Graph analytics is no longer just for the Facebooks or Twitters of the world. Get ready—you can do this too!
The more I talk to enterprise data analysts and data scientists from a wide range of industries, the more convinced I am that graph analysis is the key for unlocking insights hidden in enterprise big data. Visual-based analytics can like graph analysis can reveal unexpected and unknown relationships, patterns and connections in structured and unstructured data sets and expose valuable business advantages.
Don’t take my word for it. Stop by Intel booth #101 at Strata Conference for a demo of new graph analytics technologies, and then let us know how data visualization technologies can help you unlock big data insights at your business. Watch the video Realizing Value from Big Data with Graph Analytics with Ted Willke, Intel’s principal engineer and GM of Graph Analytics, to learn how graphic processing technologies can help you gain insights into your business relationships and connections that can make your company more competitive.
Graph analytics is ground-breaking because traditional data analytics technologies are largely blind to the unique insights contained within big data. Graph processing applies algorithms to data that query and mine relationship patterns and uncover hidden connections and anomalies. Graph analytics provides a highly visible form of analysis that’s particularly effective for pulling insights out of Web content, social media, images, system and call logs, and data from the Internet of Things (#IoT).
Graph analysis essentially models data as a network, which lets you intuitively visualize connections and patterns among data. The visual nature of graph analytics makes it easy to pinpoint and calculate critical attributes like the shortest paths, central influencers, and important sub-networks.
For instance, take a look at this map of my LinkedIn contacts. Here you can see how graph analytics has taken data about my contacts and charted it as a network of relationships, with me as the hub. It points out a number of central influencers – handy for me to know if I need to get information dispersed quickly – and also shows unexpected links between individuals in different networks, which can point me to the quickest link between groups of people.
A graph analysis from LinkedIn uncovers the hidden relationships and networks among my contacts.
In real-life business scenarios, graph analytics can help provide insights into improving customer targeting, with graphs pointing the way toward clearer segmentation of social network data to foster personalized recommendations and slow customer churn. Graph analytics can also be used to detect patterns in such areas as genetic analysis and fraud detection, and provide insights into security forensics and health treatment optimization.
Intel’s graph analytics technologies take tools and processes previously available only to the most technically sophisticated organizations, and from them creates a set of integrated, end-to-end analytics capabilities that address a broad range of big data business challenges. The technology unifies workflows and provides a common, simple-to-use interface to deliver big data graph analytics to a much broader audience.
However, these capabilities, which will feature in future Intel products, deliver much more than pretty pictures. Intel’s graph analytics technologies make advanced analytics easier for both experts and new practitioners by unifying the tools for data cleansing, graph construction, graph query, and graph analytics into an integrated platform that runs on Intel’s Hadoop distribution. Specifically, this is the unification of Graph Builder with the Apache Titan* & Giraph* projects.
It’s time for graph analytics to go mainstream.
Please seek @TimIntel out at the conference as I tweet man-on-the-street impressions so we can exchange tweets, MTs & RTs. Follow Tim and the Big Data community for Intel at @TimIntel & @IntelHadoop.