With the development of science and technology, scientific literature is mounting to an increasingly large scale. How to quickly and accurately seek the research topics, influential scholars, and high-level papers in a specific domain from the vast amount of publications remains an enormous challenge. The visual analysis method for domain literature data based on word representation model employed word2vec to recommend domain-related keywords by the similarity between word vectors, and filters the domain-related papers according to these keywords. Then it utilized the BERTopic model to extract topics from the abstracts of domain papers. Next, the values for paper impact were calculated using PageRank, and the values for author influence were calculated using Author-Rank, the author impact evaluation method, taking into account the order of authorship, the number of publications, and the impact of papers. Finally, the multi-view collaborative and interactive visualization approach could help researchers gain a quick understanding and analysis of specific areas from multiple perspectives, such as topics word frequency, topics evolution, literature impact, citation relationships, and author impact. The method can be applied to literature data analysis in the field of “food safety”, and the results and user tests can validate this method.