Sentence examples for items representation from inspiring English sources

Based on the experiments, we elect and discuss the feature extraction technique for items' representation that performed better among the others.

Thus, the items' representations aim to describe items by their characteristics and the collective appreciation of users toward them.

In order to find similar items, a similarity measure is employed between the previously described items' representations.

This step, which is detailed in the "Sentiment analysis and item representations construction" section, is performed for each item present in the database, producing a set of items' representations.

In the recommendation module, the items' representations are finally analyzed alongside the ratings provided by the users.

Our proposal focuses on the development of methods that produce items' representations based on users' reviews for recommender systems.

At the end of the feature extraction module, the resulting set is used by the items' representations generation module.

We then plan to apply these users' vectors alongside the items' representations in the recommendation process. 1 https://stanfordnlp.github.io/CoreNLP/.

In the content-based approach [2], users' profiles are matched with items' representations using a similarity measure.

This article proposes a recommender system that uses users' reviews to produce items' representations that are based on the overall sentiment toward the items' features.

Another possibility of extension is to apply the items' representations in different attribute aware recommendation algorithms or to apply the system into different data domains.