Support device, support method and support program
AssigneeNippon Telegraph and Telephone Corporation
InventorTakeshi Yamada
Life of this patent
Estimating the term…
Abstract
In assistance processing for assisting in describing a document corresponding to a portion in a document, a calculation unit calculates a similarity degree between a sentence in a document and a document described correspondingly to a portion in the document. When the calculated similarity degree is equal to or greater than a predetermined threshold, an extraction unit extracts the sentence in the document in association with the document described correspondingly to the portion in the document.
Claims6
An assistance apparatus comprising one or more processors configured to:
- identify a portion of a document and impart a tag to the identified portion;
- extract one or more test items from the portion to which the tag has been imparted;
- calculate a similarity degree between i) each sentence included in the document and ii) each of the one or more test items;
- obtain a set of extracted sentences that are relevant sentences of the one or more test items comprising:
- for each sentence included in the document:
- determine that the similarity degree between the sentence and a corresponding test item included in the one or more test items is equal to or greater than a predetermined threshold;
- and
- extract the sentence in the document as a relevant sentence of the corresponding test item;
- identify, from the set of extracted sentences, a plurality of sentences whose similarity degrees to a same test item are equal to or greater than the predetermined threshold;
- group the plurality of sentences as relevant sentences of the same test item;
- and
- extract a second test item from the grouped plurality of sentences.
An assistance method performed in an assistance apparatus for assisting in describing a document corresponding to a portion in a document, the assistance method comprising:
- identifying a portion of a document and impart a tag to the identified portion;
- extracting one or more test items from the portion to which the tag has been imparted;
- calculating a similarity degree between i) each sentence included in the document and ii) each of the one or more test items;
- obtaining a set of extracted sentences that are relevant sentences of the one or more test items comprising:
- for each sentence included in the document:
- determining that the similarity degree between the sentence and a corresponding test item included in the one or more test items is equal to or greater than a predetermined threshold;
- and
- extracting the sentence in the document as a relevant sentence of the corresponding test item;
- identifying, from the set of extracted sentences, a plurality of sentences whose similarity degrees to a same test item are equal to or greater than the predetermined threshold;
- grouping the plurality of sentences as relevant sentences of the same test item;
- and
- extracting a second test item from the grouped plurality of sentences.
A non-transitory computer readable medium storing one or more instructions causing, in processing for assisting in describing a document corresponding to a portion in a document, a computer to execute:
- identifying a portion of a document and impart a tag to the identified portion;
- extracting one or more test items from the portion to which the tag has been imparted;
- calculating a similarity degree between i) each sentence included in the document and ii) each of the one or more test items;
- obtaining a set of extracted sentences that are relevant sentences of the one or more test items comprising:
- for each sentence included in the document:
- determining that the similarity degree between the sentence and a corresponding test item included in the one or more test items is equal to or greater than a predetermined threshold;
- and
- extracting the sentence in the document as a relevant sentence of the corresponding test item;
- identifying, from the set of extracted sentences, a plurality of sentences whose similarity degrees to a same test item are equal to or greater than the predetermined threshold;
- grouping the plurality of sentences as relevant sentences of the same test item;
- and
- extracting a second test item from the grouped plurality of sentences.
Dependencies are inferred from the claim text.