Energy big data query method and system

The invention relates to an energy big data query method and system, belongs to the technical field of big data, and solves the problem of large difference between actual query content and expected query content due to semantic difference between query keywords and accurate keywords. The method comprises the steps of generating a first keyword set by utilizing an encoder neural network model based on an original query text, and sending the first keyword set to a data owner by utilizing a PIR technology; according to the first keyword set, accurate query keywords are selected from a keyword index database to generate a second keyword set, and the second keyword set is returned to the data querier; the second keyword set is sent to the data owner through the PIR technology again, then retrieval is conducted in the energy big database based on the second keyword set, an encrypted private data query result is returned to the data querier, and the data owner comprises a keyword index database and the energy big da