Article Evaluation_Recommender system

Content: In general, all the contents are relevant to the article topic(recommender system), multiple examples are cited to explain the statement, and nothing distracted. However, the article could be improved by removing the data collection part and the Netflix Prize part, which are irrelevant to the topic. The article follows a good structure of Overview - Approaches - Different kinds of recommender systems - Performance measures - Other kinds of recommender systems - Reproducibility. However, the structure could be improved. It could explain different kinds of recommender systems after the overview part, then introduce approaches and performance measures, instead of splitting the variety of recommender systems into two separate parts. In addition, the definition of the recommender system is in only one sentence. The definition part needs to be explained more and go into details because it defines the topic of the article.

Tone: The article maintains neutral tones. Claims in the articles are unbiased. No viewpoints are overrepresented or underrepresented, and it both covers the risks and benefits in a neutral way.

Sources: The article mentioned adequate citation and references from multiple neutral and reliable sources, and cite multiple examples such as Pandora, Netflix and Last.fm. However, the citation still has many things to be modified. First, 7 of 89 sources don't have a "doi" or the retrieved links. Second, some of the sources are out-of-date, such as Breese J.S.(1998) and Sarwar B.(2000). Last, the citations are not in the same format. For example, for the same source from Empirical Analysis of Predictive Algorithms for Collaborative Filtering, the No.32 citation write in the format of "First name, middle name, last name" but the No.41 citation write in the format of "Last name, middle name, first name", and the article put this source into two different number of sources.

Talk: The article is rated as class C by WikiProject Business, WikiProject Computing/Software/Hardware, WikiProject Marketing & Advertising, WikiProject Retailing, WikiProject Statistics and WikiProject Technology. 5 of 6 conversations are about modifying the external links, and the other one is about definition clarification. And the latest conversation is on Dec.14, 2017, which is not so updated.