|
|
Vorlesungsverzeichnis >> Fakultät Sozial- und Wirtschaftswissenschaften >>
|
PWB-PT-VS: Methodenworkshop Computational Social Science (ENTFÄLLT)- Dozent/in
- Maximilian Noichl
- Angaben
- Seminar
Rein Präsenz
Zeit und Ort: n.V.
- Voraussetzungen / Organisatorisches
- Schedule
28.10.2022, 14:00–18:00: Einführung: Auswahl der Impulsreferate, Abgleich von Vorkenntnissen
4.11.2022, 14:00–17:00: Impulsreferate
11.11.2022, 14:00–18:00: Whirlwind tour: Tools & Resources
Ideen-Papiere eingereicht bis 20. November, Individuelle Konsultationen, vor der folgenden Sitzung
9.12.2022, 11:00–18:00: Kollaborative Projektarbeit: Datensammeln
20.1.2023, 12:00–18:00: Kollaborative Projektarbeit: Datenanalyse
3.2.2023, 14:00–18:00: Abschlusssitzung, Präsentation der Resultate
Einreichung der term-paper. Abgabefrist: 31.03.2023
Registration:
In FlexNow, please register directly for the „Prüfung“ (examination), NOT for the „Lehrveranstaltung“ (course participation). Registration opens on Tuesday, 04/10/2022 10:00 and closes on Sunday, 30/10/2022 23:59.
In case you are unable to register in FlexNow:
If the winter term 2022/23 is your first term studying in Bamberg and you have not yet been provided with access to FlexNow or if the registration period has closed already, please get in touch with Mrs Kohlmann (secretariat Political Theory)
Assessment:
Term Paper
Please note:
- Please subscribe to the political science mailing list ( politikwissenschaft-liste.sowi@uni-bamberg.de) via iam.uni-bamberg.de navigating to "Verteilergruppen Eigene Mitgliedschaften".
- Please also sign up for the VC course „Politikwissenschaft studieren in Bamberg“.
- Inhalt
- The increased amount of digital traces societies leave behind, in conjunction with the drastic increase in the computational power available to process them, has led to the emergence of an interdisciplinary endeavor commonly called computational social science (CSS). CSS today is located at the intersection of political science, sociology, computer science, complexity science, and in part overlaps with the similarly constituted digital humanities.
The methods of CSS include network analysis, natural language processing, agent based modelling, and many others. They are often employed in interlocking fashion and integrated into sometimes quite complex processing pipelines, which serve to analyze and visualize disparate data domains to reveal social structures and dynamics.
In this course, students will get the opportunity to develop their own methodological CSS-toolkit, and explore how it integrates into their own political theorizing: In the first phase of the course, we will survey the literature for interesting methodological innovations and evaluate them together.
In particular, students can expect to encounter:
- Models for text-representation – topic modeling, transformer models, Sentiment Detection, …
- Unsupervised machine-learning – dimensionality reduction, clustering, …
- Network analysis – clustering, centrality-measures, …
- Working with social media data
- Working with parliamentary records
- Working with digitized archival data
- Working with citation data, analyzing disciplinary relations, …
Following their interests, students choose an article from a pool of preselected readings, and give a short, focused presentation on the article’s goals and methodologies. Inspired by these presentations, students then propose their own projects (in groups or individually). The instructor will provide individual feedback on the proposals, and together we try to move them towards feasible projects of appropriate scope.
In two afternoon-length collaborative hackathon sessions focused on data-procurement and dataanalysis, students then realize their projects, possible in overlapping group-work. As a term-paper, students submit an extended abstract (1500 words + graphics & code/methodology) describing their work. The final grade consists of 30% classroom-participation, 20% quality of proposal and 50% term-paper.
Successful participation in the course does not presuppose that students are familiar with programming techniques. While not a programming-course in itself, the course might give people unfamiliar with computational techniques a first opportunity to get some experiences. Students can choose their own technological frameworks for their projects. The instructor will be able to support them when working in Python, R or NetLogo.
- Englischsprachige Informationen:
- Title:
- Workshop Computational Social Science
- Credits: 8
- Zusätzliche Informationen
- Erwartete Teilnehmerzahl: 15
- Institution: Lehrstuhl für Politikwissenschaft, insbes. Politische Theorie
Hinweis für Web-Redakteure:
Wenn Sie auf Ihren Webseiten einen Link zu dieser Lehrveranstaltung setzen möchten, verwenden Sie bitte einen der folgenden Links:Link zur eigenständigen Verwendung Link zur Verwendung in Typo3
|
 |
|
|
UnivIS ist ein Produkt der Config eG, Buckenhof |
|
|