Generating vocal user interface code from a data meta-model

摘要

A method, computer program and system for generating code of the vocal interface of a software application by reading the persistent data of an application designed by a data meta model using a design language such as UML (Unified Modeling Language) defining a hierarchical organization of classes, their attributes and the relations between classes.;Similarly to the wizard for graphical user interface the vocal interface code comprises dialog unit code, which is Forms when VoiceXML language is used, to allow at execution time, entry of data for the application in a guided way.

主权项

1. A method executed on a computer for generating a vocal interface code for a software application, said method comprising:
receiving in a computer system an entity-relationship data model of said software application, said entity-relationship data model describing data of said software application and including classes, class attributes, and relations between classes; and analyzing the entity-relationship data model to automatically produce code of a voice-enabled wizard for acquiring by user interaction one or more values of respective class attributes of classes described in said entity-relationship data model, said code of the voice-enabled wizard being expressed in a dialog-oriented language, said code of the voice-enabled wizard being produced by:
collect all the classes in the data model and loop on all the classes collected; andfor each class read from the entity-relationship model, identified as being neither a class used for an interface of the software application nor an abstract class, creating, from the class descriptions of the entity-relationship model, a structure of a dialog unit code as a respective form in said dialog-oriented language for acquisition of values for the corresponding attributes of the respective class;collect all the attributes of the current class and its ancestors and loops on all the attributes; andfor each attribute of an identified class read from the entity-relationship model, if the identified class has a single relationship to one other class or if the identified class has a multiple relationship to other classes but the multiple relation does not apply to the respective attribute, creating in the dialog unit code corresponding to the identified class, code in said dialog-oriented language for a direct acquisition of value for the respective attribute else, creating in the dialog unit code, a sub-dialog code in said dialog-oriented language for enabling selection of a dialog unit code of the other classes; andfor each form created in said dialog-oriented language, identifying whether the created form is an entry form or an internal form by reading the relationships associated to the corresponding class in the entity-relationship model, and adding the created form to the voice-enabled wizard as an entry form or as an internal form according to the identification;wherein creating in the dialog unit code, code for a direct acquisition of value for the attribute comprises:create a prompt code for direct acquisition fo value for the attribute;read the attribute type form the entity-relationship model.