Reading task processing method and related equipment

Abstract

The invention relates to the technical field of artificial intelligence, and provides a reading task processing method and related equipment, and the reading task processing method comprises the steps: obtaining reading information, target problem information associated with the reading information, and historical dialogue information; executing an iterative processing step until the historical dialogue information of all rounds is processed, and taking the current moment information representation of the reading information determined based on the historical dialogue information of the last round as the first information representation of the historical dialogue information in the reading information; determining a second information representation of the target problem information in thereading information based on the first information representation; and determining reply information for replying to the target question information in the reading information based on the second information representation. The embodiment of the invention is beneficial to improving the accuracy of predicting the reply result based on the target question information.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, and in particular, the present application relates to a reading task processing method and related equipment. Background technique [0002] With the development of artificial intelligence technology, artificial intelligence technology has been applied in many technical fields. When artificial intelligence technology is applied to the field of conversational reading comprehension, it is generally used to allow machines to perform reading comprehension tasks. The reading comprehension task mainly refers to letting the machine answer questions related to the reading information according to the specified reading information. When the machine performs the reading comprehension task, it is necessary to model the relationship between question information and reading information, question information and historical dialogue information. [0003] In the prior art, when modelin...

AI Technical Summary

Problems solved by technology

Among them, the splicing method is based on the pre-trained language model, using the self-attention mechanism of the model to simultaneously model the relationship between question information and reading information, question information and historical dialogue information; but because the model has a maximum input sequence Due to the limitation of length, when the content of reading information is large or the number of rounds of historical dialogue information is large, it is impossible to consider all the dialogue history, which limits the performance of the model and makes the answer results predicted by the model based on question information less accurate.

The history encoding method is based on the HAE (History Answer Embedding) principle. This method loses the time depth information of historical dialogue information (the distance between each round of historical dialogue information and the current round of dialogue information), and it is easy for the model to model When learning a shortcut to predict data, for example, in the existing data set, the position of the predicted answer result of each round of dialogue in the reading information is positively correlated with the historical dialogue round, that is, the answer result of the i-th round of question information is often in the The answer result of the i-1 round of question information is not far behind, so the model cannot learn the expected semantic analysis ability, resulting in a low accuracy of the answer result predicted by the model based on question information

Images