Computer-readable recording medium storing learning program, learning method, and learning device
A non-transitory computer-readable recording medium storing a learning program that causes a computer to execute a process, a process includes acquiring a first input sentence and a first summary sentence into which the first input sentence is summarized, generating a second summary sentence to which the first summary sentence is partially-changed, and executing machine learning for a model to generate at least one summary sentence that corresponds to at least one input sentence, respectively, in response to an input of the at least one input sentence, such that a first probability of generating the first summary sentence in response to the input of the first input sentence becomes higher than a second probability of generating the second summary sentence in response to the input of the first input sentence, based on the first input sentence, the first summary sentence, and the second summary sentence.
1. A non-transitory computer-readable recording medium storing a learning program that causes a computer to execute a process, the process comprising:
acquiring a first input sentence and a first summary sentence into which the first input sentence is summarized;
generating a second summary sentence to which the first summary sentence is partially-changed; and
executing machine learning for a model to generate at least one summary sentence that corresponds to at least one input sentence, respectively, in response to an input of the at least one input sentence, such that a first probability of generating the first summary sentence in response to the input of the first input sentence becomes higher than a second probability of generating the second summary sentence in response to the input of the first input sentence, based on the first input sentence, the first summary sentence, and the second summary sentence.
2. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the model is capable of calculating a probability that each of a plurality of words appears in the at least one summary sentence, and is capable of generating the at least one summary sentence, based on the calculated probability, and
wherein the machine learning includes:
calculating a third probability with which each of a plurality of words appears in the first summary sentence, and calculating the first probability and the second probability, based on the third probability, with the model, and
updating the model based on the first input sentence, the first summary sentence, and the second summary sentence in a case where the first probability is higher than the second probability.
3. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the process generates the second summary sentence by randomly selecting a named entity that appears in the first summary sentence, by masking the named entry, and by searching a training data group for the at least one summary sentence partially similar to the first summary sentence that includes the masked named entry.
4. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the process generates the second summary sentence by changing a numerical value included in the first summary sentence to another numerical value.
5. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the process generates the second summary sentence by randomly selecting a word included in the first summary sentence and by changing the word to another word other than the selected word.
6. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the at least one summary sentence that corresponds to a target input sentence is generated by using the model.
7. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the model is a neural network.
8. A learning method that causes a computer to execute a process, the process comprising:
acquiring a first input sentence and a first summary sentence into which the first input sentence is summarized;
generating a second summary sentence to which the first summary sentence is partially-changed; and
executing machine learning for a model to generate at least one summary sentence that corresponds to at least one input sentence, respectively, in response to an input of the at least one input sentence, such that a first probability of generating the first summary sentence in response to the input of the first input sentence becomes higher than a second probability of generating the second summary sentence in response to the input of the first input sentence, based on the first input sentence, the first summary sentence, and the second summary sentence.
9. A learning device comprising:
a memory; and
a processer coupled to the memory and configured to:
acquire a first input sentence and a first summary sentence into which the first input sentence is summarized;
generate a second summary sentence to which the first summary sentence is partially-changed; and
execute machine learning for a model to generate at least one summary sentence that corresponds to at least one input sentence, respectively, in response to an input of the at least one input sentence, such that a first probability of generating the first summary sentence in response to the input of the first input sentence becomes higher than a second probability of generating the second summary sentence in response to the input of the first input sentence, based on the first input sentence, the first summary sentence, and the second summary sentence.