Auto-summary generator and filter

• US 8,108,398 B2
• Filed: 06/29/2007
• Issued: 01/31/2012
• Est. Priority Date: 06/29/2007
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Guday, Shai, O'Rourke, Bret P., Miller, John Mark, Alkove, James Morris, Wilson, Andrew David
• Primary Examiner(s): Rones, Charles
• Assistant Examiner(s): Chojnacki, Mellissa M
• Application Number:
  US11/771,445
• Publication Number:
  US 20090006369A1
• Time in Patent Office: 1,677 Days
• Field of Search: 707/737, 707/739
• US Class Current: 707/739
• CPC Class Codes: G06F 16/345 Summarisation for human users

Abstract

A system that facilitates data presentation and management includes at least one database to store a corpus of data relating to one or more topics. The system further includes a summarizer component to automatically determine a subset of the data over the corpus of data relating to at least one of the topic(s), wherein the subset forms a summary of at least one topic.

Claims

1. A system that facilitates data presentation and management, comprising:

one or more processors;

a memory that includes components that are executable by the one or more processors, the components including;

at least one database to store a corpus of data relating to one or more topics; and

a summarizer component to;

automatically determine a subset of the data in the corpus of data relating to at least one of the one or more topics;

recognize one or more words in the corpus of data that indicates the corpus of data includes a previously generated summary; and

form the subset of data into a summary of the at least one topic, wherein words of the previously generated summary are given more weight during formation of the summary than other words in the subset of data.

2. The system of claim 1, further comprising a profile component to control data generated by the summarizer component.

3. The system of claim 2, the profile component includes a user type component to determine whether a user is a novice user or an experienced user, a recreational component to determine recreational activity preferences of the user, an artistic component to determine artistic interests of the user, a media component to determine types of media that are included in the summary, a time component to determine when the summary is to be generated, a calendar component to provide calendar and time data to the time component, a geographical component to determine an amount of detail to include in the summary based on a geographical location of the user, or a general settings component to control additional information included in the summary.

4. The system of claim 1, further comprising a filter component to control data generated by the summarizer component.

5. The system of claim 4, the filter component includes a language component, a selection component, a policy component, or a rules component.

6. The system of claim 1, the summarizer component further comprises a word clues component to locate one or more clues, a word snippets component to analyze a sequence of words, a key word component to focus an analysis near or within a proximity of a key word, a learning component to train the summarizer component to analyze the corpus of data, a profile component to format data generated by the summarizer component according to a user profile, a filter component to filter data generated by the summarizer component according to one or more filtering criteria, or a statistical component to monitor a frequency that a keyword occurs or a distance of a word to other words.

7. The system of claim 1, further comprising one or more controls to operate with the summarizer component.

8. The system of claim 7, the controls include a length component to regulate a length of the summary, a preferences component to regulate a format of the summary, a processing time component to regulate an amount of processing time the summarizer uses to analyze the corpus of data and a speed at which the summary is generated, a thumbnail component to regulate a presentation of the summary, a machine translation component to translate the summary from a first language to a second language, and a learning constraint component to regulate an algorithm use by the summarizer to generate the summary.

9. The system of claim 1, the summarizer component operates on mixed media data to generate one or more summaries.

10. The system of claim 9, the mixed media data is associated with text data, numeric data, audio data, image data, or video data.

11. The system of claim 9, the summarizer component generates summarized data having at least two forms of mixed media.

12. The system of claim 1, further comprising an input component to capture data from a location.

13. The system of claim 12, further comprising a component to automatically summarize the data from the location.

14. The system of claim 13, further comprising a component to store summarized data as an annotation from the location.

15. The system of claim 1, further comprising a component to filter or summarize photographic libraries which are tailored to the user'"'"'s current interests.

16. The system of claim 15, the user'"'"'s current interests are determined from a user profile or by a component that monitors user activities over time.

17. The system of claim 1, further comprising a component concatenate summarized data from at least two topics.

18. A method to generate summary data for a user, comprising:

one or more processors;

reading data from a database that includes a corpus of data;

recognize one or more words in the corpus of data that indicates the corpus of data includes a previously generated summary;

determining an amount of time available to process the corpus of data for summary generation based on a user preference; and

automatically summarizing the corpus of data into at least one subset of data to generate a summary such that an accuracy of detail in the summary is proportional to the amount of time available to process the corpus of data, wherein words included in the previously generated summary are given more weight during generation of the summary than other words in the at least one subset of data.

19. The method of claim 18, further comprising employing a filter or a user profile to summarize the corpus of data.