# Proposed final Argumentation Mining seminar outline

I think we can give a good first introduction to the general topic of Argumentation Mining following this structure, which is based partly on the structure of the textbook itself and partly on the the definition of argumentation mining in ch. 1.2.

## Proposed outline
1. Introduction
- what argumentation is (information mostly from ch. 1, but possibly also essential concepts from ch. 2-3)
- what argumentation _mining_ is (ch. 1, especially steps mentioned in 1.2)
- why argumentation mining (from ch. 1.3 and ch. 10.3)
- structure of the rest of the talk (based on steps in 1.2)
2. ADUs
- presentation of running example (maybe a microtext?)
- (distinguishing argumentative from non-argumentative text/spans (ch. 5.1))
- (explanation of the task)
- (interactive annotation experiment)
- (presentation of various approaches)
- segmenting text into ADUs (ch. 5.2)
- explanation of the task
- interactive annotation experiment
- presentation of various approaches
3. Claims (based on ch. 5.3)
- definition of claim
- finding claims
- explanation of the task
- interactive annotation experiment
- presentation of various approaches
- definition of central claim
- finding the central claim(s)
- explanation of the task
- interactive annotation experiment
- presentation of various approaches
4. Supporting and objecting statements (based on ch. 6, similar structure as above: definition of a concept, annotation experiment, presentation of existing approaches)
5. Argumentation structure/relations between ADUs (based on ch. 7, maybe disregarding parts on dialogue. Similar structure, if possible including one last annotation experiment)
6. (Assessing arguments (based on parts of ch. 8))
7. Summary

## Division of tasks
I propose that each of us excepts the first and the last person to speak present one section.
The first one can take 1 and 2, as 1 is very introductory and 2 should not take too long.

The last one could take 5 and 7, as 7 is supposed to just be a quick recap.
The potential 6th section is not very important in my opinion, and just like chapter 9 I don't think it counts as "mining", but if we decide to go with it we should probably reconsider who says what.

## Annotation experiments
I like the idea of making the audience try to annotate a text, but I think it might be impractical to make them do that on paper and then collect answers if we want to make them perform several annotation tasks.

I think we can instead use something like Kahoot (the software that was used for the quiz at the retreat) and make the audience _select_ one between different possible annotation. If we are clever with designing the answers, they can resemble different approaches to the task at hand and can be a good starting point to explain each of them right after the mini-experiment itself.

## Limitations
- based on the textbook, no idea if/when/how much to refer to the papers we read
- not so much on argumentation schemes
- nothing on argument generation, but I think that really would just be a plus