* This is the deadline for all submissions except where an approved extension is in place.

Late submissions received within 5 working days of the deadline will be capped at 40%. Late submissions received later than 5 days after the deadline will receive a mark of 0.

It is your responsibility to check that your submission has uploaded successfully and obtain a submission receipt.

Your work must be done by yourself (or your group, if there is an assigned groupwork component) and

comply with the university rules about plagiarism and collusion. Students suspected of plagiarism, either   of published or unpublished sources, including the work of other students, or of collusion will be dealt with according to University guidelines (https://www.dur.ac.uk/learningandteaching.handbook/6/2/4/).

Your report must:

1.   Document how data was collected, and briefly state any relevant modules or toolkits that were used. If using SPARQL queries not contained in your code (other than the one provided), you must include these as separate files within your code submission.

2.   Describe your implementation and the models constructed.

3.   Summarize your results in appropriate ways, including use of suitable visualizations.

4.   Critically evaluate the models. This must involve critical analysis of the adequacy of the modelling (e.g. what are the assumptions the models rely on, and do they all hold; are there factors or biases that might invalidate conclusions drawn?), but should also involve some comparison with external data and/or published social/political science research on relevant subjects.

5.   Discuss what conclusions can be drawn from the model.

In this assignment, please note that:

.    You are not expected to resolve issues of lack of recorded data (or incorrectly recorded data) in additional data sources such as Wikidata. However, you are expected to account for the potential effects of these in your analysis, and should attempt to quantify these where possible.

.    You are allowed to reuse existing code – this includes open source modules and toolkits, Stack

Overflow responses, code described in online tutorials, etc., provided that whenever you do this, you clearly indicate what you have reused and exactly where it came from (including a URL for an online source). You should document smaller instances of reuse (e.g. short code examples from Stack Overflow) as comments in your code, and longer ones both in your code and as citations to the source in your report. Marks will be awarded for the original parts of your work (i.e. your extensions to and adaptations or developments of any reused material).

. Copying any material (this includes code and text) in your submission from any source without clear acknowledgement is likely to constitute plagiarism, which can have serious consequences as described in the Teaching and Learning Handbook:

https://www.dur.ac.uk/learningandteaching.handbook/6/2/4/

.    Marks will not be awarded under the “Technical depth/clarity of implementation” sections for parts of your code that reimplement code available in widely used libraries or modules.

.    You may use any programming language suitable for the task; use of Python 3.x is recommended where possible.

.    In writing your report, you should assume that the reader is familiar with everything covered

during the lectures for this module. Concepts and techniques important to your implementation which were not introduced during the module should be explained.

.    The word limit is exclusive of figures and references. You may include a maximum of 20 figures   (recommended: around 5-10 figures). Please use a 12pt font, single-spacing, and A4 pages. Your report should begin with an introduction, and does not require a separate abstract.

Submission (2 files)

.    A written report as a PDF file.

.    An archive (.zip / .tar.gz) containing the code for your implementation, including instructions for how to run it in a README.TXT file. If any data that your code relies on is over 10 Mb in size, include instructions in this file for how to obtain it and do not include it in the archive. Your code is not marked separately, but serves as supporting evidence for the criteria indicated in the marking scheme.