Language and Modeling Flashcards
What is human cognition?
The study of how information is acquired, represented and processed in the human brain.
What is the goal of computational modeling?
To simulate a cognitive process via computational tools and techniques, and use the model to explain the observed behavior.
What is fast mapping?
When a new concept is learned based on minimal exposure to a given unit of information.
–> Young children can easily map novel words to novel objects in a familiar context
Define and differentiate between these empirical findings and the theoretical account. (Use fast mapping & vocabulary learning as an example)
- Empirical findings - what we observe.
- How do children behave in specific situations?
- How does their knowledge of word meaning develop over time? - Theoretical account - explanation for the observed behavior
- What type of info. do children use for learning words?
- What learning mechanisms do they rely on?
What happens in an experiment?
Controls the environment, making sure that all the participants experience the same environment
Observe the resulting behavior
What is a theory?
A detailed description of the cognitive process
Aims to predict the behavior
What does a model do?
A model simulates the environment. implements the cognitive processes, and generates behavior
What does the use of computational modeling in cognitive processes enable us to do?
- Study cognitive processes through simulation
- Evaluate the plausibility of existing cognitive theories
- Explain observed human behavior
- Predict behavioral patterns
What is the symbolic modeling framework?
Uses symbolic systems of knowledge representation & logic-based inference techniques
What are the advantages of symbolic modeling?
Transparent and accessible
What is the connectionist modeling framework?
- Uses distributional representations of knowledge
- Parallel processing of data
Name an advantage and a problem with connectionist modeling.
Advantage: can precisely replicate cognitive processes
Problem: cannot scale up to the real world; have only be used in toy worlds
What is probabilistic modeling?
Modeling framework that combines descriptiveness of symbolic modeling with flexibility of connectionist modeling
In symbolic modeling, how is knowledge represented?
As a set of symbols and their propositional relations
In symbolic modeling, what is the learning and processing mechanism?
Knowledge is processed an updated via general rules and under certain constraints