Kingland Blog

Using Text Analytics to Uncover Insights within Three Common Use Cases

Written by Matt Good | 6/13/19 9:30 PM

Every day you have 8 to 10 hours (on average) to complete work. Most of the professionals we work with spend more than two of those hours a day reading content such as news stories, regulatory documents, annual reports, internal business documents, and more. They set up Google Alerts, scan news headlines, and skim articles for any mention of an event that could trigger a new opportunity or cause risk to their organization. 

All of this takes time. And when human comprehension and retention levels hover around 50 to 60 percent, we have to question how accurate we can be when quickly and accurately synthesizing information from various sources in order to make complex business decisions. 

How can we automate some of these processes to save time and valuable resources and improve the comprehension of unstructured text?

Let's start by discussing three primary scenarios where we see corporate actions and other critical news and regulatory events impacting clients. 

Risk Monitoring

Each year, hundreds of thousands of corporate actions and other regulatory events are announced by domestic and international entities that may need to be acted upon within enterprise processes. The secret to successfully acting upon them involves seeing the clues or patterns that reveal themselves over and over again.

Monitoring news, filings, market information and more is a tedious and necessary part of risk monitoring. You value knowing when a corporate action or regulatory event takes place that suggests a potential change in the risk posture of an entity-of-interest. To be able to act on these events, businesses have enterprise processes that require information about the entities, people and products they care about. 

As these events take place, they're captured by documents from multiple sources and in varying degrees of legalese. Risk analysts are reading hundreds of documents related to the entities they care about. Organizations constantly desire to sift through the noise of unstructured data and improve efficiency by examining contextually relevant information from various sources. 

Data Management

Corporate actions and regulatory events challenge data management professionals to create a pristine "golden source" for key operational, client and compliance data programs. The data from these events becomes obsolete, outdated and incorrect as soon as another event occurs. 

Similar to the risk monitoring scenario above, a data management professional, in a given year, will scan through thousands of documents and news sources that describe corporate actions and regulatory events. Few of the documents are actually actionable, but many leading companies are willing to live with that inefficiency when hundreds of downstream enterprise processes are impacted by a missed action or event.

Client and Market Opportunities

Sales opportunities are everywhere, but competition is fierce. Teams want to identify a group, either entities or people, of clients or prospective clients. They need to understand if the entity or person is likely to be receptive to a new product or service. Events such as the entity launching a new product, making a positive financial announcement or selling off a division may be relevant for such opportunities depending on the source and use case.

These sales professionals and their supporting teams are searching through a variety of news articles with any mention of their prospective client. Many of these articles fail to produce a new opportunity, but the competition is doing the same thing, and speed is of utmost importance. The first to offer a new opportunistic product or service is likely the first to close the deal. 

Kingland Text Analytics Suite

Regardless of the use case, it takes time to accurately uncover insight. Synthesizing information from hundreds (or even thousands) of sources and documents is time consuming, challenging and consistently inefficient. 

AI techniques such as Natural Language Processing (NLP) and Named Entity Recognition (NER) can provide cost reduction and an increase in the speed and accuracy of extracting information from trusted sources. Text Analytics can accelerate existing enterprise processes by automating the scanning and summarization of relevant stories or regulatory filings based on trained "reading" models, or extract data from such content for direct integration with your enterprise data processes.

At Kingland, our clients accurately discover and extract value from their unstructured data. The Kingland Text Analytics Suite continuously monitors, identifies and prioritizes information to enhance the tasks associated with risk, compliance and operations experts. By accurately extracting unstructured data from PDFs, web pages, legal documents, prospectuses and more, clients can act on the entities, people, relationships and many other critical data attributes to meet changing risk, regulatory and operational requirements. 

We’ve measured its effectiveness, improving manual process efficiency with real-time updates and alerts and an accuracy rating exceeding 90 percent - even up to 99 percent in narrowly focused, highly tuned data extraction use cases.

How much time do you want to save by improving your manual enterprise processes? Ask how we can help.