Semantic Role Labelling Flashcards

1
Q

What is a semantic role?

A

It is a role that arguments of a predicate take in an event

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are some semantic role datasets?

A

PropBank and FrameNet

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is semantic role labelling?

A

It is the task of assigning roles to spans of text in sentences

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is a thematic role?

A

It captures the semantic commonality around participants of event. There are many roles, typically 12

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are some problems with thematic roles?

A

There are lots of different themes/events that lead to a lot of thematic roles

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What has been developed instead of thematic roles?

A

Generalised semantic roles

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What are the options for generalised semantic roles?

A

Defining them heuristically (PropBank)

Defining roles specific to groups of verbs or nouns (FrameNet)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

In PropBank, what are semantic roles tied to?

A

They are tied to specific verb sense predicates

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

In PropBank, what does each verb have?

A

Each verb has N arguments:

arg0 is usually the PROTO-AGENT, arg1 is usually the PROTO-PATIENT

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Explain what is shown in the image

A

In the image, it shows the first word sense for the words agree and fall and their arguments. For agree, argument 0, the PROTO-AGENT, is the agreer, while argument 1, the PROTO-PATIENT, is the proposition, and argument 2 is the other entity agreeing

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What else can argument have in PropBank?

A

They can have argument modifiers, known as ArgM

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is shown in the image?

A

These are different types of argument modifiers in PropBank

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

How is FrameNet different to PropBank?

A

It associates semantic roles with frames rather than verb predicates

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is a frame?

A

A frame is a background knowledge structure that unites a set of words

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What are frame elements?

A

They are a set of semantic roles and the predicates that use these roles

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Explain what is shown in the image

A

This is a frame and its frame element. The frame is change_position_on_a_scale. The description contains some frame elements that are within the brackets. It gives a description of some components that relate to the concept and each of these components are a frame element

17
Q

What are core and non core roles in FrameNet?

A

Core roles are frame specific, while non-core roles are some more general properties that can occur across a number of different frames, such as time, location etc

18
Q

What is feature based SRL?

A

It is one approach to SRL. The training set is PropBank or FrameNet, the input is a set of sentences from these training sets. A node is a node of the parse tree. We use a supervised classifier such as CRF to predict SRL for a node. The input for the classifier, X is a feature template for a specific (node, predicate) in the parse tree. Y is a SRL tag.

19
Q

What is feature-based SRL typically used with?

A

They are used with a pipeline where you prune the parse tree, identify nodes that are/are not arguments, then classify using a multi-class classifier for nodes

20
Q

What is global consistency?

A

Semantic roles of constituents are not independent, so per node local role classification is not optimal. Global consistency step is added as a second pass

21
Q

What are some feature templates for SRL?

A

Governing Predicate, phrase type of constituent, Headword of constituent + POS, Path in parse tree from constituent to predicate, Voice of clause, Position of constituent, subcategorisation of the predicate, NE type, first and last word

22
Q

What is Neural SRL?

A

It is a neural method to perform semantic role labelling

23
Q

What is a benefit of the Neural SRL model?

A

It does not require the parse tree as input, only a sentence and instead the parse tree can be embedded within the layers of the NN

24
Q

What is the input for SRL?

A

A sentence with some separators and an identifier for the predicate

25
Q

What is the output for Neural SRL?

A

The BIO tags for SRL

26
Q

How does the Neural SRL model work?

A

Inputs go through embedding layer. We include a flag for the word that is a predicate. This enters the biLSTM layer, as output we pass to a MLP with softmax to classify each of the labels.

27
Q

In Neural SRL, how does the BiLSTM encoder work?

A

Concatenate the BERT embedding and the predicate flag embedding

Can use transformer instead of BiLSTM

28
Q

In Neural SRL, how does the MLP + softmax decoder work?

A

Concatenate the hidden layer output for each word and hidden layer output for the predicate.

Can use CRF instead of MLP