Back to results
US 9,165,556 B1Granted
Methods and systems related to audio data processing to provide key phrase notification and potential cost associated with the key phrase
AssigneePredictive Business Intelligence, LLC
InventorsJamie Sugar, Kleber Gallardo, Phil Williams
Life of this patent
Estimating the term…
Abstract
Methods and systems related to analyzing audio data and performing quantitative analysis and reporting related to the audio data. Audio data may be investigated and phrases of the audio data identified. In some implementations, phrases are identified based on the likelihood of an occurrence or non-occurrence of the phrase. In some implementations, phrases are identified based on a cost associated with a search phrase.
Claims22
A computer implemented method of analyzing content originating from an audio source, comprising:
- identifying an audio file, wherein said audio file is representative of spoken content of at least one speaker;
- identifying a key phrase, wherein said key phrase includes one or more words of interest;
- identifying a cost, wherein said cost is based on a value of either the presence of or the absence of said key phrase in said audio file;
- identifying a candidate phrase in said audio file, wherein said candidate phrase is representative of one or more words spoken by said at least one speaker and present in said audio file;
- associating a confidence level with said candidate phrase, wherein said confidence level is based on a probability that said candidate phrase matches said key phrase;
- identifying a threshold confidence level;
- determining whether said confidence level satisfies said threshold confidence level;
- providing a notification when said confidence level satisfies said threshold confidence level, said notification providing an indication of the presence of said key phrase in said audio file;
- wherein the providing of said notification is at least partially based on said cost;
- and
- identifying a cost adjusting characteristic of said audio file;
- and
- determining an adjusted cost, wherein said adjusted cost is an alteration of said cost and is based on determining said confidence level satisfies said threshold confidence level and identifying said cost adjusting characteristic of said audio file.
10 more elements
A computer implemented method of analyzing content originating from an audio source, comprising:
- receiving an audio file, wherein said audio file is representative of spoken content of at least one speaker;
- identifying a key phrase, wherein said key phrase includes one or more words which a user has interest in identifying in said audio file;
- receiving a cost, wherein said cost is indicative of a value of either the presence of or the absence of said key phrase in said audio file;
- identifying a candidate phrase in said audio file, wherein said candidate phrase is representative of one or more words spoken by said at least one speaker in said audio file and is a likely match for said key phrase;
- identifying a threshold cost value;
- determining whether said cost satisfies said threshold cost value;
- providing a notification when said cost satisfies said threshold cost value, said notification providing an indication of the presence of said key phrase in said audio file;
- and
- identifying a cost adjusting characteristic of said audio file;
- and
- determining an adjusted cost, wherein said adjusted cost is an alteration of said cost and is based on determining said confidence level satisfies said threshold confidence level and identifying said cost adjusting characteristic of said audio file.
8 more elements
A system, comprising:
- a first analysis processor, said first analysis processor operable to execute instructions stored in memory, comprising instructions to:
- receive a first audio file segment, wherein said first audio file segment is representative of spoken content of at least one speaker;
- receive a key phrase, wherein said key phrase includes one or more words of interest;
- identify a cost, wherein said cost is based on a value of either the presence of or the absence of said key phrase in an audio source;
- identify a first candidate phrase in said audio file segment, wherein said first candidate phrase is representative of one or more words spoken by said at least one speaker and present in said audio file segment;
- associate a first confidence level with said first candidate phrase, wherein said first confidence level is based on a probability that said first candidate phrase matches said key phrase;
- identifying a first threshold confidence level;
- determine whether said first confidence level satisfies said threshold confidence level;
- provide a first notification when said first confidence level satisfies said first threshold confidence level, said first notification providing an indication of the presence of said key phrase in said audio file;
- wherein the providing of said first notification is at least partially based on said cost;
- and
- identify a cost adjusting characteristic of said audio file;
- and
- determine an adjusted cost, wherein said adjusted cost is an alteration of said cost and is based on determining said confidence level satisfies said threshold confidence level and identifying said cost adjusting characteristic of said audio file.
11 more elements
A computer implemented method of analyzing content originating from an audio source, comprising:
- identifying an audio file, wherein said audio file is representative of spoken content of at least one speaker;
- identifying a key phrase, wherein said key phrase includes one or more words of interest;
- identifying a cost, wherein said cost is based on a value of either the presence of or the absence of said key phrase in said audio file;
- identifying a candidate phrase in said audio file, wherein said candidate phrase is representative of one or more words spoken by said at least one speaker and present in said audio file;
- associating a confidence level with said candidate phrase, wherein said confidence level is based on a probability that said candidate phrase matches said key phrase;
- identifying a threshold confidence level;
- determining whether said confidence level satisfies said threshold confidence level;
- providing a notification when said confidence level satisfies said threshold confidence level, said notification providing an indication of the presence of said key phrase in said audio file;
- identifying a cost adjusting characteristic of said audio file;
- and
- determining an adjusted cost, wherein said adjusted cost is an alteration of said cost and is based on determining said confidence level satisfies said threshold confidence level and identifying said cost adjusting characteristic of said audio file
- wherein the providing of said notification is at least partially based on said adjusted cost.
9 more elements
Dependencies are inferred from the claim text.