expert.ai Targets White Whale of Unstructured Data
February 7, 2021
out of 10 chief data officers (CDOs) agree that management of unstructured
language data – including text from business documents and emails – must be
addressed in the next 12 months. Yet, few are prepared with the natural language
processing (NLP) and natural language understanding (NLU) experience and tools
needed to make a transition work.
A new report published by expert.ai surveyed data and analytics decision makers
to reveal how teams are faring as they attempt to guide their companies towards
Value of unstructured data
There is growing realization across enterprises that unstructured language data
is not merely a by-product of operations but a vital resource to be mined for
actionable insights. The ability to extract value from unstructured data is what
will separate businesses from their competition via better Net Promoter Scores
and reduced manual document handling and extraction costs. NLP and NLU
technology have been proven the key to doing so.
Despite this, only 8% of data teams have completed NLP and NLU projects within
their business that would enable them to fully unlock the value of their
unstructured language data. More than a third (34%) of data teams have started
activating plans for NLP projects. Nearly a quarter (24%) are still defining
their plans but are not ready to activate them.
The report, titled “Harnessing the Power of Unstructured Data with NLP and NLU”
was prepared by The AI Journal. It was fueled by a survey of CDOs, which also
revealed the top three most popular types of NLU solutions are platform (44%),
open-source (34%) and cloud vendor offerings (34%).
AI skills in demand
Nearly all CDOs (96%) see delivering business impact through AI as their top
concern in 2022. However, while two-thirds of organizations claim to be
knowledgeable about AI, they often lack employees with the skills to build and
execute programs. Organizations have found it difficult to acquire the necessary
talent, whether through internal training or external recruitment.
Given the low supply of data skills on the market, it’s no surprise that
organizations looking to fill their skills gaps choose not to seek external
expertise as a first option. With that said, the timing of AI projects is an
important consideration. Companies that have made definitive AI plans but have
not yet activated them are more likely to look for external expertise (58%)
versus upskilling methods.
When data teams identify holes in their team members’ knowledge and
understanding, the most common solution is to upskill through training. This was
the primary method for every specific knowledge area including AI (51%), NLP
(41%) and NLU (35%). Few organizations felt online content or mentoring from
team leaders were viable methods for bridging the skills gap.
data is the white whale of the business world. It represents the great majority
of enterprise data but is extremely difficult to extract value from,” said Marco
Varone, founder and chief technology officer, expert.ai. “Those that can do so
effectively put themselves in a prime position to make more intelligent business
decisions and operate with greater efficiency.
“To make the most of unstructured data, AI and NLP must be priorities, added
Varone. However, historical approaches to AI and NLP no longer suffice. To
succeed, you need the right approach, the right expertise and a focus on the
right data. Natural language understanding is the answer to these broad language
The survey was conducted among 116 decision makers where data/analytics is a
large part of their role. Research took place across the USA and Europe. The
interviews were conducted online by Sapio Research in October 2021.