Comment reviewing is a valuable form of social learning in Video-Based Learning (VBL) environments as it helps learners expand on what they learned from videos and explore aspects of a topic they might otherwise overlook. However, asking all learners to review the same comments regardless of their levels of knowledge may not be beneficial. When comments do not contribute meaningfully to learning, learners may become less motivated to review them or approach the task with less seriousness, potentially leading to a lack of attentiveness in comment reviewing. To address this issue, we present an approach for recommending comments to learners based on their knowledge. We compare various measures of engagement and learning using data collected from the same first-year university course through our VBL platform over two different years: 2023 (control group, with identical comments to review for all learners) and 2024 (experimental group, with personalized comments to review). Our findings indicate that learners who received personalized comments interacted more with the VBL platform as measured by the number of visits to the comment rating pages, and higher number of ratings and replies to comments. Our intervention resulted in higher learning gain for learners, particularly for those who requested more recommendations.