Главная
Study mode:
on
1
Intro
2
Today's talk
3
Interpreting a sentence
4
The real situation
5
Other examples of ambiguity
6
Labelled examples
7
Extracting features
8
Statistical natural language processing
9
Labelled data is hard to obtain
10
Result: unequal access
11
Human language learning
12
Must computers learn like humans?
13
But language isn't "in the world"
14
Research programme
15
Learning biases
16
Stronger bias = less data
17
Example problem: segmentation
18
Word segmentation
19
Statistical learning experiment
20
Testing for learning
21
How do they do it?
22
What about real language?
23
Another strategy
24
A model for segmenting words
25
The right bias can help
26
The Dirichlet process model
27
Output of the system
28
Words aren't marbles
29
Improved system
30
Where else can these ideas help?
31
Continuing work
32
Meaning as translation
33
Results so far
34
Conclusions
35
Acknowledgements
Description:
Explore language learning in humans and machines through this insightful 46-minute talk by Dr. Sharon Goldwater, a Reader at the University of Edinburgh's School of Informatics. Delve into the challenges of natural language processing, the importance of understanding human language acquisition, and the potential for incorporating human-like learning biases into computational systems. Discover how this approach could address the resource-intensive nature of current language processing methods and expand their applicability to a wider range of the world's languages. Examine examples from Dr. Goldwater's research, including word segmentation and meaning as translation, to gain a deeper understanding of the complexities of language learning and the innovative solutions being developed in the field.

Language Learning in Humans and Machines - Making Connections to Make Progress

Alan Turing Institute
Add to list