I have implemented Cheng et al, “Generate natural language explanations for recommendation”-EARS-2019. In this paper, the authors propose a hierarchical sequence-to-sequence model (HSS) with auto- denoising for personalized recommendation and natural lan- guage explanation generation. In particular, the paper makes the following contributions: 1. Authors propose a hierarchical generation model, which is able to collaboratively learn over multiple sentences from different users for explanation sen- tence generation. 2. Based on item feature words extracted from reviews, The authors propose a feature-aware attention model to implicitly select explanation sentences from reviews for model learning, and we further introduce a feature atten- tion model to enhance the feature-level personality of the ex- planations. For details see the report