Methods and systems for measuring semantics in communications

• US 9,269,353 B1
• Filed: 12/05/2012
• Issued: 02/23/2016
• Est. Priority Date: 12/07/2011
• Status: Active Grant

• Current Assignee: Lingo IP Holding, LLC
• Original Assignee: Manu Rehani, Warren L. Wolf
• Inventors: Rehani, Manu, Wolf, Warren L.
• Primary Examiner(s): Desir, Pierre-Louis
• Assistant Examiner(s): Tzeng, Forrest F
• Application Number:
  US13/706,093
• Time in Patent Office: 1,175 Days
• Field of Search: 704/9
• US Class Current: 1/1
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding

Abstract

A content is divided into tokens. Tokens can represent sounds, symbols, letters, words, sentences, paragraphs, or documents, among other possibilities. Probable measures of meaning (PMMs) are determined for each token. Given a target token, a proximity function for other tokens can be calculated, which in turn can be used to modify the PMMs for the other tokens. These modified PMMs for the other tokens can then be used in conjunction with the PMM for the target token to generate a relative measure of meaning (RMM) for the target token.

Claims

1. A system, comprising:

a server;

a probable measure of meaning (PMM) database stored on the server, the PMM database storing PMMs for a plurality of tokens;

an input port to receive a content;

a token determination unit on the server operative to determine a token in said content;

a relative measure of meaning (RMM) calculation unit on the server operative to calculate a RMM for said token, factoring in a first probable measure of meaning (PMM) from the PMM database for said token and a plurality of second PMMs from the PMM database for a plurality of second tokens in said content using a proximity function (PF) according to

2. A system according to claim 1, wherein the PMM database is stored on a second server, said second server and the server connected via a network.

3. A system according to claim 1, further comprising a proximity function database.

4. A system according to claim 3, wherein the proximity function database is stored on the server.

5. A system according to claim 3, wherein the proximity function database is stored on a second server, said second server and the server connected via a network.

6. A method, comprising:

receiving a content at a server over a network;

identifying a first token in the content using the server;

determining a first probable measure of meaning (PMM) for the first token from a PMM database stored on the server;

identifying at least two second tokens;

determining a second PMM for each of the at least two second tokens from the PMM database stored on the server;

using a proximity function (PF) to modify each of the second PMMs for each of the at least two second tokens, the proximity function factoring in the first token and a distance between the first token and each of the at least two second tokens; and

generating a relative measure of meaning (RMM) for the first token, the RMM factoring in the first PMM for the first token and each of the modified PMMs for each of the at least two second tokens, according to

7. A method according to claim 6, wherein:

identifying a first token in the content using the server includes identifying a first sentence in content using the server; and

identifying a second token includes identifying a second sentence.

8. A method according to claim 6, wherein:

identifying at least two second tokens includes identifying a second token preceding the first token and a third token following the first token;

determining a second PMM for each of the at least two second tokens from the PMM database includes determining the second PMM for the second token from the PMM database and determining a third PMM for the third token from the PMM database;

using a proximity function to modify each of the second PMMs for each of the at least two second tokens includes using the proximity function to modify the second PMM for the second token and to modify the third PMM for the third token, the proximity function factoring in the first token and the distance between the first token and the second token and third token; and

generating a RMM for the first token includes generating the RMM for the first token, the RMM factoring in the first PMM for the first token, the modified PMM for the second token, and the modified PMM for the third token.

9. A method according to claim 6, wherein:

identifying at least two second tokens includes identifying five second tokens preceding the first token and five third tokens following the first token;

determining a second PMM for each of the at least two second tokens from the PMM database includes determining a PMM for each of the five second tokens and the five third tokens from the PMM database;

using a proximity function to modify each of the second PMMs for each of the at least two second tokens includes using the proximity function to modify the second PMMs for each of the five second tokens and the five third tokens, the proximity function factoring in the first token and the distance between the first token and each of the five second tokens and the five third tokens; and

generating a RMM for the first token includes generating the RMM for the first token, the RMM factoring in the first PMM for the first token and the modified PMM for each of the five second tokens and the five third tokens.

10. A method according to claim 6, further comprising using the RMM to modify the PMM for the first token in the PMM database.

11. An article, comprising a non-transitory storage medium, said non-transitory storage medium having stored thereon instructions that, when executed by a machine server, result in:

receiving a content at a server over a network;

identifying a first token in the content using the server;

determining a first probable measure of meaning (PMM) for the first token from a PMM database stored on the server;

identifying at least two second tokens;

determining a second PMM for each of the at least two second tokens from the PMM database stored on the server;

using a proximity function (PF) to modify each of the second PMMs for each of the at least two second tokens, the proximity function factoring in the first token and a distance between the first token and each of the at least two second tokens; and

generating a relative measure of meaning (RMM) for the first token, the RMM factoring in the first PMM for the first token and the each of modified PMMs for each of the at least two second tokens, according to

12. An article according to claim 11, wherein:

identifying a first token in the content using the server includes identifying a first sentence in content using the server; and

identifying a second token includes identifying a second sentence.

13. An article according to claim 11, wherein:

identifying at least two second tokens includes identifying a second token preceding the first token and a third token following the first token;

determining a second PMM for each of the at least two second tokens from the PMM database includes determining the second PMM for the second token from the PMM database and determining a third PMM for the third token from the PMM database;

using a proximity function to modify each of the second PMMs for each of the at least two second tokens includes using the proximity function to modify the second PMM for the second token and to modify the third PMM for the third token, the proximity function factoring in the first token and the distance between the first token and the second token and third token; and

generating a RMM for the first token includes generating the RMM for the first token, the RMM factoring in the first PMM for the first token, the modified PMM for the second token, and the modified PMM for the third token.

14. An article according to claim 11, wherein:

identifying at least two second tokens includes identifying five second tokens preceding the first token and five third tokens following the first token;

determining a second PMM for each of the at least two second tokens from the PMM database includes determining a PMM for each of the five second tokens and the five third tokens from the PMM database;

using a proximity function to modify each of the second PMMs for each of the at least two second tokens includes using the proximity function to modify the second PMMs for each of the five second tokens and the five third tokens, the proximity function factoring in the first token and the distance between the first token and each of the five second tokens and the five third tokens; and

generating a RMM for the first token includes generating the RMM for the first token, the RMM factoring in the first PMM for the first token and the modified PMM for each of the five second tokens and the five third tokens.

15. An article according to claim 11, said non-transitory storage medium having stored thereon further instructions that, when executed by the server, result in using the RMM to modify the PMM for the first token from the PMM database.