Back to results
US 2003/0195872 A1Published application
Web-based information content analyzer and information dimension dictionary
AssigneeNone recorded
InventorPaul Senn
Life of this patent
Estimating the term…
Loading patent…
Loading patent…
AssigneeNone recorded
InventorPaul Senn
Estimating the term…
The purpose of the invention described below is to make the information available via the Internet (broadly defined to include intranets and extranets) more useful for applications which need to distinguish not only the factual content, but also other dimensions of information content. One example of a type of information content which is not factual is emotional content Current search engines allow for keyword searches, but it is quite cumbersome to use these engines to pick out sites with specific emotional tones. For-instance, suppose one wanted to use a currently existing search engine to pick out all of the English language internet sites containing predominantly negative references to Microsoft corporation. The English language is complex enough that the task of forming the right keyword and phrase list is formidable. Simply looking for adjectives like “bad” or “disappointing” or “frustrating”, etc combined with the keyword Microsoft will yield thousands of false matches. Such a search would also miss many overall negative sites because the methods of expressing negativity in the English language are so varied.
A method of searching a network, such as the Internet, using a computer operatively coupled for communication with the network, comprising:
A method of searching on the Internet, or other network, comprising:
A method of creating an Information Dimension Dictionary for use in searching a network, such as the Internet, using a computer operatively coupled for communication with the network, comprising:
Dependencies are inferred from the claim text.
Drawings are temporarily unavailable. Open the PDF at USPTO
From the US filing history. Foreign family members aren’t listed.
Backward citations are indexed for grants only.
Retrieval from the web: indexing; Web crawling techniques
Querying, e.g. by the use of web search engines: query formulation
Querying, e.g. by the use of web search engines: presentation of query results
Computing and data processing
Term end not estimated
The present invention is directed to a search engine for searching the Internet, or like networks, for a specific dimension, such as emotional content, which is not available using current, conventional search engines. The problem is not specific to Microsoft, the English language or negative references. The problem exists any time a desire to scope searches based on emotional tone is present, in any language. This invention provides a general solution to this problem by insulating the user from the complex algorithms required to come up with a reasonable set of matches for a search of this type. The user enters the specific keywords for the domain he is interested in (Microsoft, Rugby, flying, whatever it may be) and then picks the “information dimension” he is interested in. An example of an information dimension would be a specific emotional tone (happy, angry, tired, sad, etc).
The search engine and system of the present invention will perform a series of searches (one or more) using Content Retrieval Objects (CROs) described below to form the correct search keys. The system will then make use of Content Analysis objects (CAOs) to determine which search results contain the tone of interest and the strength of the tone in each result. CAOs may arrive at their determination using a variety of means. One method used by CAOs could be the comparison of the search results with a repository of audio, visual or text patterns which are relevant to the Information Dimension of interest. This document will describe a version of such a repository which is an extension of a standard dictionary, and which we will refer to as the Information Dimension Dictionary. The CAOs do not return simply yes or no answers to the question of which sites match the criteria, but instead return a summary of search results including a rating, a reference count and a confidence level. The rating describes the level of, for-instance positive emotional tone for a given site or sites and the reference count gives the number of sites the rating is based on. The confidence level describes the level of confidence in the rating.
Another element which can be measured by CAO's is the contrast index. The contrast index describes the difference in strength and direction of emotional tone between two chosen information domains in a given information dimension. For-instance, Sun Corporation is a large competitor to Microsoft Corporation, so one might be interested in the contrast between the reference index for Sun and the reference index for Microsoft within the information dimension of positive emotional tone.
The invention will be more readily understood with reference to the accompanying drawing, wherein:
The FIG. 1 is a schematic showing the major components of the system of the invention;
FIG. 2 is a flowchart of the Content Retrieval Objects (CRO);
The search invention of the invention is preferably used for the Internet 10, in the figure, to which are operatively coupled Web Sites 12, a system-administrator browser 16, an end-user browser 18, a WICA 20 (Web-Based Information Content Analyzer) described hereinbelow, and a WICA database 22, also described hereinbelow. The following describes the content of the WICA Site 20 and the WICA database.
“Content Retrieval Objects”
Referring to FIG. 2, the role of the Content Retrieval Object (CRO) is to create Search Results that can be passed on to the Content Analysis Engine (CAE), which will determine a Rating for the results in the Information Dimension and Information Domain of interest (Blocks 30-34). The CRO can operate in two modes: Broad Search Mode and Targeted Search Mode (Blocks 36, 38). In Broad Search Mode, the CRO takes the keywords chosen by the user which define the Information Domain, and simply performs a search using those keywords. The Search Results are then directly passed on to the CAE. Broad Search Mode simply performs a search in the same manner as the common search engines (such as AltaVista), in operation today (Block 40). Targeted Search Mode (Blocks 42-46), however, is a unique feature of this invention, which we shall now describe.