Methods for automatically generating natural-language news items from log files and status traces

• US 8,090,727 B2
• Filed: 05/20/2008
• Issued: 01/03/2012
• Est. Priority Date: 05/20/2008
• Status: Active Grant

• Current Assignee: Check Point Software Technologies Limited
• Original Assignee: Check Point Software Technologies Limited
• Inventors: Lachtarnik, Marina, Raviv, Gil
• Primary Examiner(s): THAI, HANH B
• Application Number:
  US12/123,544
• Publication Number:
  US 20090292719A1
• Time in Patent Office: 1,323 Days
• Field of Search: 707/736, 707/754, 707/804, 704/9, 704/257
• US Class Current: 707/754
• CPC Class Codes: G06F 40/237 Lexical tools

Abstract

Methods, for automatically generating natural-language news items from log files, including the steps of: gathering at least one data record; filtering at least one data record according to at least one rule to produce at least one filtered data set; aggregating at least one filtered data set; analyzing at least one filtered data set for at least one statistical trend; and automatically generating a news item based on at least one statistical trend. Preferably, the method further includes the step of: customizing the news item based on a relative importance of at least one statistical trend. Preferably, the method further includes the step of: performing a drill-down analysis on at least one statistical trend. Most preferably, the method further includes the step of: enriching the news item based on the drill-down analysis. Preferably, the method further includes the step of: embedding at least one graphical element into the news item.

Claims

1. A method for automatically generating a natural-language news item, the method comprising the steps of:

(a) gathering at least one data record;

(b) filtering said at least one data record according to at least one rule to produce at least one filtered data set;

(c) aggregating said at least one filtered data set, in a storage device;

(d) analyzing, by a processor, said at least one filtered data set for at least one statistical trend; and

(e) automatically generating, by said processor, a news item based on said at least one statistical trend, wherein the automatically generating includes;

(i) selecting a news-item template from a plurality of templates;

(ii) selecting at least one keyword based on said at least one statistical trend; and

(iii) inserting said at least one keyword into said news-item template.

2. The method of claim 1, the method further comprising the step of:

(f) customizing said news item based on a relative importance of said at least one statistical trend.

3. The method of claim 2, wherein said relative importance is based on a history of said at least one statistical trend.

4. The method of claim 1, the method further comprising the step of:

(f) performing a drill-down analysis on said at least one statistical trend.

5. The method of claim 4, the method further comprising the step of:

(g) enriching said news item based on said drill-down analysis.

6. The method of claim 1, the method further comprising the step of:

(f) embedding at least one graphical element into said news item.

7. The method of claim 1, the method further comprising the step of:

(f) assessing said news item for a relative ranking.

8. The method of claim 7, wherein said relative ranking is based on a record of previous news items.

9. The method of claim 1, wherein said step of automatically generating further includes:

(ii) obtaining at least one data parameter from said at least one statistical trend; and

(iii) inserting said at least one data parameter into said news-item template.

10. The method of claim 1, wherein said selecting of said at least one keyword further includes selecting said at least one keyword from a list of synonymous keywords.

11. The method of claim 10, wherein said selecting from said list of synonymous keywords is random.

12. The method of claim 10, wherein said selecting from said list of synonymous keywords is based on keyword usage in a record of previous news items.

13. The method of claim 1, wherein said selecting is based on a relative severity of said at least one statistical trend.

14. The method of claim 13, wherein said relative severity is based on a history of said at least one statistical trend.

15. The method of claim 1, wherein said step of analyzing includes performing at least one numerical operation on said at least one filtered data set.