Method and device for classifying questions based on artificial intelligence

1. A method for classifying questions based on artificial intelligence, comprising:

acquiring, by at least one computing device, text content of a question input by a user;

performing, by the at least one computing device, a word segmentation process on the text content to obtain a plurality of segmentations;

acquiring, by the at least one computing device, hidden representation vectors of the plurality of segmentations;

generating, by the at least one computing device, a first vector of the text content according to the hidden representation vectors; and

determining, by the at least one computing device, a target responder corresponding to the question from a number M of preset responders according to the first vector and a preset classification model, wherein the number M of preset responders are robots, each designed to automatically respond to different classifications of questions, where M is a positive integer; and

appointing, by the at least one computing device, the target responder to the user;

wherein determining, by the at least one computing device, a target responder from a number M of preset responders corresponding to the question according to the first vector and a preset classification model comprises:

converting, by the at least one computing device, the first vector into a second vector according to the number M, wherein a length of the second vector is M;

converting, by the at least one computing device, the second vector into M probabilities by the preset classification model;

determining, by the at least one computing device, a maximum probability among the M probabilities; and

selecting, by the at least one computing device, a responder corresponding to the maximum probability as the target responder from the preset responders according to the maximum probability.

2. The method according to claim 1 , wherein acquiring, by the at least one computing device, hidden representation vectors of the plurality of segmentations comprises:

acquiring, by the at least one computing device, word vectors of the plurality of segmentations; and

converting, by the at least one computing device, the word vectors into the hidden representation vectors based on a Neural Network Model.

3. The method according to claim 2 , wherein the Neural Network Model is obtained by:

acquiring, by the at least one computing device, data of sample words;

training, by the at least one computing device, the data with a text deep representation model word2vec, to obtain the word vectors of the sample words; and

performing, by the at least one computing device, a train by using a loss function as a training target function and by using a marked corpus and the word vectors of the sample words as sample data, to obtain the Neural Network Model.

4. The method according to claim 1 , wherein acquiring, by at least one computing device, text content of a question input by a user comprises:

determining, by the at least one computing device, how the question is inputted by the user;

if the question is inputted via speech, performing, by the at least one computing device, a speech identification on the speech inputted;

if the question is inputted via text, acquiring, by the at least one computing device, the text content of the question.

5. A device for classifying questions based on artificial intelligence, comprising:

a processor;

a memory, configured to store instructions executable by the processor; wherein the processor is configured to:

acquire text content of a question input by a user;

perform a word segmentation process on the text content to obtain a plurality of segmentations;

acquire hidden representation vectors of the plurality of segmentations;

generate a first vector of the text content according to the hidden representation vectors;

determine a target responder corresponding to the question from a number M of preset responders according to the first vector and a preset classification model, wherein the number M of preset responders are robots, each designed to automatically respond to different classifications of questions, where M is a positive integer; and

appoint the target responder to the user;

wherein the processor is configured to determine a target responder corresponding to the question from a number M of preset responders according to the first vector and a preset classification model by acts of:

converting, by the processor, the first vector into a second vector according to the number M, wherein a length of the second vector is M;

converting, by the processor, the second vector into M probabilities by the preset classification model;

determining, by the processor, a maximum probability among the M probabilities; and

selecting, by the processor, a responder corresponding to the maximum probability as the target responder from the preset responders according to the maximum probability.

6. The device according to claim 5 , wherein the processor is configured to acquire hidden representation vectors of the plurality of segmentations by acts of:

acquiring word vectors of the plurality of segmentations; and

converting the word vectors into the hidden representation vectors based on a Neural Network Model.

7. The device according to claim 6 , wherein the processor is configured to obtain the Neural Network Model by acts of:

acquiring data of sample words;

training the data with a text deep representation model word2vec, to obtain the word vectors of the sample words; and

performing a train by using a loss function as a training target function and by using a marked corpus and the word vectors of the sample words as sample data, to obtain the Neural Network Model.

8. The device according to claim 5 , wherein the processor is configured to acquire text content of a question input by a user according to acts of:

determining how the question is inputted by the user;

if the question is inputted via speech, performing a speech identification on the speech inputted;

if the question is inputted via text, acquiring the text content of the question.

9. A non-transitory computer readable storage medium, with a computer program stored, wherein the computer program is executed by a processor to achieve a method for classifying questions based on artificial intelligence, and the method comprises:

acquiring text content of a question input by a user;

performing a word segmentation process on the text content to obtain a plurality of segmentations;

acquiring hidden representation vectors of the plurality of segmentations;

generating a first vector of the text content according to the hidden representation vectors;

determining a target responder from a number M of preset responders corresponding to the question according to the first vector and a preset classification model, wherein the number M of preset responders are robots, each designed to automatically respond to different classifications of questions, where M is a positive integer; and

appointing the target responder to the user;

wherein determining a target responder corresponding to the question from a number M of preset responders according to the first vector and a preset classification model comprises:

converting, by the processor, the first vector into a second vector according to the number M, wherein a length of the second vector is M;

converting, by the processor, the second vector into M probabilities by the preset classification model;

determining, by the processor, a maximum probability among the M probabilities; and

selecting, by the processor, a responder corresponding to the maximum probability as the target responder from the preset responders according to the maximum probability.

10. The non-transitory computer readable storage medium according to claim 9 , wherein acquiring hidden representation vectors of the plurality of segmentations comprises:

acquiring word vectors of the plurality of segmentations; and

converting the word vectors into the hidden representation vectors based on a Neural Network Model.

11. The non-transitory computer readable storage medium according to claim 10 , wherein the Neural Network Model is obtained by:

acquiring data of sample words;

training the data with a text deep representation model word2vec, to obtain the word vectors of the sample words; and

performing a train by using a loss function as a training target function and by using a marked corpus and the word vectors of the sample words as sample data, to obtain the Neural Network Model.

12. The non-transitory computer readable storage medium according to claim 9 , wherein acquiring text content of a question input by a user comprises:

determining how the question is inputted by the user;

if the question is inputted via speech, performing a speech identification on the speech inputted;

if the question is inputted via text, acquiring the text content of the question.