting discoveries from read content, including things like sentiment and topic summarization. Additionally, the Inventor persona has deeper involvement in the opportunity of leveraging historical data for predictive benefit, using unsupervised learning to surface trends that can be further analyzed and decided upon.
  • The CEO makes the big decisions, leveraging output from the Scholar and Inventor to ultimately make the decisions on data to extract, or predictions to act upon. The true power of Cognitive comes in to play here, through surfacing these decisions, predictions, and all supportive material for providing ultimate assistance to analyst users.

We put our Cognitive personas to work for the benefit of scaling SME analyst user capabilities by removing the drudgery of poring over an unending amount of content for only the tidbits they need. And while we work in software and not robotics hardware, this cult classic clip from the movie Short Circuit has always inspired my vision for our Cognitive personas. Think of them as reading through collected content at blazing speed, ready to assist us on the extracted data and decisions from the content. Like the Short Circuit robot, the personas will quickly demand “MORE INPUT!”

Stay tuned to additional blogs covering our Cognitive personas, as I’ll take a deeper dive into each one!

Topics: Cognitive, Text Analytics, Enterprise Data

Subscribe to Email Updates

Recent Posts