Data processing method and device, equipment and medium

[0025] Example 1: figure 1 A schematic diagram of a data processing process provided by an embodiment of the present invention, the process includes:

[0026] S101: Determine the target entity contained in the text data to be processed.

[0027] The data processing method provided by the embodiment of the present invention is applied to an electronic device, and the electronic device may be an intelligent device such as a robot, or a server.

[0028] In a practical application scenario, when a user needs to query certain information, the query information can be input to the smart device. The query information may be an input query question, for example, "Who is A's wife?", or a descriptive sentence, for example, "A is the spokesperson of the world brand C". Among them, there are many ways for the user to input query information, which may be input by voice, or by inputting query information in the form of text on the display screen of the smart device, and of course, inputting

[0041] Embodiment 2: In order to simplify the process of entity linking and provide the efficiency, flexibility, and generalization ability of entity linking, on the basis of the above embodiment, in this embodiment of the present invention, the entity linking model completed by pre-training is based on Text data, knowledge records, and the target type corresponding to the knowledge records, to determine whether the knowledge records match the target entities contained in the text data, including:

[0042] Through the coding network in the entity linking model, the text vector corresponding to the text data, the attribute vector corresponding to the knowledge record, and the category vector corresponding to the target type are respectively determined; and

[0043] Through the decoding network in the entity linking model, based on the text vector, attribute vector, and category vector, it is determined whether the knowledge record matches the target entity contained in the text dat

[0113] Embodiment 3: In order to simplify the process of entity linking and provide the efficiency, flexibility, and generalization ability of entity linking, on the basis of the above embodiment, in this embodiment of the present invention, the entity linking model is trained in the following manner:

[0114] Obtain any sample data in the sample set. The sample data includes sample text data, sample knowledge records corresponding to the sample text data, and sample types corresponding to the sample knowledge records. The sample data corresponds to a label, and the label is used to identify the sample knowledge record and the sample text. Whether the entities contained in the data match;

[0115] Through the original entity linking model, based on the sample data, determine the recognition result of whether the sample knowledge record matches the entity contained in the sample text data; and

[0116] Based on the labels and the recognition results, the original entity link model