HUMSENTPROC
Uncovering the nature of human sentence processing: a computational/experimental approach
| Coordinatore | UNIVERSITY COLLEGE LONDON
Organization address
address: GOWER STREET contact info |
| Nazionalità Coordinatore | United Kingdom [UK] |
| Totale costo | 179˙603 € |
| EC contributo | 179˙603 € |
| Programma | FP7-PEOPLE
Specific programme "People" implementing the Seventh Framework Programme of the European Community for research, technological development and demonstration activities (2007 to 2013) |
| Code Call | FP7-PEOPLE-2009-IEF |
| Funding Scheme | MC-IEF |
| Anno di inizio | 2010 |
| Periodo (anno-mese-giorno) | 2010-09-01 - 2012-08-31 |
Partecipanti
| # | ||||
|---|---|---|---|---|
| 1 |
UNIVERSITY COLLEGE LONDON
Organization address
address: GOWER STREET contact info |
UK (LONDON) | coordinator | 179˙603.20 |
Mappa
Word cloud
Esplora la "nuvola delle parole (Word Cloud) per avere un'idea di massima del progetto.
Obiettivo del progetto (Objective)
'The project’s research objective is to further our understanding of how people process language. This is accomplished by developing and implementing a range of computational models of sentence processing, each embedding different psychological mechanisms. Next, the models' ability to account for a large body of experimental data is evaluated. On the basis of how the model(s) fit the data, we can isolate the most plausible psychological mechanisms for sentence processing. The notion of 'word surprisal' is used to link the models to the data. Formally, word surprisal is defined in terms of informativeness of a word in sentence context. Quantitative measures of surprisal can therefore be extracted from different probabilistic language models. Empirically, surprisal can be viewed as the extent to which the word came unexpected to a reader or listener, having a measurable effect during sentence processing. To capture both aspects of word surprisal, the project combines psycholinguistic experimentation and computational modelling. On the experimental side, word-reading times and ERP data will be collected to obtain measurements of the surprise experienced by readers on encountering words. On the modelling side, several probabilistic sentence-processing models will be implemented to obtain estimates of word surprisal which serve to predict the empirical data. Matching the model's surprisal estimates to the experimental findings will identify the model(s) that provide(s) the most accurate description of human sentence processing. This will inspire the development of a more advanced model that simulates sentence processing more adequately. The main training objective is for the applicant to obtain the necessary knowledge, skills and experience to set up and perform psycholinguistic reading-time and EEG experiments. This complements his current expertise in computational methods, putting him in an ideal position to perform further integrative research in cognitive science.'