Graph databases gaining in popularity, but confusion still clouds market
Companies are beginning to use graph databases for a variety of business applications, but the market is still evolving.
“We see a lot of experimentation, but not a lot of production use,” said Carl Olofson, research vice president, data management software research at International Data Corp. (IDC). “I have spoken with firms that are using this technology for [business] critical uses, though, in areas like fraud detection, law enforcement, and big data analysis and ‘wrangling.’”
One use that comes up frequently is that of pattern recognition for security purposes, including detection of suspicious activity on such things as online bank accounts and credit cards.
“They are used by law enforcement and intelligence services to graph networks of associations of persons of interest,” Olofson said. “They are used for social media analysis in a similar manner, to determine who are the influencers and who are the followers, and what trends excite the influencers.”
Graph databases are also used for network analysis, such as for detecting weak spots in power grids or overworked parts of a computer network, and for data analysis by data scientists.
“They also fit well with geo-spatial analysis, since spatial structures are essentially graphs anyway,” Olofson said. “These various use cases require a variety of technologies. Some need to be highly scalable, some need to be very fast. Some emphasize the depth and complexity of the graph over other factors.”
There is still confusion in the market, Olofson said, in part caused by the many vendors of other types of database management systems that offer some graph support features.
"This may inhibit adoption, as users try to find an ‘all-in-one’ solution,” Olofson said. “One thing that should propel future growth and adoption is cognitive computing. Graphs are ideal for the capturing and manipulation of the kinds of data structures that cognitive systems employ.”