Brain text processing model: A new approach based on conceptual dependency stories
-
1
University of West Bohemia, Department of Computer Science and Engineering, Czechia
One of the possibilities how to model text processing by the human brain is to model story processing, whereas these stories include verbal communication. Interesting model of story structure and format was introduced by Roger C. Schank [1]. This model and similar ones are described also in [2]. This problem is connected with "Artificial Intelligence: Conceptual Dependency".
The aim of this article is to find and test a new experimental text processing in human brain model. There is a program that allows teaching computer simple stories. These stories consist of sequence of events in virtual world and of text communication. In this virtual world there are intelligent software agents controlled by program and lifeless objects. Events and virtual world are added for better simulation of word semantics. Following problems are partially solved: data model simulating human brain memory, saving stories from scripts to data model, imitation of similar stories and data model and stories visualization.
Stories are stored in semantic network. This network is composed of nodes and oriented edges with weights. The type of nodes can be special or standard. Special nodes are predefined at the beginning of the program runtime and represent letters and attributes. Standard nodes are created by story learning and each such node is defined by another nodes. Whole network is divided into a few parts where each part is divided into layouts. Whole network represents neural network of memory in human brain, nodes represent neurons, edges represent synapses and the division of the network represent division of a brain memory.
The special feature of this solution is used. The oriented edges only connect nodes and don't carry semantic information as in the Shank's model and as in the many others models. The weight of each edge will be just one scalar value. The meaning of each node is defined only by edges connecting this node to other nodes and by attributes as if a node is one of a sequence or trustworthiness, etc. The meaning is defined by a connection with special nodes. The next special feature is universality. Special nodes can represent letters as well as frequences or a position of pixel in pixel matrix or a position and type of chess figure on chessboard. The program at this time can process events, changes of internal states and also the text communication between agents in the simple virtual world. Later the program will be extended to support fuzzy logic, trustworthiness, language discrimination, etc.
Test of the quality of story processing is as follows: at the beginning the agent will be trained with a few stories and in next step we will test on how many stories the agent suggests right finishing of story chopped in the middle. At this time methods for story finishing are in the development state.
References
1. Roger C. Schank, Larry Tesler: A conceptual dependency parser for natural language, Association for Computational Linguistics Morristown, NJ, USA, 1969
2. Elaine Rich: Artificial intelligence, The University of Texas at Austin, NY, USA, 1983
Conference:
Neuroinformatics 2009, Pilsen, Czechia, 6 Sep - 8 Sep, 2009.
Presentation Type:
Poster Presentation
Topic:
Computational neuroscience
Citation:
Kratochvil
P and
Kleckova
J
(2019). Brain text processing model: A new approach based on conceptual dependency stories.
Front. Neuroinform.
Conference Abstract:
Neuroinformatics 2009.
doi: 10.3389/conf.neuro.11.2009.08.090
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
22 May 2009;
Published Online:
09 May 2019.
*
Correspondence:
Petr Kratochvil, University of West Bohemia, Department of Computer Science and Engineering, Pilsen, Czechia, krato@kiv.zcu.cz