Systems and methods for quote extraction
View Patent ↗Embodiments of the present disclosure may be utilized to analyze a content item comprising text to identity: a quote, a named entity that is the source of the quote (e.g., a person or organization such as a company), and identification information for the named entity (such as a title of the person giving the quote). Quotes may also be ranked to determine, for example, the best quotes to include in an article.
1. A method comprising:
receiving, by a processor, content items, each of the content items associated with a topic;
analyzing, by the processor, each content item to identify a plurality of quotes and to identify a named entity for each of the quotes, each named entity associated with a respective quote being a source of the respective quote;
extracting, by the processor, keywords in each quote;
ranking, by the processor, the quotes based on a relevance of the keywords to the topic of the content items;
determining, by the processor, a subset of the quotes based on the ranking;
generating, by the processor, the article using the subset of quotes and the content items, the generating the article comprising:
generating a summary page for the article, the summary page including a subset of the content items, at least one selected quote from the subset of quotes, and a link to one or more detail pages, the summary page displaying text of the selected quote and a named entity associated with the selected quote,
generating the one or more detail pages, the detail pages including a remainder of the content items not included in the subset of the content items, and
combining the summary page and the one or more detail pages to form the article, the combining comprising arranging the summary page for initial display upon receipt of the article; and
transmitting, by the processor, the article to a client device for display, the display comprising displaying the summary page upon receipt of the article and displaying the one or more detail pages upon detecting a selection of the link to one or more detail pages or detecting a scrolling of the article.
2. The method of claim 1 , wherein the analyzing of each content item further comprises identifying identification information for each named entity, wherein the identification information for each named entity includes first information that is descriptive of the respective named entity and second information that is descriptive of the first information.
3. The method of claim 2 , wherein the identification information for a named entity includes a geographical location associated with the corresponding named entity.
4. The method of claim 1 , wherein the ranking of the quotes is further based on a length of each quote.
5. The method of claim 1 , wherein the ranking of the quotes is further based on a respective position of each quote within a content item.
6. The method of claim 1 , wherein the ranking of the quotes is further based on type of words used in each quote.
7. The method of claim 1 , wherein the identifying of the named entity further comprises performing a coreference resolution with respect to content in the content item.
8. The method of claim 1 , wherein the ranking of a quote in the quotes is further based on information available in the content item for a named entity associated with the quote.
9. The method of claim 8 , wherein at least a portion of the information available for the named entity associated with the quote is retrieved from sources other than the content item containing the quote.
10. A computer system comprising:
a processor;
a display;
a storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising:
content item receiving logic executed by the processor for receiving content items, each of the content items associated with a topic;
content item analyzing logic executed by the processor for analyzing each content item to identify a plurality of quotes and to identify a named entity for each of the quotes, each named entity associated with a respective quote being a source of the respective quote;
keyword analyzing logic executed by the processor for extracting keywords in each quote;
quote ranking logic executed by the processor for ranking the quotes based on a relevance of the keywords to the topic of the content items;
quote determining logic executed by the processor for determining a subset of the quotes based on the ranking;
article generating logic executed by the processor for generating the article using the subset of quotes and the content items, the generating the article comprising:
generating a summary page for the article, the summary page including a subset of the content items, at least one selected quote from the subset of quotes, and a link to one or more detail pages, the summary page displaying text of the selected quote and a named entity associated with the selected quote,
generating the one or more detail pages, the detail pages including a remainder of the content items not included in the subset of the content items, and
combining the summary page and the one or more detail pages to form the article, the combining comprising arranging the summary page for initial display upon receipt of the article; and
article transmitting logic executed by the processor for transmitting the article to a client device for display, the display comprising displaying the summary page upon receipt of the article and displaying the one or more detail pages upon detecting a selection of the link to one or more detail pages or detecting a scrolling of the article.
11. The computer system of claim 10 , wherein the content item analyzing logic for analyzing each content item further comprises identification information identifying logic executed by the processor for identifying identification information for each named entity, wherein the identification information for each named entity includes first information that is descriptive of the respective named entity and second information that is descriptive of the first information.
12. The computer system of claim 11 , wherein the identification information for a named entity includes a geographical location associated with the corresponding named entity.
13. The computer system of claim 10 , wherein the quote ranking logic further ranks the quotes based on a length of each quote.
14. The computer system of claim 10 , wherein the quote ranking logic further ranks the quotes based on a respective position of each quote within a content item.
15. The computer system of claim 10 , wherein the quote ranking logic further ranks the quotes based on type of words used in each quote.
16. The computer system of claim 10 , wherein the content item analyzing logic for identifying the named entity further comprises performing a coreference resolution with respect to content in the content item.
17. The computer system of claim 10 , wherein the quote ranking logic further ranks a quote in the quotes based on information available in the content item for a named entity associated with the quote.
18. The computer system of claim 17 , wherein at least a portion of the information available for the named entity associated with the quote is retrieved from sources other than the content item containing the quote.
19. A non-transitory computer readable storage medium tangibly storing computer program instructions capable of being executed by a processor of a computer system, the computer program instructions defining the steps of:
receiving, by the processor, content items, each of the content items associated with a topic;
analyzing, by the processor, each content item to identify a plurality of quotes and to identify a named entity for each of the quotes, each named entity associated with a respective quote being a source of the respective quote;
extracting, by the processor, keywords in each quote;
ranking, by the processor, the quotes based on a relevance of the keywords to the topic of the article;
determining, by the processor, a subset of the quotes based on the ranking;
generating, by the processor, the article using the subset of quotes and the content items, the generating the article comprising:
generating a summary page for the article, the summary page including a subset of the content items, at least one selected quote from the subset of quotes, and a link to one or more detail pages, the summary page displaying text of the selected quote and a named entity associated with the selected quote,
generating the one or more detail pages, the detail pages including a remainder of the content items not included in the subset of the content items, and
combining the summary page and the one or more detail pages to form the article, the combining comprising arranging the summary page for initial display upon receipt of the article; and
transmitting, by the processor, the article to a client device for display, the display comprising displaying the summary page upon receipt of the article and displaying the one or more detail pages upon detecting a selection of the link to one or more detail pages or detecting a scrolling of the article.
20. The non-transitory computer readable storage medium of claim 19 , wherein the analyzing of each content item further comprises identifying identification information for each named entity, wherein the identification information for each named entity includes first information that is descriptive of the respective named entity and second information that is descriptive of the first information.