Unifi Now Automatically Discovers Similar Datasets and Auto-Recommends Search Queries Using AI and Advanced Natural Language Processing
Unifi Software, a leader in providing a seamlessly integrated suite of self-service data tools, announced significant new advances to its data platform to make the process of cataloging data and discovering datasets even faster and easier for business users, data stewards and data analysts alike by extending its use of artificial intelligence (AI) and natural language processing (NLP). Unifi continues to expand its use of AI to obscure the technical complexity of finding, exploring and preparing data to support data analytics self-sufficiency and improve data quality across the organization with each new platform release.
Behind the hype, the portfolio of services categorized as “AI” can provide a wide variety of data and analytic optimization benefits. The challenge is in getting started now in a market where many vendors are touting future-facing functionality that may not go to market for another 6, 12, or even 18 months. Enterprises seeking to take full advantage of AI in 2018 need to work with vendors that have a clear and transparent set of AI functions right now, such as those provided by Unifi or risk losing a year of data-driven differentiation to their competitors,” said Hyoun Park, CEO and Principal Analyst, Amalgam Insights.
The more users that engage across the Unifi Data Platform, the more intuitive data insights become as the company’s AI-engine, OneMindTM, learns to predict patterns and to recommend datasets to serve up to users. The newest release includes the ability to find similar datasets through the platform’s Dataset Explorer. When displaying a dataset in this view, users can choose to display other datasets that are the same or similar simply by clicking ‘Similar Datasets.’ The result displays the percentage of similarity based upon sample stats such as comparing properties of the primary dataset. The AI-engine parses for these similarities to build the recommendation.
For data stewards or data engineers, finding similar datasets allows duplicate datasets to easily be discovered and cleansed. Often there are datasets that are generated on a recurring basis and in most instances, the latest version will provide the highest value. In other instances, governance rules may be applied to one dataset and open in another or assigned to users with varying policies for use such as to mask PII data. In this way, a data steward or data engineer can easily find those datasets, combine information, delete or archive them based on lack of use over time.
“Business analysts naturally want to explore as much information as possible for potential value and at the same time be very efficient with their time. As they build a catalog of their data, by automatically being able to select similar datasets they can quickly move from discovery to gaining insights,” said Ayush Parashar, Co-founder and Vice President of Engineering at Unifi Software. “Through our advanced AI, we mask the complexity of finding, preparing and governing data in order to get right to exploring, asking questions and then iterating on information for analytics. Efficiency, speed, and accuracy all contribute to a company’s return on their technology investment.”
Other new key features of the Unifi Data Platform include automatic Tag Recommendations. As users explore data they can assign a Tag to a dataset to indicate what type of information it is such as ‘Sales’ or ‘Finance’ which then becomes searchable to other users later. As the AI-engine learns which datasets are of interest to a user, it serves up recommendations based on Tags of the same nature. Tags are then used to improve scoring and add context to the ontology or business glossary and metadata information of a dataset.
Unifi has also dramatically expanded its use of Natural Language Processing in its search feature to auto-complete a query. For example, as a user begins to type, “Find sales…” Unifi will complete the query “forecasts through this quarter,” as it has learned over time that information is frequently requested by that user or team. An extension of that capability is to show the relationship between those aspects in a dataset such as, ‘Show me the permissions’ would indicate the governance rules applied to that dataset.
“The use of artificial intelligence and natural language processing is a key technology differentiator in today’s crowded market of data catalog and data preparation tools. A high level of data self-service dramatically reduces the burden on IT teams to support data access requests from across the organization,” said Dave Wells, Practice Director, Data Management at Eckerson Group. “Data is a raw commodity and an intelligent data catalog helps a company to mine the data that is most relevant to its users and add context to it – and that’s a real tangible value to a business. The Unifi Data Platform is a leader in innovative and intelligent data cataloging using artificial intelligence and natural language processing.”
The newest release of the Unifi Data Platform is generally available now.
Recommended Read: What Makes Some Companies Heroes and Others Villains?