MANAGING KNOWLEDGE COMPLEXITY THROUGH VISUALIZATION

Qlik has the ability to combine multiple data sources easily, load them into its repository, and associate these sources. “Since all the data is stored in memory,” said Denise LaForgia, Qlik’s senior director of product marketing, “response time is very rapid. As the user hones in on a datapoint or field, all of the data updates automatically on those relationships.”

Qlik’s acquisition of Attunity in 2019 contributed to its data integration capabilities, and the company has built out its analytics features internally as well. “Augmented analytics that assists and guides the users, along with predictive analytics for business users, helps trigger action off the data,” noted LaForgia. Qlik’s augmented analytics capabilities leverage AI to deliver automated insight generation, authoring assistance, and natural language interaction. For example, a user can select the desired type of analysis, and Qlik auto-generates the visualizations. This support is particularly helpful in enabling business users to quickly and easily explore complex datasets.

Qlik’s visualization capability was developed internally and continues to expand. The ability to explore the data offers benefits that are not available through traditional BI analytics. “In retail sales analysis, users can not only see products that are sold and at what volume, but can look at certain products that are not being sold in a particular market,” LaForgia continued. “They can dig into that information to see if there is an association with particular sales reps, for example, or with some other factor. People can discover relationships that they might not have thought to ask about before seeing it visualized, and gain insights much more easily than they could have by making a series of queries.”

Graph database visualization

Graph databases are a step above the static, simple graphs and charts found in market research reports and business documents. Well-suited to presenting complex data that is relationship-based, the connections among entities are what’s stored. The node and edge visualization showing entities and their connections is one form in which the data can be presented. The initial concept is relatively easy to understand, but once the number of datapoints increases, so does the difficulty of presenting the data clearly. Graph database vendors generally provide basic visualizations as part of their product, but these are usually limited in the amount of data they can present at once and in their degree of interactivity.

Cambridge Intelligence, which sees graphs as models, provides a software developers kit (SDK) for graph data visualizations that works with all the leading graph databases and provides a sophisticated set of visualizations. KeyLines, its flagship product, is a JavaScript toolkit, and ReGraph is a similar kit for React, a JavaScript library for developing user interfaces. A third product, KronoGraph, shows patterns across time. Developers use these products to build applications for end users.

Common use cases for these tools include intelligence analysis, anti-fraud initiatives, cybersecurity, and supply chain analytics. “A visual layer provides an understanding of the connections in large datasets that would not be possible otherwise,” said Corey Lanum, chief product evangelist at Cambridge Intelligence. “When there is a lot of data, the ability to filter in or out certain portions of the visualized data or identify connections in groups of data elements is essential to zeroing in on what’s relevant.”