International Business Machines Corporation

Elton B Sherwin, Jr

Balakrishna Raghavendra Iyer

Bhaskar Sinha

John M Lucassen

Paul S Cohen

Teresa A Meriwether

Frank Vincent Castellucci

Vincent M Stanford

Roger M Miller

Ora J Williamson

Jorge L Vizcaino

A method and system are disclosed for reducing perplexity in a speech recognition system within a telephonic network based upon determined caller identity. In a speech recognition system which processes input frames of speech against stored templates representing speech, a core library of speech templates is created and stored representing a basic vocabulary of speech. Multiple caller-specific libraries of speech templates are also created and stored, each library containing speech templates which represent a specialized vocabulary and pronunciations for a specific geographic location and a particular individual. Additionally, the caller-specific libraries of speech templates are preferably processed to reflect the reduced bandwidth, transmission channel variations and other signal variations introduced into the system via a telephonic network. The identification of a caller is determined upon connection to the network via standard caller identification circuitry and upon detection of a spoken utterance, that utterance is processed against the core library, if the caller's identity cannot be determined, or against a particular caller-specific library, if the caller's identity can be determined, thereby greatly enhancing the efficiency and accuracy of speech recognition by the system.

Method and system for reducing perplexity in speech recognition via

A continuous speech recognition and voice response system provides a natural sounding and effective interactive, speech-driven dialogue from a data processing system. A concatenation of words into phrases and sentences improves recognition and mimic natural language processing. The system uses speaker-independent, continuous-speech to initiate the playback of audio files. The system employs high-speed context switching to modify the active vocabulary and applies high-speed context switching to modify or activate Audio WAV voice response files. The system uses dialogue history to activate selected context, Baukus-Naur Form (BNF) grammars and WAV files and provides phrase or sentence long dialogue prompts to improve accuracy. The system also provides audio prompts to improve accuracy and provides speech-activated buttons to navigate between menus.

Continuous speech recognition and voice response system and method to

A method and system for reducing perplexity in a speech recognition system based upon determined geographic location. In a mobile speech recognition system which processes input frames of speech against stored templates representing speech, a core library of speech templates is created and stored representing a basic vocabulary of speech. Multiple location-specific libraries of speech templates are also created and stored, each library containing speech templates representing a specialized vocabulary for a specific geographic location. The geographic location of the mobile speech recognition system is then periodically determined utilizing a cellular telephone system, a geopositioning satellite system or other similar systems and a particular one of the location-specific libraries of speech templates is identified for the current location of the system. Input frames of speech are then processed against the combination of the core library and the particular location-specific library to greatly enhance the accuracy and efficiency of speech recognition by the system. Each location-specific library preferably includes speech templates representative of location place names, proper names, and business establishments within a specific geographic location.

Method and system for location-specific speech recognition

Describes novel methods for compressing data character strings into 'storage optimized indices' (SOIs) and stores their adaptive Ziv Lempel (AZL) indices, called 'evolution based indices' (EBIs), in fields in corresponding entries in a SOI dictionary. The method also compresses data using the SOI dictionary, which accesses the corresponding EBIs for representing the compressed data. The EBIs are put into storage, or transmitted to a receiving location. Greater data compression processing efficiency is obtained by using the SOI dictionary than is available using prior types of AZL dictionaries. The disclosure further describes methods for decompressing EBI indices into corresponding phrases at a receiving location using either a conventional AZL dictionary or a SOI dictionary after translating received EBIs into SOIs. Also described is a submethod for phrase length determination for use in the decompression process. Using this phrase length detection, received phrases are decompressed in their reverse character order through a buffer, or directly in a target storage area, or in a combination of the latter two techniques. The phrase length may be stored in corresponding dictionary entries to avoid a length tracing process in the invention.

Transmission compatibility using custom compression method and hardware

A method of transmitting compressed data using a Ziv-Lempel compression/expansion algorithm, using an adaptive Ziv-Lempel (AZL) dictionary modified to a mature state. The mature state is signaled by a time to freeze signal sent as a switch-over signal from a transmitting location to each receiving location. These signals freeze and synchronize the AZL dictionaries at both locations, and starts a translation of the frozen AZL dictionary to a static SZL dictionary--at least at the transmitting location. The SZL dictionary is then used to compress records being transmitted. An index translation process is generates translation information to allow the receiving locations to decompress SZL indices into original characters. The AZL-to-SZL dictionary translation process re-organizes the frozen AZL to an SZL dictionary. The SZL process is used until either the end of the inputted sequence, or a time to unfreeze signal is generated. An SZL to AZL switch-over signal is generated in response to the time to unfreeze signal, which in turn signals a switch over back to the AZL process and invokes the saved frozen AZL dictionary to be used to until mature to on the current input data stream at which time the AZL is frozen and a switch-over signal is provided and a new SZL is generated.

Growing community of inventors

Stamford, CT, United States of America

Elton B Sherwin, Jr