Intel IT has developed a context-aware recommender system (CARS) to address big data predictive analytics challenges involving expanding data warehouses, constantly changing contextual parameters, and fast response times.
The CARS manages the massive amounts of data associated with recommendation engines—information filtering systems that predict the rating of products and services— and adds the intelligence of immediate contextual parameters, such as time of day, location, and weather. As a result, users receive more relevant recommendations that are based on a combination of their historical preferences and contextual parameters.
By building the CARS with Intel® Distribution for Apache Hadoop* software (Intel® Distribution) we have accomplished the following:
• Shortened time to market.
• Expanded revenue-generation opportunities.
• Built a reusable recommendation engine.
We built the initial CARS architecture for our location-based-services BU, which recommended coupons to customers using a mobile navigation application. An eight-week pilot program showed that customers using our solution exchanged 45 percent more coupons than a control group that received coupons based on location only.
Since the success of the initial pilot, the CARS has been repurposed for a sales and marketing BU that needed assistance with promoting the right product to the right customer at the right time. Because the CARS is flexible and applicable in a variety of use cases, other BUs are exploring ways to use it to take advantage of its big data predictive analytics capability.
Read the full Using Apache Hadoop* for Context-Aware Recommender Systems White Paper.