| CPC G06F 16/383 (2019.01) [G06F 16/3329 (2019.01); G06F 16/3347 (2019.01)] | 18 Claims |

|
1. A question-answer processing method, comprising:
acquiring an input question and a candidate question-answer pair, wherein the candidate question-answer pair comprises a candidate question and a candidate answer matching the candidate question;
inputting the input question into a first model to obtain an input question vector;
calculating a similarity between the input question vector and a candidate question vector, and calculating a relevance between the input question vector and a candidate answer vector, wherein the candidate question vector is a vector corresponding to the candidate question, and the candidate answer vector is a vector corresponding to the candidate answer; and
determining a matching degree of the candidate answer and the input question according to the similarity and the relevance;
wherein the calculating the relevance between the input question vector and the candidate answer vector comprises:
calculating the relevance between the input question vector and the candidate answer vector by a formula 2, wherein the formula 2 is:
![]() rel(vq, va) is the relevance, vq is the input question vector, va is the candidate answer vector, vqT·va is the transpose of the input question vector and the dot product of the candidate question vector, |vq| is the length of the input question vector, |va| is the length of the candidate answer vector.
|