Searching alternative data sources

Patent metadata

Field	Value
Publication number	US-10127307-B2
Application number	US-201414217671-A
Country	US
Kind code	B2
Filing date	Mar 18, 2014
Priority date	Nov 28, 2012
Publication date	Nov 13, 2018
Grant date	Nov 13, 2018

Abstract

A method and system for searching alternative data sources include monitoring a first communications source broadcasting unstructured data, and a second communications source broadcasting structured data. The method further includes generating text from the unstructured data and from the structured data collected, and parsing the generated text. The method also includes defining a search phrase, and analyzing the generated or parsed text for semantically relevant text in relation to the search phrase. The method also includes selecting the semantically relevant text.

First claim

What is claimed is:

1. A method for searching sources of data, comprising steps of:
capturing unstructured broadcast data from a broadcast source by using a listening device, wherein the broadcast source includes one or more of a television broadcast source and a radio broadcast source;
converting the captured unstructured broadcast data to a first structured data, the converting comprising detecting an audio component of the captured unstructured broadcast data and converting the audio component to structured text;
receiving a second structured data from a source different from the broadcast source;
capturing a digital image from a source different from the second structured data source, wherein the digital image includes graphical text elements recognizable as text by optical character recognition software;
converting the graphical text elements of the digital image to a third structured data, the third structured data comprising structured text from the graphical text elements;
generating text of contents of each of the first structured data, the second structured data, and the third structured data and storing the generated text on a searchable data storage device;
parsing the stored text by (i) collecting only phrases relevant to a specified field and (ii) grouping phrases recognized as comprising temporal references;
receiving a search phrase and a selection of a particular language from a user;
semantically analyzing the parsed text by ignoring any text not in the particular language;
searching the semantically analyzed text using the received search phrase;
generating search results based on searching the semantically analyzed text; and
providing the search results for communication to the user.

2. The method of claim 1 , further comprises:
identifying, by the processor, one or more advertisement components of the stored text, wherein the searching is based on the one or more advertisement components.

3. The method of claim 1 , wherein the second structured data includes one or more of streaming audio and streaming video.

4. The method of claim 1 , wherein the unstructured broadcast data includes one or more of analog audio and analog video having an analog audio component.

5. The method of claim 1 , wherein the second structured data includes one or more emails.

Assignees

• IBM

Inventors

• Duong Quang T
• Gordon David M
• Schmidt David L
• Will Scott A

Classifications

• G06F17/3002

  Physics · mapped topic

• G06F17/30634 Primary

  Physics · mapped topic

• G06F16/41 Primary

  Indexing; Data structures therefor; Storage structures · CPC title

• G06F16/33 Primary

  Querying · CPC title