Machine Learning for Text by Charu C. Aggarwal

Author:Charu C. Aggarwal
Language: eng
Format: epub, azw3, pdf
Publisher: Springer International Publishing, Cham

8.2.5 Application: Recommender Systems with Ratings and Text

Content-based recommender systems use the textual descriptions of items to learn user propensities about particular items. Ratings indicate the degree of like or dislike of users towards items. In such cases, the data for each user is converted into a user-specific text classification problem. The training documents for each user-specific classification problem correspond to the descriptions of items rated by that user, and the dependent variable is its item-specific rating from that user. This training data can be used to learn a user-specific classification or regression model for rating prediction.

However, content-based systems do not use the collaborative power of like-mined users to make predictions. A different class of recommendation methods, referred to as collaborative filtering methods, use the similarities in rating patterns between users and items to make predictions. Let R be an m × n ratings matrix R over m users and n items. The matrix R = [r ij ] is massively incomplete, and only a small subset O of the ratings in R are observed:

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.