This is an old revision of the document!
I found this early draft of the graph transducer work in April 2015 and decided to offer it online because, in retrospect, this is a much better document than any of the published papers. There should be a corresponding AT&T Bell Laboratoriesa technical report (HA615600-960701-01TM) sleeping in the AT&T archives.
Abstract: We present here a general architecture for building Automatic Document Analysis Systems. This architecture is composed of a succession of modules transforming graphs describing lower-level hypotheses on the documents into graphs describing higher level hypotheses. This architecture generalizes techniques used in Neural Networks, Optical Character Recognition, Natural Language Processing and Speech Recognition.
transducer-1996.djvu transducer-1996.pdf transducer-1996.ps.gz
@misc{bottou-1996,
author = {Bottou, L\'{e}on and Bengio, Yoshua and {LeCun}, Yann},
title = {Draft report: Document Analysis with Transducers},
year = {1996},
month = {July},
optnote = {Made available on http://leon.bottou.com/publications in 2015},
url = {http://leon.bottou.org/papers/bottou-1996},
}