Searching on the web gets better with semantic technology. How about searching in your local supermarket?

Point Inside launched a private ad network offering this week that blends indoor shopper- and product-location technologies with proprietary semantic matching features to understand how what’s on users’ digital grocery lists can connect to what’s on a store’s shelves – and to opportunities to deliver in-context deals to consumers who opt in at the appropriate point in time.

“People are bringing smart phones into stores, and that means brick and mortar retailers can now do what online retailers have been taking advantage of – putting ads and suggestions in front of people” based on data insight, says Point Inside CMO Todd Sherman. The new service gives the store control of the network, dubbed nSide, through integration with their own mobile apps, letting the store or brands that buy space on its network deliver coupons, discounts, recipes or how-to videos correlating to shoppers’ electronic lists and their in-store location.

Currently, retailers such as supercenter Meijer use its technology as an opt-in service to help shoppers pinpoint the location of items in their stores and the most efficient routes to get to them. They can search for sale items and coupons, too, but with the new platform, those connections are automatically made for and presented to them.

Point Inside’s take on semantic technology tends to be more along the lines of synonym lookups around search to match, for example, a supermarket’s classification of a product with the various ways a user might describe it on an electronic grocery list. Machine learning algorithms continue to learn more from consumers the more they shop using an electronic list, from the repeated use of product descriptions, Sherman says. Point Inside also can provide recommendations of additional items a consumer might like to buy based on what’s already on her current list, search information, and past history as well as associated item relationships in its database. Sherman also sees as a logical next step bringing natural language processing to the picture to help in interpreting users’ information.

But the more exciting technology may have less to do with breaking ground in search or recommendation capabilities per se and more with how their use is enabled. Its micro-location capabilities let it use a combination of available WiFi networks, spatial indexing and other proprietary technologies to understand from access points and mobile phone sensor fusion the indoor mapping of a store and exactly where shoppers are inside it to provide those relevant offers – say, 50 percent off a particular peanut butter brand for those with intent to buy a jar – when the consumer is nearby to take advantage of them. “We combine all that deep recommendation technology we have with location information, where you are in the store, so that it’s relevant and within proximity, which drives sales,” Sherman says. “When it comes time to give someone an ad, we don’t give them an ad for something in aisle 37 when they are in aisle 1.”

The private ad network, Sherman believes, has advantages over more general solutions such as Google Maps indoor mapping can support. “A private ad network gives us the ability to zero in on the behaviors of an Albertson’s or a Stop and Shop’s customers, to understand them and mold the network around those behaviors vs. a generalized ad network like Google’s that doesn’t really have the ability to differentiate effectively.”