Event naming
First Claim
1. A machine-implemented method comprising:
- identifying an event for a particular category based on a first quantity of documents classified as relevant to the particular category in a first period of time compared with a second quantity of documents relevant to the particular category in a previous, second period of time;
identifying a first set of key words in the first quantity of documents and a second set of key words in the second quantity of documents;
based on a comparison of the first set of key words to the second set of key words, identifying a set of event keywords more prevalent in the documents classified as relevant to the particular category in the first period of time than in the documents classified as relevant to the particular category in the second period of time;
calculating scores for each of the first quantity of documents based on appearances of the event keywords in the documents;
automatically selecting a highest scoring document as a representative document for the event and identifying a link to the representative document;
automatically determining whether the representative document is still accessible at the identified link; and
when the representative document is no longer accessible at the identified link, replacing the highest scoring document with a next highest scoring document as the representative document for the event.
5 Assignments
0 Petitions
Accused Products
Abstract
Some embodiments provide a machine-implemented method. The method identifies an event for a particular category based on a number of documents classified as relevant to the particular category in a particular period of time. Based on content of the documents classified as relevant to the particular category, the method identifies a set of keywords for the event. The method uses the keywords to automatically select a representative document for the event. Some embodiments store a link to the representative document and automatically determine whether the particular document is still accessible at the link. When the document is no longer accessible at the link, the method replaces the document with a backup document as the representative document for the event.
114 Citations
17 Claims
-
1. A machine-implemented method comprising:
-
identifying an event for a particular category based on a first quantity of documents classified as relevant to the particular category in a first period of time compared with a second quantity of documents relevant to the particular category in a previous, second period of time; identifying a first set of key words in the first quantity of documents and a second set of key words in the second quantity of documents; based on a comparison of the first set of key words to the second set of key words, identifying a set of event keywords more prevalent in the documents classified as relevant to the particular category in the first period of time than in the documents classified as relevant to the particular category in the second period of time; calculating scores for each of the first quantity of documents based on appearances of the event keywords in the documents; automatically selecting a highest scoring document as a representative document for the event and identifying a link to the representative document; automatically determining whether the representative document is still accessible at the identified link; and when the representative document is no longer accessible at the identified link, replacing the highest scoring document with a next highest scoring document as the representative document for the event. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A machine readable medium storing a program which when executed by at least one processor maintains a stored link to a document, the program comprising sets of instructions for:
-
identifying an event for a particular category based on a first quantity of documents classified as relevant to the particular category in a first period of time compared with a second quantity of documents relevant to the particular category in a previous, second period of time; identifying a first set of key words in the first quantity of documents and a second set of key words in the second quantity of documents; based on a comparison of the first set of key words to the second set of key words, identifying a set of event keywords more prevalent in the documents classified as relevant to the particular category in the first period of time than in the documents classified as relevant to the particular category in the second period of time; calculating scores for each of the first quantity of documents based on appearances of the event keywords in the documents; automatically selecting a highest scoring document as a representative document for the event and identifying a link to the representative document; automatically determining whether the representative document is still accessible at the identified link; and when the representative document is no longer accessible at the identified link, replacing the highest scoring document with a next highest scoring document as the representative document for the event. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17)
-
Specification