AI, the “human data” interface

Artificial intelligence as an assistant to data AI, the “human data” interface

Artificial intelligence as a technology has been somewhat abstract so far and is still in the background. Software maker Tableau wants to use it as an intermediary between the user and the data collector and uses algorithms for “Ask Data” and “Explain Data” for this.

Companies in this topic

Artificial intelligence can help with data.

(Photo: pinkeyes –

AI originally comes from the world of IT and data, and people are known to not. From this point of view, it is clear that AI is used as a translator between the user and data assessment tools. Tableau now wants to use AI, specifically Ask Data and Explain Data, to help employees ask questions about data “in a very natural way, just by typing it in.” As a result – at least that is the claim – they get a visual answer. Even if they have no idea about programming, says Henrik Jorgensen, DACH regional director at Tableau Software.

create context

Ask Data uses algorithms to automatically locate and index data sources. For example, the algorithm knows that a query in sales data for “US shoes” requires filters such as “shoes” or “USA” to be applied. “This is done by combining statistical knowledge about the source of the data with contextual data about real facts,” explains the country manager. “Shoes” is a common value in the “Product category” field and “America” is a synonym for the word “United States.” This inherent support for synonyms allows users to gain insights while using different expressions for the same field, such as ‘sales’ or ‘booking’.

See also  Lufthansa was concerned about travel restrictions in China

Find the pattern

Explanation of Data speeds up the analysis process and helps users quickly discover factors affecting data. This is possible without preparation, data preparation, and data modeling, Jorgensen says. Until now, users have had to manually search for possible explanations to confirm influencing factors or explain data points and outliers. With annotation data, on the other hand, users simply choose a data point in the visualization and Bayesian algorithms automatically evaluate hundreds of possible patterns and interpretations. The Data Explanation module then provides the most relevant explanations with the highest statistical significance as an interactive visualization.

Sales example

In practice, this can be used as follows: the sales team can use predictive modeling to identify the most profitable selling opportunities. The algorithm can predict the probabilities of a purchase even though it does not possess the critical (intuitive) knowledge of an account manager who knows the business relationship through personal customer service. Or vice versa: if an account manager can also use AI-assisted data science, then they can run different scenarios and see which options are likely to go down best with their customers in doubt.

Data science for all users

Difficult tasks such as resource allocation, prioritization, staff deployment, and logistics must be able to be solved by more staff. “This is very valuable, for example in a data-driven approach to marketing and sales with opportunity capturing, time-to-closing forecasts, and many other CRM-related use cases that most data science teams cannot prioritize,” says Jorgensen.

channel table

In Tableau, the channel consists of approximately 1,200 Tableau distributor partners. Tableau service partners implement customized solutions for clients on the Tableau platform. Technology partners have tools to collect, store, embed, and connect critical data for a variety of business areas.

See also  Collegiate football predictions: Florida hanging in six New Year's despite losing, North Carolina joins the field

(ID: 47494639)

Leave a Reply

Your email address will not be published.