CPC G06Q 30/0627 (2013.01) [G06Q 30/0282 (2013.01); G06Q 30/0631 (2013.01); H04L 67/306 (2013.01); G06N 20/00 (2019.01)] | 31 Claims |
1. A computer-implemented method for recommending content to a target user, the method being performed by one or more processors, the method comprising the steps of:
(a) identifying a target user;
(b) identifying a set of data, said set of data including at least:
(i) a rating by said target user for a rated product,
(ii) a review by a reviewer for said rated product,
(iii) a review by said reviewer for an unrated product, and
(iv) an identity of said reviewer;
(c) for said rated product:
(i) identifying one or more n-grams contained in said review by said reviewer, and
(ii) linking said one or more of said n-grams identified in step (c)(i) with said identity of said reviewer who provided said review containing said n-grams;
(d) after step (c), training a classification or prediction model using the data linked in step (c) and said rating by said target user for said rated product; and
(e) after step (d), determining a predicted rating for said unrated product wherein said classification or prediction model determines said predicted rating for said unrated product based upon the presence of said one or more n-grams in said review by said reviewer of said unrated product.
|