Human-computer interaction method and system
A human-computer interaction and intent technology, applied in the field of human-computer interaction methods and systems, can solve problems such as inaccurate understanding, and achieve the effects of fast system processing, simple operation, and saving processing time for labeling
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0094] The similarity between sentences A and B can be calculated using the following formula:
[0095] sim(A,B)=α*semanticSim(A,B)+β*syntaxSim(A,B)+γ*classSim(A,B)
[0096] Where α+β+γ=1, α>β,γ
[0097] In addition, the neural network can also be used to calculate. After the sentence is vectorized, CNN, RNN or RNN+attention (attention cycle neural network) is used to train the similarity model by calculating the Euclidean distance or cosine angle between two sentences. So as to get the similarity of two sentences. The calculation of this embodiment is simple and easy to explain.
Embodiment 2
[0099] Sentences with the same intent are considered similar sentences, sentences with different intents are considered dissimilar, and the trained model can calculate the similarity between the two sentences.
[0100] The similarity between sentence A and sentence B can be expressed by the following simple formula:
[0101] sim(A,B)=f(Wx1+b,Wx2+b), X1 and X2 are the vectors of sentence A and sentence B respectively, W and b are neural network parameters, and f is the similarity calculated by Euclidean distance or cosine angle degree function. This embodiment requires a large amount of corpus for training, so the accuracy is high.
Embodiment 3
[0103] Sentences and intentions under the same intention are considered similar, and sentences and intentions under different intentions are considered dissimilar. The trained model can calculate the similarity between sentences and intentions.
[0104] The similarity between sentence A and intent C can be expressed by the following simple formula:
[0105] sim(A,C)=f(Wx1+b,Wx2+b), X1 and X2 are the vectors of sentence A and intention C respectively, W and b are neural network parameters, and f is the similarity calculated by Euclidean distance or cosine angle degree function. This embodiment is based on a large amount of corpus training, and the calculation speed is fast, which can effectively improve the response speed of the system.
[0106] The system can use the method of Embodiment 1 to calculate the similarity in the initial stage. After gradually accumulating a large amount of corpus, it can transition to the method of Embodiment 2 or Embodiment 3. When the performanc...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


