Data-driven reporting is quickly becoming essential for companies worldwide. As such, companies are using natural language generation (NLG) and machine learning technologies to generate actionable insights and make informed decisions.
Steve Dock, Senior Vice President and Research, Content and Publishing Technology Owner at Moody’s, offered insights into NLG and machine learning technologies during his keynote presentation to Argyle’s CIO membership at the 2018 Chief Information Officer Leadership Forum in New York on June 19. In his presentation, Dock discussed the impact of NLG for business professionals across all industries.
NLG empowers businesses to create reports based on massive amounts of structured and unstructured data. It takes machine learning and artificial intelligence (AI) technologies to new heights, as it enables business professionals to instantly produce in-depth reports and gain the insights they need to make fast, data-driven decisions.
At Moody’s, NLG is crucial because it helps the company perform comprehensive financial analysis. Moody’s also continues to explore ways to upgrade its NLG efforts to ensure it can produce extensive reports that match or exceed its clients’ expectations. “Our ultimate goal is to have machine-written content that either supplements human content or is better than human content,” Dock stated. “It is a man on the moon-type exercise … but I’m bullish on [NLG].”
Thanks to NLG, Moody’s can understand financial data across an entire industry. NLG allows Moody’s to create reports to help companies assess industry trends, and as a result, obtain insights that they can use to differentiate themselves from the competition.
There is no shortage of data available to Moody’s and other businesses, and NLG helps all companies produce reports based on small and large data sets. That way, companies can use NLG to understand why customers may choose one brand over another, how customers are engaging with a brand and much more. In addition, companies frequently use NLG to analyze business trends, but the technology may enable companies to make data-based predictions in the years to come.
“At the end of the day, there are very data-driven narratives,” Dock said. “The difference-maker is rear-view mirror versus future prediction. Most exceptional machine-based writing can tell the story behind data, but it is not so great at evolving and providing insights from data.”
Although NLG is powerful, scalability remains a chief concern for businesses. Having the ability to integrate NLG into all areas of a business sometimes is difficult. If a company integrates NLG into its operations over time, it can identify any potential issues and quickly address these problems. Then, as a company streamlines data management, it can use NLG to produce reports based on specific use cases.
“[NLG] projects tend to be hard-wired to data,” Dock noted. “Each use case is very related to the data that you’re creating … and there may be a very heavy [data] maintenance cost.”
Going forward, companies are on the lookout for ways to make NLG context-aware. As NLG innovations are introduced, businesses can take advantage of context-aware reports. By doing so, companies can use context-aware reports to drive ongoing business innovation and improvement.
“We are struggling with how we extend [NLG] and make it more scalable … and how we merge [many] models into one model,” Dock stated. “We want to make machine-written content more context-aware.”
With NLG, business professionals can review a decision from all angles and determine the best course of action based on myriad information. NLG ensures business professionals can create reports consistently and use these reports to make informed decisions, faster than ever before.
“NLG provides assembly processes that have different decision points,” Dock indicated. “It enables you to replace [assembly processes], so if I generate the same report 100 times, I ultimately get 100 derivatives.”
To optimize the value of NLG, business professionals must understand the importance of metrics, too. Using pertinent key performance indicators (KPIs) and other metrics puts a business in the best position to succeed. Because if a company uses NLG reports to analyze metrics, it can drive meaningful performance improvements across all departments. Therefore, NLG reports can help a business learn from its past performance, identify problem areas and resolve issues before they escalate.
“To do [NLG] right, you have to understand all the metrics … and know all the ways that you’re going to express yourself,” Dock said. “And you need to be able to explain a model’s performance.”