Alignment generation device and alignment generation method
A computer generates a plurality of encoded sentences in the first language by encoding a plurality of sentences in the first language in prescribed units. Next, the computer generates a plurality of encoded sentences in the second language by encoding a plurality of sentences in the second language, each of which is associated with each of the plurality of sentences in the first language, in the prescribed units. The computer generates alignment information based on a code included in each of the plurality of encoded sentences in the first language and a code included in an encoded sentence in the second language, which is associated with each of the plurality of encoded sentences in the first language. The alignment information indicates an alignment between a plurality of codes in the first language and a plurality of codes in the second language.
1. A non-transitory computer-readable recording medium having stored therein an alignment generation program that causes a computer to execute a process comprising:
generating a first plurality of encoded sentences by encoding, in accordance with prescribed units, a plurality of sentences in a first language;
generating a second plurality of encoded sentences by encoding, in accordance with the prescribed units, a plurality of sentences in a second language, each of which is associated with each of the plurality of sentences in the first language; and
generating relationship information indicating relationships between a first plurality of codes in the first plurality of encoded sentences and a second plurality of codes in the second plurality of encoded sentences, in accordance with a number of combinations of a first code included in a third plurality of codes and a second code included in a fourth plurality of codes, the combinations of the first code and the second code being included in a first plurality of combinations of each of the third plurality of codes and each of the fourth plurality of codes, the first plurality of combinations being generated for each of a second plurality of combinations generated by selecting each of the first plurality of encoded sentences and each of the second plurality of encoded sentences, the third plurality of codes being included in the selected first encoded sentence and the fourth plurality of codes being included in the selected second encoded sentence.
2. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the first code is associated with the second code.
3. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the encoding of the plurality of sentences in the first language includes generating first index information having a logical value indicating whether each of the first plurality of codes is included in each of the first plurality of encoded sentences,
the encoding of the plurality of sentences in the second language includes generating second index information having a logical value indicating whether each of the second plurality of codes is included in each of the second plurality of encoded sentences, and
the generating of the relationship information is executed in accordance with the first index information and the second index information.
4. The non-transitory computer-readable recording medium according to claim 1 ,
wherein the prescribed units correspond to a string in which a representation of a word or a plurality of words is normalized.
5. An alignment generation device comprising:
a memory configured to store a plurality of sentences in a first language and a plurality of sentences in a second language, each of which is associated with each of the plurality of sentences in the first language; and
a processor coupled to the memory and the processor configured to
generate a first plurality of encoded sentences by encoding, in accordance with prescribed units, the plurality of sentences in the first language,
generate a second plurality of encoded sentences by encoding, in accordance with the prescribed units, the plurality of sentences in the second language, and
generate relationship information indicating relationships between a first plurality of codes in the first plurality of encoded sentences and a second plurality of codes in the second plurality of encoded sentences, in accordance with a number of combinations of a first code included in a third plurality of codes and a second code included in a fourth plurality of codes, the combinations of the first code and the second code being included in a first plurality of combinations of each of the third plurality of codes and each of the fourth plurality of codes, the first plurality of combinations being generated for each of a second plurality of combinations generated by selecting each of the first plurality of encoded sentences and each of the second plurality of encoded sentences, the third plurality of codes being included in the selected first encoded sentence and the fourth plurality of codes being included in the selected second encoded sentence.
6. The alignment generation device according to claim 5 ,
wherein the first code is associated with the second code.
7. The alignment generation device according to claim 5 ,
wherein the processor is configured to:
generate first index information having a logical value indicating whether each of the first plurality of codes is included in each of the first plurality of encoded sentences;
generate second index information having a logical value indicating whether each of the second plurality of codes is included in each of the second plurality of encoded sentences; and
generate the relationship information in accordance with the first index information and the second index information.
8. The alignment generation device according to claim 5 ,
wherein the prescribed units correspond to a string in which a representation of a word or a plurality of words is normalized.
9. An alignment generation method comprising:
generating, by a processor, a first plurality of encoded sentences by encoding, in accordance with prescribed units, a plurality of sentences in a first language;
generating, by the processor, a second plurality of encoded sentences by encoding, in accordance with the prescribed units, a plurality of sentences in a second language, each of which is associated with each of the plurality of sentences in the first language; and
generating, by the processor, relationship information indicating relationships between a first plurality of codes in the first plurality of encoded sentences and a second plurality of codes in the second plurality of encoded sentences, in accordance with a number of combinations of a first code included in a third plurality of codes and a second code included in a fourth plurality of codes, the combinations of the first code and the second code being included in a first plurality of combinations of each of the third plurality of codes and each of the fourth plurality of codes, the first plurality of combinations being generated for each of a second plurality of combinations generated by selecting each of the first plurality of encoded sentences and each of the second plurality of encoded sentences, the third plurality of codes being included in the selected first encoded sentence and the fourth plurality of codes being included in the selected second encoded sentence.
10. The alignment generation method according to claim 9 ,
wherein the first code is associated with the second code.
11. The alignment generation method according to claim 9 ,
wherein the encoding of the plurality of sentences in the first language includes generating first index information having a logical value indicating whether each of the first plurality of codes is included in each of the first plurality of encoded sentences,
the encoding of the plurality of sentences in the second language includes generating second index information having a logical value indicating whether each of the second plurality of codes is included in each of the second plurality of encoded sentences, and
the generating of the relationship information is executed in accordance with the first index information and the second index information.
12. The alignment generation method according to claim 9 ,
wherein the prescribed units correspond to a string in which a representation of a word or a plurality of words is normalized.
13. A non-transitory computer-readable recording medium having stored therein a translation program that causes a computer to execute a process comprising:
generating first encoded text data by encoding, in accordance with prescribed units, text data in a first language;
translating the first encoded text data into second encoded text data in accordance with relationship information indicating relationships between a first plurality of codes in a first plurality of encoded sentences generated by encoding, in accordance with the prescribed units, a plurality of sentences in the first language and a second plurality of codes in a second plurality of encoded sentences generated by encoding, in accordance with the prescribed units, a plurality of sentences in a second language; and
generating text data in the second language by decoding the second encoded text data,
wherein the relationship information is generated in accordance with a number of combinations of a first code included in a third plurality of codes and a second code included in a fourth plurality of codes, the combinations of the first code and the second code being included in a first plurality of combinations of each of the third plurality of codes and each of the fourth plurality of codes, the first plurality of combinations being generated for each of a second plurality of combinations generated by selecting each of the first plurality of encoded sentences and each of the second plurality of encoded sentences, the third plurality of codes being included in the selected first encoded sentence and the fourth plurality of codes being included in the selected second encoded sentence.