Method for searching a file having a format unsupported by a search engine

• US 6,327,589 B1
• Filed: 06/24/1998
• Issued: 12/04/2001
• Est. Priority Date: 06/24/1998
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Blewett, William John, Hawley, Eric, Gerlt, David C., Moore, Charles G.
• Primary Examiner(s): HOMERE, JEAN RAYMOND
• Application Number:
  US09/105,123
• Time in Patent Office: 1,259 Days
• Field of Search: 707/3, 707/5, 707/102, 707/513, 707/531, 704/9
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

Searching a file in a format unsupported by a search engine by creating term-topic links with associated probabilities. A file is retrieved comprising a compressed HTML file or a webpage. The file is parsed to retrieve data associated with title tags and body tags. In addition, user queries are received so that the user may associate a query with the title data. Term-topic links are created by linking terms from the retrieved data and the query with a topic. Heuristics are then used to determine the probability associated with each term-topic link. Term-topic links having a term containing nouns are assigned a higher probability than verbs, verbs are assigned a higher probability than adjectives, and adjectives and adverbs are assigned the same probability. The term-topic links are trained by adjusting the assigned probabilities based on a user defined query and an associated target topic.

Claims

1. A method of searching a file by using terms in the file to find related topics in the file, comprising the steps of:

(a) receiving the file to be searched;

(b) parsing the file into selected ones of the terms related to a certain one of the topics;

(c) creating term-topic links for the selected terms, wherein each term-topic link associates one of the selected terms to the certain topic; and

(d) if a selected term matches a term in an existing term-topic link,

creating a term-topic list by associating the matching term and related terms with a term-topic link, wherein each term-topic list is related to a term-topic link.

2. The method of claim 1, further comprising the step of repeating steps (b) and (c) for each remaining one of the topics.

3. The method of claim 1, further comprising the steps of:

identifying synonyms for each selected term; and

creating term-topic links for each of the identified synonyms by associating each identified synonym with the certain topic.

4. The method of claim 1, wherein the file is a compressed HTML file and prior to performing the step (b), further comprising the step of decompressing the compressed HTML file.

5. The method of claim 1, wherein the file is a path with an associated address of a webpage, and the step of parsing the file comprises the steps of:

retrieving the webpage; and

parsing the webpage into selected ones of the terms related to the certain topic.

6. The method of claim 1, further comprising the step of assigning a probability to each term-topic link.

7. The method of claim 6, wherein the step of assigning the probability is performed by speech heuristics on the term in each term-topic link.

8. The method of claim 7, wherein the step of assigning the probability by speech heuristics comprises the steps of:

assigning nouns a higher probability than verbs;

assigning verbs a higher probability then adjectives; and

assigning adjectives and adverbs the same probability.

9. The method of claim 1, prior to performing the step of creating term-topic links, further comprising the steps of:

identifying stop-terms, wherein stop-terms comprise a predefined set of common words; and

removing the identified stop-terms from the selected terms.

10. The method of claim 1, wherein the file is a query.

11. A method of searching a file, comprising the steps of:

receiving the file to be searched;

identifying title tags and body tags contained in the file;

identifying terms contained in data associated with the title tags and in data associated with the body tags, wherein the data associated with the title tag is called a topic of the file;

creating term-topic links by associating each identified term with the topic, wherein each identified term is related to the topics and

if a term matches a term in an existing term-topic link,

creating a term-topic list by associating the matching term and related terms with a term-topic link, wherein each term-topic list is related to a term-topic link.

12. The method of claim 11, wherein the file is a compressed HTML file and prior to performing the step of identifying the title tags and body tags, further comprising the step of decompressing the compressed HTML file.

13. The method of claim 11, wherein the file is an paths and its associated address of a webpage, and the step of identifying title tags and body tags in the file comprises the steps of:

retrieving the webpage; and

identifying title tags and body tags contained in the webpage.

14. The method of claim 11, further comprising the steps of:

identifying synonyms for each identified term; and

creating term-topic links for each of the identified synonyms by associating each identified synonym with the certain topic.

15. The method of claim 11, further comprising the step of assigning a probability to each term-topic link.

16. The method of claim 15, wherein the step of assigning the probability is performed by speech heuristics on the term in each term-topic link.

17. The method of claim 16, wherein the step of assigning the probability by speech heuristics comprises the steps of:

assigning nouns a higher probability than verbs;

assigning verbs a higher probability then adjectives; and

assigning adjectives and adverbs the same probability.

18. The method of claim 11, prior to performing the step of creating term-topic links, further comprising the steps of:

identifying stop-terms, wherein stop-terms comprise a predefined set of common words; and

and removing the identified stop-terms from the identified terms.

