The invention belongs to the technical field of enterprise information
intelligent management, and provides an enterprise information intelligent question and answer
robot, which is characterized in that dispersed user questions are subjected to word segmentation, stop word removal, vectorization, clustering and the like through an NLP technology, the questions are automatically summarized into unified standard questions, high-
score answers authenticated by employees are integrated to form a high-quality resource
library, and the high-quality resource
library is stored in a
database; by analyzing the questioning frequency and the questioned priority calculated according to the interval rule, the high-frequency focus of attention of an enterprise is accurately captured, the priority response of core questions is ensured, the staff
information retrieval time is shortened, meanwhile, dual matching sorting is performed in combination with the current business priority of the enterprise, and the sorting consistency is dynamically judged through the Spearman
correlation coefficient, so that the efficiency is improved. And when the double matching sorting is inconsistent, a comprehensive sorting strategy is adopted, so that the efficient response of a daily high-frequency problem is ensured, knowledge strongly related to the core business can be preferentially pushed in the business peak period, and the
information supply is deeply matched with the enterprise operation
rhythm.