User Review Summarization in Russian

The prevalence of online shopping has made it the foremost method for purchasing various goods. The online customer reviews play a crucial role in providing valuable insights into customers’ interests and knowledge of the product. Recent articles have focused on extracting sentiment and aspect information from reviews and incorporating it into the review summarization process. Unfortunately, the majority of these methods have primarily worked with the English language and have been evaluated exclusively on English language datasets. Based on recent advances on the topic, this work researches review summarization methods in application to the Russian language. We collect a corpus of Russian reviews and evaluate the models on manually created summaries and summaries from an existing English language dataset in the same domain. Specifically, the best ROUGE-1/2/L scores of 33.87/2.87/12.20 on the collected data are achieved by fine-tuned AceSum and the best ROUGE-1/2/L scores of 30.67/6.88/18.55 on Space dataset are achieved by PlanSum. Additionally, we investigate the impact of fine-tuning these models on a small subset of data entries from different dataset.