Message transcription, voice query and query delivery system
A message transmission system accepts a telephone call from a user who wishes to send an e-mail message, send an SMS message, perform an Internet query or retrieve his or her electronic mail. The voice call is transcribed and the message is sent, or the question in the voice call is transcribed and answered by an agent. Any number of agents connect to a central site over an Internet connection and transcribe messages or answer queries in an assembly line like fashion. In addition, a Web query delivery system accepts a query or statement from a user; the query is transcribed, classified, and then broadcast over any medium to any number of experts or web sites that desire to answer the particular type of query received. The entire query is delivered to an expert or web site who provides a full answer to the user.
1. A method of transcribing a message comprising:
receiving a voice message from a user at a computer server along with a recipient address;
recording said voice message as a voice file at said computer server;
executing a speech recognition engine to transcribe said voice file into a text file;
sending said voice file and said text file to a computer workstation operated by an agent;
receiving a corrected text file from said computer workstation at said computer server, said corrected text file being a more accurate transcription of said voice file than said text file;
comparing said text file with said corrected text file to determine the accuracy of said speech recognition engine;
determining that said accuracy of said speech recognition engine is sufficient and that future voice messages from said user need not be corrected by an agent;
routing said corrected text file and said recipient address to a delivery means; and
delivering said corrected text file to a recipient using said recipient address.
2. A method of transcribing a message comprising:
receiving a voice message from a user at a computer server along with a recipient address;
recording said voice message as a voice file at said computer server;
executing a first speech recognition engine to transcribe said voice file into a first text file;
executing a second speech recognition engine to transcribe said voice file into a second text file;
comparing, by said computer server, said first text file and said second text file to determine accuracy of words and phrases;
producing, by said computer server, a hybrid text file based upon said first text file and said second text file, said hybrid text file being is more accurate than said first text file or said second text file,
sending said voice file and said hybrid text file to a computer workstation operated by an agent;
receiving a corrected text file from said computer workstation at said computer server, said corrected text file being a more accurate transcription of said voice file than said hybrid text file;
routine said corrected text file and said recipient address to a delivery means; and
delivering said corrected text file to a recipient using said recipient address.
3. A method as recited in claim 2 wherein said step of comparing uses accuracy information for a first phrase from said first text file along with accuracy information for a corresponding second phrase from said second text file to determine the accuracy of said first and second phrases.
4. A method of transcribing a message comprising:
receiving a voice message from a user at a computer server along with a recipient address;
recording said voice message as a voice file at said computer server;
executing a speech recognition engine to transcribe said voice file into a text file;
sending said voice file and said text file to a computer workstation operated by an agent;
receiving a corrected text file from said computer workstation at said computer server, said corrected text file being a more accurate transcription of said voice file than said text file;
routing said corrected text file and said recipient address to a delivery means;
delivering said corrected text file to a recipient using said recipient address;
feeding said text file and said corrected text file into said speech recognition engine, said corrected text file being flagged as the correct version of said text file;
building a user speech profile based upon said step of feeding by said speech recognition engine; and
storing said user speech profile for said user in association with said speech recognition engine.