Jump page for the Amazon search demo

What is your name?

What is your Amazon search query?

Some recent searches

Thu, Mar 8, 2012 02:14:13 PM PST "orbit professional gutter cleaner" info 16
Thu, Mar 8, 2012 02:12:49 PM PST "patra beaded mesh sheath dress" info 16
Thu, Mar 8, 2012 02:12:29 PM PST "elizabeth arden flawless finish mousse makeup natural 02 nib sealed" info 16
Thu, Mar 8, 2012 11:55:50 AM PST "pickles" info 8
Thu, Mar 8, 2012 11:53:03 AM PST "men's XL shoes" info 15
Thu, Mar 8, 2012 11:51:53 AM PST "Disney glasses" info 9
Wed, Dec 11, 2013 05:28:39 AM PST "55 gallon passion lube" info 0
Thu, Mar 8, 2012 11:55:39 AM PST "white kitchen timer" info 8
Thu, Mar 8, 2012 02:13:07 PM PST "kenneth cole reaction black leather wide band men's casual dress watch rk1103" info 3
Thu, Mar 8, 2012 11:52:09 AM PST "grey mouse pad" info 9
Tue, Mar 18, 2014 11:29:36 AM PDT "teal kitchen timer" info 0
Wed, Mar 15, 2017 08:48:56 AM PDT "find a pen" info 8
Thu, Mar 8, 2012 02:13:14 PM PST "authentic christian louboutin moustique 11 crepe satin black heels" info 16
Thu, Mar 8, 2012 11:53:21 AM PST "men's XL white shoes" info 10

About the demo

We use natural language processing to determine the main product the user wants, as well as various distinctive attributes of it. We also use natural language processing on each returned Amazon product and its browse node relationship, to decide what object it is and what attributes it makes available. We combine that information to decide how appropriate an item is for the user's request. The items are then displayed in an ordering that better reflects the user's intent.

When Amazon cannot produce a reasonable array of products to match the user's actual request, we take the main item and what we consider to be the important attributes being requested, and make a subset search request that is more likely to be matched.

Link to All Recent Searches
Link to Pitch