cognitive processes - cs187 · cognitive processes cs187 dr andr´e grun¨ ing department of...

19
Cognitive Processes CS187 Dr Andr´ e Gr¨ uning Department of Computing University of Surrey Email: [email protected] SS 2008 Dr Andr´ e Gr¨ uning Cognitive Processes

Upload: others

Post on 20-Aug-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Cognitive ProcessesCS187

Dr Andre Gruning

Department of ComputingUniversity of Surrey

Email: [email protected]

SS 2008

Dr Andre Gruning Cognitive Processes

Page 2: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Imaging Part

Part

Part

Dr Andre Gruning Cognitive Processes

Page 3: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Overview

Organisation of world knowledge in terms of categories

(Why categories?)

Membership in a category? Typicality?

How similar are two categories?

What are the relations between categories? / Organisation ofcategories.

Theories of Categorisation

Dr Andre Gruning Cognitive Processes

Page 4: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Typicality

Sentence verification

In order to probe world knowledge and the relations between itsdifferent pieces, subjects are asked to tell whether sentence like thefollowing are true:

1 A canary is a bird.

2 An ostrich is a bird.

3 A potato is a tree

4 A rifle is a tree

Interesting? Yes, at least when you record error rates and reactiontimes.

Dr Andre Gruning Cognitive Processes

Page 5: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Typicality

Sentence Verification – Results

There is a Typicality Effect:

Question (1) (“Canary”) is on average answered faster thenquestion (2) (“Ostrich”)

There is a Category Similarity Effect:

It takes longer to reject (3) (“potato”) as untrue than (4)(“rifle”)

⇒ World knowledge is somehow structured. It takes differenttimes to access different pieces in different contexts

Dr Andre Gruning Cognitive Processes

Page 6: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Organisation of Categories

with the help attributesfrom Alan J. Parkin, Essential Cognitive Psychology, Psychology Press, Hove, 2005, p. 159.

Dr Andre Gruning Cognitive Processes

Page 7: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Organisation of Categories

Means of transportation.

On the black board.

Dr Andre Gruning Cognitive Processes

Page 8: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Relations between Categories

Relations between Categories

Superordinate: “animal” contains “fish”

Subordinate: “canary” is a “bird”

Siblings: “Shark” and “Salmon” are both members of “fish”

Default Inheritance: a category inherits attributes from itssuperordinate by default: “canary” can fly (from “bird”),breathes (from “animal”)

Overwriting Attributes: an ostrich cannot fly, though a “bird”can in general”

First approach to similarity: how many nodes to travel fromone category to the next.

Doesn’t predict the reaction times right.

Can’t explain typicality effects

Dr Andre Gruning Cognitive Processes

Page 9: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Attribute-based Categories

Attribute-based Categories and Similarity

there are core attributes (“must have” ones), (e.g. “horse” isfour-legged), defining in a strict sense

there are prototypical ones (“typical ones), (e.g. a “bird” canfly”, those typical one that spring to mind first.

for comparing categories:

check how many prototypical attributes agrees,if overlap not sufficient: check only on core attributes.

Problems:

What number of common attributes?Inheritance: what number of overwrite to allow?

Different definition of categories than by attributes?

Dr Andre Gruning Cognitive Processes

Page 10: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Categories

What is a “cup”?

Some attributes:

one can drink liquids from it

mainly hot liquids like tea, coffee, hot chocolate

round

has a handle

made of ceramics

not too big

Dr Andre Gruning Cognitive Processes

Page 11: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

When is a cup a cup?

A psychophysical experiment: Subjects are asked how they wouldname the objects shown.taken from reproduction in Ellis and Hunt: Fundamentals of Cognitive Psychology, McGraw Hill, Boston, 1993, p.

210.

Dr Andre Gruning Cognitive Processes

Page 12: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Context dependence of Categories

Context dependence

Subjects who saw the picture on the slide before were either askedto imagine the objects to be filled with

with liquids (solid lines)

with food (dashed lines)

taken from reproduction in Ellis and Hunt: Fundamentals of Cognitive Psychology, McGraw Hill, Boston, 1993, p.211.]

Dr Andre Gruning Cognitive Processes

Page 13: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Context dependence of Categories

Results from Experimental Cognitive Science

Categories boundaries depend on context.

Categories are fuzzy.

Categories membership can be gradual.

Dr Andre Gruning Cognitive Processes

Page 14: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Categories

Theories of Categorisation

Exemplar Theory

Attribute Theory

Prototype Theory

. . .

Dr Andre Gruning Cognitive Processes

Page 15: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Categories

Attribute Theory

A category is defined through a lists of attributes and features

You have to live with exceptions:

Category “bird”Attribute “can fly”Instance “penguin”: attribute: “cannot fly”

works a bit like default inheritance in OO programming

subcategories/instances can overwrite defaults fromsupercategory

to test categories membership, just check the features. It’s adiscrete theory.

difficulties with “goodness-of-example” or “typicality effects”:

when all features/attributes are met, then recognition of“not-so-typical” instance should be as fast as very “central”instances: this is not the case!

Dr Andre Gruning Cognitive Processes

Page 16: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Categories

Prototype theory

A category is defined through a prototypeA prototype is like a quintessence or summary of all instances of thecategory you have ever encountered (or imagined)It is like an “ideal” member of the category, not necessary one thatexists in the real world.think of an ideal tree or ideal chair!There is some measure of the distance of an instance to theprototype: it’s a metric theory.The smaller the distance the more typical the instanceThink of the mountain landscape of a Hopfield network: distance isthe time it takes the ball (representing the instance) to roll down tothe valley (the prototype)Can deal easier with reaction times and typicality effects.Can deal better with fuzzy category membership.Fits better in with the “generation, reconstruction” paradigm ofmemory retrieval (⇒Ian Wells, last week)Smooth transition from exemplar based to prototype based categoriesin development (think of the irregular verbs!)

Dr Andre Gruning Cognitive Processes

Page 17: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Categories

Exemplar theory

all instances are stored

allows you to assess also the variability of the categories

no abstraction/generalisation over instances takes place

needs a lot of storage: memory is overloaded, so doesn’tsupport one main advantage of categorisation: datacompression.

maybe employed when you are building up a new category andso far have only a few instances for it.

Dr Andre Gruning Cognitive Processes

Page 18: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Summary

World knowledge as expressed in categories

Categories are organised in a (single?) hierarchy

Categories can be defined in terms of attributes or prototypes.

However there are often too many exceptions to allow forstrict and hard rules or clear category boundaries.

Dr Andre Gruning Cognitive Processes

Page 19: Cognitive Processes - CS187 · Cognitive Processes CS187 Dr Andr´e Grun¨ ing Department of Computing University of Surrey Email: a.gruning@surrey.ac.uk SS 2008 Dr Andr´e Gruning¨

Dr Andre Gruning Cognitive Processes