X hits on this document

PDF document

The split-fovea model of visual word - page 12 / 78

329 views

0 shares

0 downloads

0 comments

12 / 78

Chapter 1. Introduction and Background

7

Figure 1.3: The Seidenberg and McClelland (1989) Model: Triangle model with lay-

ers representing orthography, phonology, meaning and context. Activation spreads be-

tween layers via the hidden layer between them

the success of connectionist models, since every serious reading model should be able to simulate the two different forms of dyslexia.

The

Seidenberg

and

McClelland

(1989)

Model

The

Seidenberg

and

McClelland

(1989) model was one of the first successfull connectionist networks used to model

reading processes. The whole model (also called ’triangle model’) different layers (excluding the hidden layers) as shown in figure 1.3:

consists

of

four

  • Orthography: The orthographic form the words.

  • Phonology: The phonological form of the words.

  • Meaning: The semantics of the words.

  • Context: The context in which the word appears.

Only the bottom part of the whole model, i.e. the mapping between orthography and phonology was implemented computationally. The model was implemented as a feed- forward connectionist networks, which did not include feedback from the phonological

Document info
Document views329
Page views343
Page last viewedSun Jan 22 18:35:05 UTC 2017
Pages78
Paragraphs3087
Words17414

Comments