International Business Machines Corporation

Marco Noel

Andrew Ronald Freed

Victor Povar

Ryan Brink

Aaron T Smith

Shane M Kilmon

Vamshi Krishna Thotempudi

Jasmeet Singh

Martha Holloman

Mohammad Gorji-Sefidmazgi

Daniel Zyska

Aishwarya Hariharan

In an approach for training data enhancement for an interactive response system, a processor retrieves a set of training data including a set of intents, a set of entities, and a set of utterances that map to each intent. A processor determines iteratively a root verb among the set of utterances for each intent. A processor to determine a set of new intents based on analysis of the determined root verb by performing a pairwise iteration and similarity score over the set of intents. A processor determines iteratively one or more new entities for each new intent. A processor generates a set of new training data based on the set of new intents and entities.

Training data enhancement

A method, system, and computer program product for recognizing, and training to recognize, alphanumeric speech data inputs. The method may include segmenting a data input into a sequential n-gram chunks based on a predetermined rule, where the data input is received through speech recognition. The method may also include receiving metadata regarding characteristics of the data input. The method may also include generating a language model based on the metadata. The method may also include generating a first set of language model variations of the data input. The method may also include training the language model based on at least the first set of language model variations. The method may also include generating one or more alternatives for the data input using the trained language model. The method may also include transmitting an output including the one or more alternatives for the data input.

Speech recognition and training for data inputs

Methods, computer program products, and systems are presented. The methods, computer program products, and systems can include, for instance: determining, in performance of an interactive voice response (IVR) session, prompting data for presenting to a user, and storing text based data defining the prompting data into a data repository; presenting the prompting data to the user; receiving return voice string data from the user in response to the prompting data; generating a plurality of candidate text strings associated to the return voice string of the user; examining the text based data defining the prompting data; augmenting the plurality of candidate text strings in dependence on a result of the examining to provide a plurality of augmented candidate text strings associated to the return voice string data; and evaluating respective ones of the plurality of augmented candidate text strings associated to the return voice string data; and selecting one of the augmented candidate text strings as a returned transcription associated to the return voice string data.

Contextualized speech to text conversion

An approach to correcting transcriptions of speech recognition models may be provided. A list of similar sounding phonemes from associated with the phonemes of high frequency terms may be generated for a particular node associated with a virtual assistant. An utterance may be transcribed and receive a confidence score regarding the correctness of the transcription based on audio metrics and other factors. The phonemes of the utterance can be compared to the phonemes of the high frequency terms from the list and a score for the matching phonemes and similar sounding phonemes can be determined. If it is determined the sounds similar score for a term from the high frequency term list is above a threshold, the transcription can be replaced with the term, providing a corrected transcription.

Improving speech recognition transcriptions

Processing data for speech recognition by generating hypotheses from input data, assigning each hypothesis, a score according to a confidence level value and hypothesis ranking, executing a pass/fail grammar test against each hypothesis, generating replacement hypotheses according to grammar test failures, assigning each replacement hypothesis a score according to a number of hypothesis changes, and providing a set of hypotheses, wherein the set comprises at least one replacement hypotheses.

Growing community of inventors

CA

Marco Noel