Amazon Technologies, Inc.

Alexey Korobko

Emilio Ian Maldonado

Subhrangshu Nandi

Type ambiguity detection methods and systems are provided to predict a browse node classification for items in an electronic catalog and determine ambiguity between browse nodes based at least partly on the predicted browse node classifications for the items. For example, the type ambiguity detection system may identify a plurality of search queries previously submitted by users of the electronic catalog, determine a plurality of browse nodes that are frequently selected after submitting the search queries, determine items that consumers frequently view or purchase in connection with the search queries, input the items into a classifier, receive an output predicting a proper item type or browse node for a given item, and initiate an action in response to determining ambiguity.

Growing community of inventors

Sammamish, WA, United States of America

Alexey Korobko