Nuance Communications, Inc.

Felix Calls, LLC

Yext, Inc.

Felix Call, LLC

Brent D Metz

Kevin Caffrey

Sean Joseph MacIsaac

James Robert Lewis

Linda M Boyer

Rajesh Balchandran

Ilia Mirkin

Ciprian Agapi

Howard C Lerman

A method () and system () is provided to add the creation of examples at a developer level in the generation of Natural Language Understanding (NLU) models, tying the examples into a NLU sentence database (), automatically validating () a correct outcome of using the examples, and automatically resolving () problems the user has using the examples. The method () can convey examples of what a caller can say to a Natural Language Understanding (NLU) application. The method includes entering at least one example associated with an existing routing destination, and ensuring an NLU model correctly interprets the example unambiguously for correctly routing a call to the routing destination. The method can include presenting the example sentence in a help message () within an NLU dialogue as an example of what a caller can say for connecting the caller to a desired routing destination. The method can also include presented a failure dialogue for displaying at least one example that failed to be properly interpreted to ensure that ambiguous or incorrect examples are not presented in a help message.

Method and system for conveying an example in a natural language understanding application

A method () and system () is provided to add the creation of examples at a developer level in the generation of Natural Language Understanding (NLU) models, tying the examples into a NLU sentence database (), automatically validating () a correct outcome of using the examples, and automatically resolving () problems the user has using the examples. The method () can convey examples of what a caller can say to a Natural Language Understanding (NLU) application. The method includes entering at least one example associated with an existing routing destination, and ensuring an NLU model correctly interprets the example unambiguously for correctly routing a can to the routing destination. The method can include presenting the example sentence in a help message () within an NLU dialog as an example of what a caller can say for connecting the caller to a desired routing destination. The method can also include presented a failure dialog for displaying at least one example that failed to be properly interpreted to ensure that ambiguous or incorrect examples are not presented in a help message.

A system and method for extracting and presenting useful data from calls received by a client is disclosed. The resulting 'telephone call inbox' is a way for a client view pay per call advertising as a stream of consumers with information available to understand the call activity of the consumers and for the client to navigate their call history. The system automatically filters non-consumer fraudulent calls, extracts the identity of a consumer, aggregates several calling entities into a single consumer, transcribes the call into a call stream using voice recognition software, extracts patterns and draws conclusions from the call stream, and presents a list of call streams in a user friendly set of web pages configured as the telephone call inbox. The telephone call inbox includes, for each call, the caller ID, one or more key words, phrases or major conclusions concerning the call, and the voice recognized call stream.

Telephone call inbox

A system and method for optimizing the price of a call to a business and the placement of a business' information within an ad buy on a Web page are disclosed. A call is received by a client from a potential consumer of a product. The call is transcribed into a text call stream on both the consumer and the client sides of the call. Patterns are extracted from the call. The patterns extracted may be applied against patterns relevant to the client. Patterns relevant to the client may be tied to certain goals of the business of the client that indicate successful business transactions between the consumer and the client or that indicate a service provide by the client. Extracted patterns may be used for targeting of advertisements based on a call containing patterns relevant to the client.

Media optimization using transcription analysis

A system and method for automatic assignment of outbound and inbound call identity between a salesperson and a client is disclosed. For automatic assignment of outbound call identity between a salesperson and a client, a salesperson places a call to the telephone number of the client. A phone server matches a sales group to a salesperson and an entered sales group and an NPA (i.e., area code) of the client in a database. If a match is found, the phone server substitutes a replacement caller ID for actual caller ID of the salesperson. When the client answers the call, the replacement caller ID or default caller ID is displayed on client's phone. For automatic assignment of inbound call identity between a client and a salesperson, the client places a call using the replacement caller ID. The phone server searches the database for a sales group that matches the DID of the replacement caller ID. The phone server rings the telephones of salespeople that match the sales group entry in order of the most 'efficient' salesperson of the group until one of the salespeople answers their phone.

Method and system for automatic assignment of outbound and inbound call identity

A method and system for testing code within a speech application. A test file can be automatically generated to verify the functionality of a new section of code within a graphical call flow builder application. A user can specify through a wizard two points on a path identifying the code section to be tested. The wizard can generate a test file and configure a path to a new subpath. Values are assigned to graphical call flow prompts along the path. Thus, the new code section is reached under the same path conditions for allowing repeatable testing. The system can include a test harness to test a new code section from within a context of the speech application, and a test controller for transitioning to the new code section. The test controller can run the test harness within the speech application to evaluate a functionality of the new code section.

Method and system for testing sections of large speech applications

A system and method for optimizing the price of a call to a business and the placement of a business' information within an ad buy on a Web page are disclosed. A call is received by a client from a potential consumer of a product. The call is transcribed into a text call stream on both the consumer and the client sides of the call. Patterns are extracted from the call. The patterns extracted may be applied against patterns relevant to the client. Patterns relevant to the client may be tied to certain goals of the business of the client that indicate successful business transactions between the consumer and the client or that indicate a service provide by the client. Extracted patterns may be used to adjust the price of the call based on the client's willingness to pay for a call containing patterns relevant to the client.

A system and method for tracking call activity of a plurality of businesses is disclosed. A global pool of phone numbers is allocated from which to draw numbers. A business-specific pool of numbers is established for business locations involved in advertisements using the Internet via Web pages. As a consumer begins viewing an advertisement for a business, a number is temporarily allocated from the business's pool of available numbers. As the business's pool of unallocated numbers is exhausted, a metric is used to determine whether or not to grow the pool of numbers, to show an untracked 'fallback' number for the business, or to reuse a previously allocated number. When one of the temporarily allocated numbers is called, the online consumer it was associated with is credited for making the call and the business is credited and charged for receiving the call.

System and method for tracking call activity to a set of advertised businesses

The present invention discloses a speech processing solution that utilizes an original speech recognition grammar in a speech recognition system to perform speech recognition operations for multiple recognition instances. Instance data associated with the recognition operations can be stored. A replacement grammar can be automatically generated from the stored instance data, where the replacement grammar is a statistical language model grammar. The original speech recognition grammar, which can be a grammar-based language model grammar or a statistical language model grammar, can be selectively replaced with the replacement grammar. For example when tested performance for the replacement grammar is better than that for the original grammar, the replacement grammar can replace the original grammar.

Automatic grammar tuning using statistical language model generation

A system and method for tracking call activity of a plurality of businesses is disclosed. A global pool of phone numbers is allocated from which to draw numbers. A business-specific pool of numbers is established for business locations involved in advertisements using the Internet via Web pages. As a consumer begins viewing an advertisement for a business, a number is temporarily allocated from the business's pool of available numbers. As the business's pool of unallocated numbers is exhausted, a metric is used to determine whether or not to grow the pool of numbers, to show an untracked 'fallback' number for the business, or to reuse a previously allocated number. When one of the temporarily allocated numbers is called, the online consumer it was associated with is credited for making the call and the business is credited and charged for receiving the call. This method of allocating and presenting numbers with all numbers ultimately being tied to the same business ensures that even if a consumer calls a number much later after it has been re-allocated, that the number will still ring to the correct business.

A method and system for addressing disambiguation issues in interactive applications by creating a disambiguation system for generating complex grammars that includes homonym detection and grouping, and provides optimization feedback that eliminates time-consuming and repetitive iterative steps during the grammar generation portion of the interactive application configuration.

Growing community of inventors

New York, NY, United States of America

Brent D Metz