Computational Argumentation
Course, master, summer 2019, L.079.05811
General
Lecture
- Instructor. Henning Wachsmuth
- Location. O1.258
- Time. Wednesday, 1pm – 4pm, c.t.
- First date. April 10, 2019
- Last date. July 10, 2019.
Tutorial
- Instructor. Henning Wachsmuth, Milad Alshomary
- Location. F1.544 (DISCO)
- Time. Tuesday, 9am – 11am, c.t.
- First date. April 16, 2019
- Last date. July 9, 2019
Announcements
- This course has ended.
Description
Argumentation is an integral part of both professional and everyday communication. Whenever a topic or question is subject to controversy, people consider arguments to form opinions, to make decisions, or to convince others of a certain stance. In the last years, the computational analysis and synthesis of natural language argumentation has become an emerging research area, due to its importance for the next generation of web search engines and intelligent personal assistance. Based on fundamental techniques from natural language processing, computational argumentation ranges from the mining of arguments from natural language text, over the assessment of argumentation quality, to the retrieval of arguments in web search.
The students learn both fundamentals from argumentation theory and state-of-the-art techniques from computational argumentation. The use of respective methods is practiced in assignments.
Course material
The course will cover lectures the following topics. The slides from each lecture will be put here, usually soon after the respective lecture has taken place:
-
Introduction to computational argumentation (lecture, tutorial)
-
Basics of argumentation (lecture)
-
Applications of computational argumentation (tutorial)
-
Resources for computational argumentation (lecture)
-
Mining of argumentative units (material in PAUL)
-
Mining of supporting and objecting units (material in PAUL)
-
Mining of argumentative structure (material in PAUL)
-
Assessment of the structure of argumentation (material in PAUL)
-
Assessment of the reasoning of argumentation (material in PAUL)
-
Assessment of the quality of argumentation (lecture)
-
Generation of argumentation (material in PAUL)
-
Development of an argument search engine (material in PAUL)
-
Conclusion (lecture)
Slides with meta-information:
- Organizational course information (slides)
- Topic assignment process (slides)
- Topic assignment (slides)
Other material:
- Slide template for power point (pptx)
Assessment Tests
Self-assessment
Under the following URL, you'll find a self-assessment test, which is designed to help decide whether this course is the right choice for a student or not:
https://umfragen.uni-paderborn.de/index.php/387353?lang=en
The test covers a selection of basics from linguistics, statistics, and machine learning, all of which are considered as prerequisites for the course (along with good programming skills).
The test is voluntary, fully anonymous, and will not affect admission to the course. It is only intended for one's own consideration. However, students that struggle with several questions are advised to expand their skills first and maybe rather register next year.
Required course achievement
The maximum number of course participants is limited. In case the number of registered participants exceeds the maximum, there will be a mandatory test in the beginning, to decide about admission. The 24 students with the best results can then participate in the course.
In particular, each student will get an individualized exercise sheet after the first lecture, which needs to be solved and submitted within a few days. The test is expected to take about 1-2 hours - depending on one's background knowledge. It will cover questions related to statistics, linguistics, and machine learning, all of which are covered by the course Introduction to Text Mining, among others.
Besides the test, also a certain amount of participation is required. More details will be announced in the first lecture.
Notice: These are not the graded tasks of the course. Details on the grading in the organizational course information slides.