Keyword outputting apparatus and method
A keyword analysis device obtains word vectors represented by the documents by analyzing keywords contained in each of documents input in a designated period. A topic cluster extraction device extracts topic clusters belonging to the same topic from a plurality of documents. A keyword extraction device extracts, as a characteristic keyword group, a predetermined number of keywords from the topic cluster in descending order of appearance frequency. A topic structurization determination device determines whether the topic can be structurized, by segmenting the topic cluster into subtopic clusters with reference to the number of documents, the variance of dates contained in the documents, or the C-value of keyword contained in the documents, as a determination criterion. And a keyword presentation device presents the characteristic keyword group in the subtopic cluster upon arranging the keyword group on the basis of the date information.
1. A keyword outputting apparatus comprising:
a document input device configured to input a plurality of documents having date information in a designated period;
a keyword analysis device configured to obtain a plurality of word vectors represented by the documents by analyzing a plurality of keywords contained in each of the documents;
a topic cluster extraction device configured to extract a pair of documents as a topic cluster from said documents, the pair of documents having a maximum inner product of the corresponding word vectors and belonging to an identical topic;
a keyword extraction device configured to extract a predetermined number of keywords as a characteristic keyword group from the topic cluster in descending order of appearance frequency;
a topic structurization determination device configured to determine whether the topic is configured to be structurized, by segmenting the topic cluster into subtopic clusters with reference to one of a number of the documents, a variance of dates contained in the documents, and a C-value of keyword in the documents, as a determination criterion;
a control device configured to control the topic cluster extraction device to repeatedly extract a subtopic cluster until it is not determined that a subtopic of the subtopic cluster is not configured to be structurized, and to control the keyword extraction device to extract another characteristic keyword group from the extracted subtopic cluster; and
a keyword presentation device configured to present the characteristic keyword group in the subtopic cluster upon arranging the keyword group on the basis of the date information.
2. An apparatus according to claim 1 , further comprising a keyword structurization device configured to structurize keywords by classifying or hierarchizing keyword groups of the topic cluster and the subtopic cluster.
3. An apparatus according to claim 2 , wherein the keyword structurization device classifies and hierarchizes the keyword group by using cooccurrence information based on commonality of an appearance source documents of keywords.
4. An apparatus according to claim 1 , further comprising a genre estimation device configured to estimate a genre of one of the topic cluster and the subtopic cluster and attach a genre name thereto.
5. An apparatus according to claim 1 , further comprising a search query generating device configured to generate a search query on the basis of a keyword group of one of the topic cluster and the subtopic cluster.
6. An apparatus according to claim 5 , wherein when a keyword is selected from a keyword group presented by the keyword presentation device, the search query is assigned to the keyword.
7. An apparatus according to claim 5 , which further comprises a device configured to structurize keywords by classifying or hierarchizing keyword groups of the topic cluster and the subtopic cluster, and in which
the search query generating device generates a search query by combining a plurality of keywords by using a structurization result of the keywords.
8. An apparatus according to claim 5 , wherein the search query generating device attaches a pre-designated specific keyword string.
9. An apparatus according to claim 5 , which further comprises a genre estimation device configured to estimate a genre of one of the topic cluster and the subtopic cluster and assign a genre name thereto, and in which
the search query generating device switches search query generating methods in accordance with the genre name estimated by the genre estimation device.
10. An apparatus according to claim 1 , further comprising a keyword storage device configured to store a keyword group extracted by the keyword extraction device together with a score of each keyword.
11. An apparatus according to claim 10 , wherein the topic cluster extraction device calculates scores at the time of extraction of the topic cluster by also using a history of scores stored in the keyword storage device.
12. An apparatus according to claim 10 , wherein an attribute is assigned to said each keyword by comparing scores of the respective keywords stored in the keyword storage device.
13. An apparatus according to claim 12 , wherein in accordance with one of presence/absence and a type of the attribute, the keyword presentation device displays a mark which makes the attribute identifiable.
14. An apparatus according to claim 12 , wherein the keyword presentation device changes a display order of the keyword group in accordance with one of presence/absence and a type of the attribute.
15. An apparatus according to claim 10 , wherein an attribute indicating that a keyword has newly arrived is assigned to the newly-arrived keyword determined by comparing with a keyword stored in the keyword storage device.
16. An apparatus according to claim 1 , wherein a document input to the document input device is a news article of a Web page on the Internet.
17. An apparatus according to claim 1 , wherein a document input to the document input device is information contained in an electronic program guide (EPG).
18. An apparatus according to claim 1 , wherein a document set contained in at least one designated period is input to the document input device.
19. An apparatus according to claim 1 , wherein the keyword extraction device calculates a score based on an attribute of a keyword.
20. An apparatus according to claim 19 , wherein the score is an appearance frequency of the keyword.
21. An apparatus according to claim 1 , wherein the keyword extraction device extracts the keyword group by using an inclusion relation between keywords.
22. An apparatus according to claim 1 , wherein the keyword presentation device comprises a display device which displays a document containing the keyword in response to predetermined operation for a keyword.
23. A non-transitory computer readable storage medium storing instructions of a computer program which when executed by a computer results in performance of steps comprising:
inputting a plurality of documents having date information in a designated period;
obtaining a plurality of word vectors represented by the documents by analyzing a plurality of keywords contained in each of the documents;
extracting a pair of documents as a topic cluster from said documents, the pair of documents having a maximum inner product of the corresponding word vectors and belonging to an identical topic;
extracting a predetermined number of keywords as a characteristic keyword group from the topic cluster in descending order of appearance frequency;
determining whether the topic is configured to be structurized, by segmenting the topic cluster into subtopic clusters with reference to one of a number of the documents, a variance of dates contained in the documents, and a C-value of keyword in the documents, as a determination criterion;
performing control to repeatedly extract a subtopic cluster until it is not determined that a subtopic of the subtopic cluster is not configured to be structurized, and to extract another characteristic keyword group from the extracted subtopic cluster; and
presenting the characteristic keyword group in the subtopic cluster upon arranging the keyword group on the basis of the date information.
24. A keyword outputting method comprising:
inputting a plurality of documents having date information in a designated period;
obtaining a plurality of word vectors represented by the documents by analyzing a plurality of keywords contained in each of the documents;
extracting a pair of documents as a topic cluster from said documents, the pair of documents having a maximum inner product of the corresponding word vectors and belonging to an identical topic;
extracting a predetermined number of keywords as a characteristic keyword group from the topic cluster in descending order of appearance frequency;
determining whether the topic is configured to be structurized, by segmenting the topic cluster into subtopic clusters with reference to one of a number of the documents, a variance of dates contained in the documents, and a C-value of keyword in the documents, as a determination criterion;
performing control to repeatedly extract a subtopic cluster until it is not determined that a subtopic of the subtopic cluster is not configured to be structurized, and to extract another characteristic keyword group from the extracted subtopic cluster; and
presenting the characteristic keyword group in the subtopic cluster upon arranging the keyword group on the basis of the date information.