19. The method of claim 11, further comprising the steps of:

receiving a query;

identifying terms contained in the query; and

creating term-topic links by associating each identified term of the query with the certain topic, wherein the identified terms of the query are related to the topics.

20. A computer readable medium having computer-executable instructions for performing the steps recited in claim 11.

21. A method of searching a help file, comprising the steps of:

receiving the help file containing help information to be searched;

identifying title tags and body tags contained in the help file;

identifying a certain one of a plurality of topics of the help file comprising data associated with the title tag;

receiving a first query;

identifying terms of the help file contained in the certain topic, in data associated with the body tag and in the query;

creating term-topic links by associating each identified term with the certain topic, wherein the identified terms are related to the certain topic;

associating a probability to each term-topic link;

receiving a second query and a target topic; and

training the term-topic links by adjusting the probability of term-topic links to reflect an association between the second query and the target topic.

22. The method of claim 21, wherein the help file is a compressed file and prior to performing the step of identifying the title tags and the body tags, further comprising the step of decompressing the help file.

23. The method of claim 21, wherein the help file is a path and its associated address of a webpage containing help information and wherein the step of identifying title tags and body tags in the help file comprises the steps of:

retrieving the webpage using a browser; and

identifying title tags and body tags contained in the webpage.

24. The method of claim 21, prior to performing the step of training the term-topic links, further comprising the steps of:

identifying synonyms for each identified term; and

creating term-topic links for each of the identified synonyms by associating each identified synonym with the certain topic.

25. The method of claim 21, further comprising the step of assigning a probability to each term-topic link.

26. The method of claim 25, wherein the step of assigning the probability is performed by speech heuristics on the term in each term-topic link.

27. The method of claim 26, wherein the step of assigning the probability by speech heuristics comprises the steps of:

assigning nouns a higher probability than verbs;

assigning verbs a higher probability then adjectives; and

assigning adjectives and adverbs the same probability.

28. The method of claim 21, prior to performing the step of creating term-topic links, further comprising the steps of:

identifying stop-terms, wherein stop-terms comprise a predefined set of common words; and

removing the identified stop-terms from the selected terms.

29. A method of training term-topic links by refining the probabilities of term-topic links to reflect an association between a given query and a target topic, comprising the steps of:

receiving the query and the target topic, wherein the target topic is related to the query;

identifying a first selected ones of terms contained in the query;

retrieving topics linked to each one of the first selected terms in the term-topic links, wherein the target topic will match one of the retrieved topics;

for each retrieved topic, assigning a combined probability to each retrieved topic by combining the probabilities associated with each one of a second selected ones of terms contained in the retrieved topic;

identifying a predefined number of retrieved topics based on the sorted combined probabilities; and

in the event the target topic is not above a predefined threshold of retrieved topics, decreasing the combined probability of one of the retrieved topics having a higher combined probability than the target topic and increasing the combined probability of the target topic.

30. The method of claim 29, wherein the step of increasing the combined probability of the target topic and decreasing the combined probability of a one of the retrieved topics, comprises the steps of:

decreasing the probability associated with a term contained in one of the retrieved topics having a higher combined probability than the target topic;

retrieving the terms associated with the target topic by using the term-topic links; and

increasing the associated probabilities of each retrieved term.

31. The method of claim 29, wherein the step of decreasing the probability associated with a term contained in a randomly selected retrieved topic is performed the number of times equal to the number of terms identified in the target topic.

32. A computer readable medium having computer-executable instructions for performing the steps recited in claim 29.

33. A method of searching a file by using terms in the file to find related topics in the file, comprising the steps of:

(a) receiving the file to be searched;

(b) parsing the file into selected ones of the terms related to a certain one of the topics;

(c) creating term-topic links for the selected terms, wherein each term-topic link associates one of the selected terms to the certain topic; and

(d) assigning a probability to each of the term-topic links by performing speech heuristics on the term in each term-topic link, wherein the step of assigning the probability by speech heuristics comprises the steps of;

assigning nouns a higher probability than verbs,

assigning verbs a higher probability then adjectives, and

assigning adjectives and adverbs the same probability.

34. A method of searching a file, comprising the steps of:

receiving the file to be searched;

identifying title tags and body tags contained in the file;

identifying terms contained in data associated with the title tags and in data associated with the body tags, wherein the data associated with the title tag is called a topic of the file;

creating term-topic links by associating each identified term with the topic, wherein each identified term is related to the topics;

and

assigning a probability to each of the term-topic links by performing speech heuristics on the term in each term-topic link, wherein the step of assigning the probability by speech heuristics comprises the steps of;

assigning nouns a higher probability than verbs,

assigning verbs a higher probability then adjectives, and

assigning adjectives and adverbs the same probability.