Title:
|
MODELING DIALOGS WITH LINGUISTIC WAVELETS |
Author(s):
|
Daniela López De Luise, Débora Hisgen, Andrés Cabrera, Marcelo Morales Rins |
ISBN:
|
978-972-8939-73-1 |
Editors:
|
Miguel Baptista Nunes, Guo Chao Peng, Jörg Roth, Hans Weghorn and Pedro Isaías |
Year:
|
2012 |
Edition:
|
Single |
Keywords:
|
Chatterbot. Morphosintactic Linguistic Wavelets. |
Type:
|
Full Paper |
First Page:
|
11 |
Last Page:
|
18 |
Language:
|
English |
Cover:
|
|
Full Contents:
|
click to dowload
|
Paper Abstract:
|
This paper describes briefly a model implemented in the conversational robot named WIH (Word Intelligent Handler) and several statistics showing how it represents information extracted from Spanish dialogs. In order to get knowledge quickly, dialogs are processed with a batch interface. The prototype has complex architecture with many components. It is organized in several memories; each one models Spanish sentences that implement Morphosintactic Linguistic Wavelets (MLW). Using a knowledge-representation mechanism, WIH can model answers automatically taking previous dialogs with or between humans. Along with test cases description are also statistical analyses of the results obtained with real dialogs. It is shown here that WIH prototype handles structures called Ece that are the kernel of its learning in a human-like fashion, and that it is able to use it as a model for answers in future dialogs. |
|
|
|
|