Sergey Nivens - Fotolia
With Tableau's Ask Data software, the BI and data visualization vendor has injected its self-service analytics platform with natural language processing to enable users to query and visualize data more easily by typing in plain text.
Tableau Ask Data was formally released on Feb. 13 as part of Tableau 2019.1, which is generally available now, the vendor said. Tableau previewed the natural language processing tool at its annual user conference in October 2018 and made many of the software's NLP functions available in a beta release then.
With Ask Data, Tableau -- one of the top self-service BI and analytics vendors -- is keeping pace with the trend toward more NLP use in enterprise analytics software, rather than innovating at the head of its group, said Krishna Roy, an analyst at 451 Research.
But the Seattle-based vendor appears to have succeeded in making NLP relatively simple for end users to work with, she said.
"NLP is becoming a tick-box item within an analytics platform, and so it's on offer -- or under development -- within most BI and analytics platforms," Roy said. "Ask Data has a Tableau take on NLP because it is genuinely easy to use, but, at the same time, sufficiently complex to be useful."
Engage with data by typing in text
Ask Data enables users to type data queries in plain language and instantly get a visual response in the Tableau platform. The goal, according to Tableau, is to make it easier for anyone, regardless of skill set or analytics expertise, to engage with data and produce analytical insights.
Krishna Royanalyst at 451 Research
"With natural language processing, we're taking that one step further by making data even easier for more people to access. I think natural language will close the gap between the people who want to and are now able to ask their questions," Francois Ajenstat, chief product officer at Tableau, said in an interview.
With the Tableau Ask Data feature, users can type colloquial questions, such as, "What were my sales this month?" And the platform will return an interactive visualization.
Users then can either type clarifying questions or use drag-and-drop gestures to further narrow results and continue to explore the data. Tableau Ask Data supports a number of synonyms and like phrases to generate insights even when searchers use different terms to represent the same field, the company said.
The Ask Data feature requires "zero setup," according to Ajenstat, and it's available at no extra charge as part of the Tableau 2019.1 release. Making NLP an integral component of the main Tableau analytics platform for free is part of what differentiates Tableau and its NLP strategy, Ajenstat said.
"Many times, this kind of technology requires significant effort by the customer to set up the data sources and set up the data model," he said. "Ask Data is available automatically, out of the box, with no setup."
The Tableau Ask Data tool also takes advantage of all of Tableau's data connectors, so users can query data without having to move it, Ajenstat said.
Another differentiator is Tableau Ask Data's cross-channel accessibility, Ajenstat asserted. Ask Data works in the Tableau Online SaaS product and in on-premises and public cloud deployments of the company's server software.
Voice queries: The next NLP frontier?
Voice-enabled queries are considered a next step for NLP technology that Tableau competitors like MicroStrategy and Qlik are starting to take advantage of. Tableau has yet to release a voice search feature, but Roy said she already sees voice as a standard capability that most BI vendors will have to provide to be relevant.
"I think voice-enabled queries will take off eventually, because -- like NLP search -- they provide an already-familiar means to query and interrogate data to get insights," Roy said. "However, I think we're a long way from reaching that goal right now."
First, voice querying technology has to become reliable, according to Roy. "I have seen voice-activated queries demoed several times by other vendors -- not Tableau -- and they haven't worked," she added.