In the big data environment, information retrieval efficiency has become a key factor restricting information processing and utilization. In order to improve information retrieval efficiency, this study specifically combines the ER (Entity Retrieval) algorithm to optimize the retrieval model and algorithm structure, thereby achieving fast and accurate information retrieval in the big data environment. In order to verify the efficiency of the method in this paper, a comparative experiment is carried out and it is found that the accuracy of the ER algorithm in information retrieval tasks is as high as 89.91%, and the recall rate is 87.23%. Compared with other traditional methods, its accuracy and recall rates have been greatly improved. Therefore, this study believes that the ER algorithm is an efficient and reliable information retrieval technology, which is of great significance for improving information processing capabilities in big data environments.