Computer knowledge representation format, system, methods, and applications

• US 7,860,815 B1
• Filed: 02/07/2007
• Issued: 12/28/2010
• Est. Priority Date: 07/12/2006
• Status: Active Grant

• Current Assignee: Venkateswara Prasad Tangirala
• Original Assignee: Venkateswara Prasad Tangirala
• Inventors: Tangirala, Venkateswara Prasad
• Primary Examiner(s): Holmes; Michael B.
• Assistant Examiner(s): Kim; David H
• Application Number:
  US11/703,975
• Time in Patent Office: 1,420 Days
• Field of Search: 706/45, 707/790, 707/791, 707/792, 704/1, 704/9
• US Class Current: 706/45
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation

Abstract

A near natural language format is used to codify knowledge (including information, programs, rules, and procedures). The codified knowledge is stored in knowledge snippet files according to a nomenclature in which each file has a name relating to a particular interrogatory and the type of the interrogatory (e.g., having a file extension of: “.what”, “.how”, “.where”, “.which”, “.when”, “.who”, or “.why”). The content of each of the knowledge snippet files comprises an answer to the interrogatory. Language systems and methods, including a run-time environment to process or execute the knowledge, are provided to use the knowledge so represented. The knowledge may also be presented in many different perspectives or views.

Claims

1. A computer-implemented method for organizing and processing knowledge in a computer-readable, near-natural language, comprising:

(a) maintaining a collection of knowledge snippet files according to a nomenclature in which each of the knowledge snippet files has a name relating to a particular interrogatory and the type of the interrogatory and content in a near-natural language comprising an answer to the interrogatory, at least one of the knowledge snippet files including a link to another of the knowledge snippet files;

(b) selecting one of the knowledge snippet files;

(c) storing in memory content of the selected knowledge snippet file; and

(d) storing in memory content of at least one other knowledge snippet file linked from the selected knowledge snippet file;

(e) dynamically consolidating inputs and outputs of different “

how”

-type knowledge snippet files, the “

how”

-type knowledge snippet files sequenced by a user into a new parent knowledge snippet file which is presented to the user with net inputs and outputs after the consolidating and de-duplicating intermediate inputs and outputs from each of child “

how”

-type knowledge snippet files.

2. The method of claim 1, further comprising presenting user-desired knowledge from a particular perspective, using at least part of the stored content;

wherein steps (c) and (d) are performed based at least in part on parsing the content of the selected knowledge snippet file and resolution, by user input, of one or more parsed token contained in a meta tag.

3. The method of claim 1, further comprising executing the selected knowledge snippet file, wherein the selected knowledge snippet file relates to a “

how”

type of interrogatory.