Providing recommendations based on user-generated post-purchase content and navigation patterns

Abstract

A method including generating a weighting vector for a first user. The weighting vector can track a weight corresponding to each feature of a plurality of features. The plurality of features can represent purchasing criteria that are common to each item in a category of items. The method also can include, in response to receiving a request from the first user to view details for a selected item, recommending to the first user, in real-time after receiving the request, one or more other items that are different from the selected item. The selected item and the one or more other items can be in a category of items that is the same. The plurality of features can be in common for each item in the category of items. Sentiment data can include a sentiment score for each feature for each item in the category of items. The one or more other items can be recommended based on the sentiment score of one or more first features of the plurality of features for each of the one or more other items exceeding the sentiment score of a corresponding one of the one or more first features for the selected item. The method additionally can include receiving a new request from the first user to view details for a new selected item that is selected from among the one or more other items. The method further can include updating the weight that corresponds to each of the one or more first features for the new selected item in the weighting vector for the first user. Other embodiments of related systems and methods are disclosed.