Oct 26, 2015 | Zack Blois
Natural Language Generation: 4 Ways NLG Drives Actionable Business Intelligence
Is your company building a “partnership” between your big data, business intelligence (BI), and artificial intelligence initiatives? If you answered “yes”, you are not alone according to our study, “State of Artificial Intelligence and Big Data in the Enterprise”.
58 percent of businesses are attempting to wrangle this combination of analytics and smart machines into actionable information (and that number is only set to grow).
Though many businesses are attempting to translate the power of data into actionable insights, lots are falling short, spending too much time manually interpreting the output of analysis and communicating results to key stakeholders. We believe this gap in process is filled by Natural Language Generation (NLG) technology, a subset of AI that transforms data into written narrative.
The following four reasons will tell you why NLG will help your company take your BI tools and analysis to the next level:
1) Combining NLG and Big Data to Scale Success
Companies across the globe are rapidly adopting data visualization tools, BI platforms, and analytic apps, but these tools are often siloed within the organizational infrastructure. Concurrently, the value of these same insights is directly tied to the recipient’s data interpretation skills.
Or as Dan Woods of Forbes puts it, “If your idea of big data is that you have a data scientist doing some sort of analysis and then presenting it through a dashboard, you are thinking far too small. The fact of the matter is that big data really can’t be understood without machine learning and advanced statistical algorithms.”
NLG provides the systematic learning and statistical algorithms needed to bridge the gap between data insights and employee understanding. It does so by using machine knowledge to turn ideas into actual sentences – turning your structured data into words with incredible scale, accuracy, and efficiency.
2) Real-time Written Communication = Better Business Decisions
In-depth, written performance reports are great, but much like a brand-new car, the value continuously depreciates the moment the ink dries on the paper. This is because the data is no longer current, which means your business decisions are being driven by dated information.
Take logistics reporting for example: a report may only be a day or two old, but within that time period, a snowstorm closed a major interstate in the Northeast US, causing your chain to be severely rerouted. Your near-term business decisions are drastically altered and your previous report is immediately obsolete. Even if it is updated in real-time, these types of reports still require interpretation, increasing the time it takes from analysis to action.
With NLG, the report becomes easy to understand and act-on, and it updates as the underlying data changes, ensuring it is up-to-date and accurate (much like how the Waze app provides up-to-date driving data to every-day commuters). This type of immediate, easy-to-consume, analysis allows your business to make the right decisions faster.
3) More Time to Complete Important Activities and Initiatives
One big knock on artificial intelligence systems are that people believe they will eliminate jobs. But, what if these advanced tools instead free up humans to do more important and unforeseen things within the business?
For perspective, per ZDNet, American farm worker numbers decreased from 10 million in 1950 to 3 million in 2010. But here’s the catch: during that same 60-year period, the modern technology industry was essentially created from scratch and increased by roughly 6.5 million workers.
Just as improvements in farming technology automated rote tasks, current improvements in analytics and AI are automating mundane data analysis and reporting processes in the transition to the innovation economy.
What does this mean for your workforce? NLG will provide them the bandwidth needed to tackle important initiatives like growth hacking or spending time on effort on high-value tasks.
4) Increased Internal and Client-facing Transparency
Let’s face it: graphs, charts, and data sets are not a universal language--even if many of us wish they were. These types of visualizations leave the data (and outcomes) open to interpretation.
However, written word and narratives are universal. Sentences, bullet-points, and paragraphs help break down interpretation barriers, analysis paralysis, and basic confusion. We call this transition from data visualization to transparent, normalized natural language insights “Data Storytelling”.
Take portfolio reviews for example. Historically, these financial reviews are very dense and leave many customers unsure what the litany of graphs and charts actually mean. However, as our CEO Stuart Frankel stated in the Harvard Business Review, “AI systems provide transparency by communicating what has happened and offering recommendations in plain English that people can immediately understand.”
See it for yourself: Believe it or not, this transparent and easy-to-understand financial report linked was not created by a human. It was instead written in seconds by a data storytelling system.