Your company manages an application that aggregates news articles from many different online sources and sends them to users. You need to build a recommendation model that will suggest articles to…

Question

Options

• ACreate a collaborative filtering system that recommends articles to a user based on the user's
• BEncode all articles into vectors using word2vec, and build a model that returns articles based on
• CBuild a logistic regression model for each user that predicts whether an article should be
• DManually label a few hundred articles, and then train an SVM classifier based on the manually

• B
  89% (24)
• C
  4% (1)
• D
  7% (2)

Explanation

The task is content-based filtering - finding articles similar to what the user is currently reading based on article content, not user history. Collaborative filtering (A) recommends based on what similar users read, which requires user history and does not directly compare article content. Logistic regression per user (C) is user-centric, computationally expensive to maintain at scale, and relies on labeled training data. Manual labeling + SVM (D) is not scalable to millions of articles and requires significant human effort. Word2Vec (B) encodes each article as a dense semantic vector by aggregating word embeddings. Articles covering similar topics will have similar vector representations, so nearest-neighbor search in the embedding space efficiently returns the most relevant articles regardless of exact wording - the ideal approach for content similarity at scale.

No community discussion yet for this question.