At&t Intellectual Property I, LP

At+t Corp.

James M Wilson

Theodore J Roycraft

Cecilia Castillo

Dilek Zeynep Hakkani-Tur

Gokhan Tur

Mazin G Rahim

Giuseppe Riccardi

Barbara B Hollister

Daniel Leon Stern

Tirso M Alonso

Ilana Bromberg

Lawrence Lyon Rose

A method, system and module for automatically validating dialogs associated with a spoken dialog service. The method comprises extracting key data from a dialog call detail record associated with a spoken dialog service, transmitting the key data as a dialog to a state-based representation (such as a finite-state machine) associated with a call-flow for the spoken dialog service and determining whether the dialog associated with the key data is a valid dialog for the call-flow.

System and method for the automatic validation of dialog run time systems

A method, system and module for automatically converting a call flow into a state-based representation are disclosed. The method comprises walking a call flow and converting each page of the call flow into a rule of a higher level representation of the call flow, augmenting the higher level representation with terminal symbols representing state variable assignments and comparisons associated with decision and computation shapes in the call flow and converting the higher level representation into a state-based representation.

System and method for converting graphical call flows into finite state machines

A method and apparatus for processing unique dialogs from a dialog system to identify paths through the dialog system, each unique dialog having a corresponding identified path and generating a visual call flow representation as a function of the identified paths of the dialog system, the call flow representation including nodes and edges connected to correspond to the identified paths.

Method for automatic graphical profiling of a system

Systems and methods for annotating speech data. The present invention reduces the time required to annotate speech data by selecting utterances for annotation that will be of greatest benefit. A selection module uses speech models, including speech recognition models and spoken language understanding models, to identify utterances that should be annotated based on criteria such as confidence scores generated by the models. These utterances are placed in an annotation list along with a type of annotation to be performed for the utterances and an order in which the annotation should proceed. The utterances in the annotation list can be annotated for speech recognition purposes, spoken language understanding purposes, labeling purposes, etc. The selection module can also select utterances for annotation based on previously annotated speech data and deficiencies in the various models.

Reducing time for annotating speech data to develop a dialog application

Growing community of inventors

Berkeley Heights, NJ, United States of America

James M Wilson