From speech to SQL queries : a speech understanding system
Salma Jamoussi, Kamel Smaı̈li, Jean‐Paul Haton · 2005
In this paper, we describe our speech understanding system which has been tested on two different applications. The proposed system is a task specific one and it concern especially oral database consultation tasks. In this work, we consider that the automatic speech understanding problem could be seen as an association problem between two different languages. At the entry, the request expressed in natural language and at the end, before the interpretation stage, the same request is expressed in term of concepts. A concept represents a given meaning; it is defined by a set of words sharing the same semantic properties. We propose a new method based on Bayesian networks to automatically extract the underlying concepts. We also propose and compare three approaches for the vector representation of words. This paper ends with a description of the post-processing step during which the SQL query corresponding to an input sentence is generated. The understanding process is then implemented in a real speech recognition engine. For the two test applications, rates of 78 % and 81 % of well formed SQL requests have been obtained.