This document describes the project as part of the Linked Data & Semantic Web course. The project is carried out in addition to the labs session of the project, and are executed in pairs (groups of 2). Each group will work on its own project based on a dataset created by the group. The lab sessions will deliver input on the steps and tools to be taken and used in the project. In the lab session the instructions and assignments will be done on the tools and datasets (not being the project datasets). In some cases the student can repeat the approach afterwards within the student's own project and its own specific dataset.

Case
The University of Twente has published several dashboards and datasets related to the Campus Buildings. These dataset publications are not in a Linked Data way, or in any other desired re-usable way of publishing. Your project will contribute to an improved data publication related to the Campus Building. The data needs to be supportive to your case which you can select yourself. For inspiration the cases can be related to environmental issues, emergency response, creating better maps, planning of educational activities, and so on.

Inspiration on Data
Just as with the case, each student has freedom to choose and create its own dataset. Only for inspiration we provide some links to data on the utwente website.

Names, addresses, abbreviations: https://www.utwente.nl/en/service-portal/campus/route-campus-map-addresses-and-buildings#building-names-and-addresses
Lecture room information: https://www.utwente.nl/en/service-portal/workplace-support/building-caretakers/lecture-rooms#therm-exams-location
Energy consumption: https://energydata.utwente.nl/
(Live)Occupancy: https://apps.utwente.nl/campuscrowdmap/?zoom=15&building-section=16
- We collected data from 2022-03-16 08:00 to 2022-03-16 09:00 about the people count (occupancy) in the UT buildings, which is available below as JSON files that follow the GeoJSON standard. If you want to use another time window for your project, please contact Joao Moreira via Teams.
- Dataset 2022-03-16 08:00: building-section-floor_prd_001.json
- Dataset 2022-03-16 08:05: building-section-floor_prd_002.json
- Dataset 2022-03-16 08:15: building-section-floor_prd_003.json
- Dataset 2022-03-16 08:15: building-section-floor_prd_004.json
- Dataset 2022-03-16 08:20: building-section-floor_prd_005.json
- Dataset 2022-03-16 08:25: building-section-floor_prd_006.json
- Dataset 2022-03-16 08:30: building-section-floor_prd_007.json
- Dataset 2022-03-16 08:35: building-section-floor_prd_008.json
- Dataset 2022-03-16 08:40: building-section-floor_prd_009.json
- Dataset 2022-03-16 08:45: building-section-floor_prd_010.json
- Dataset 2022-03-16 08:50: building-section-floor_prd_011.json
- Dataset 2022-03-16 08:55: building-section-floor_prd_012.json
- Dataset 2022-03-16 09:00: building-section-floor_prd_013.json

Dataset of the whole month (February 2022): with this data you can, for example, make analysis over building occupancy by weekday.

You can visualize these data with: https://geojson.io

Inspiration on Linking
The data can be (modeled/implemented) as one graph/dataset, or multiple, but then with linking (on model level). Additional external linking on model/instance level might be done with other vocabularies and repositories, like schema.org, wikidata or OpenStreetMap (OSM), or the Kadaster data in the Netherlands. For the latter the key register Addresses and Buildings and the key register Topography, or the Knowledge Graph in which these are integrated makes most sense (available through data.labs.kadaster.nl).

Useful resources for reusing and validating ontologies:

Linked Open Vocabularies: https://lov.linkeddata.es/dataset/lov/
OntOlogy Pitfall Scanner!: http://oops.linkeddata.es/
BioPortal: https://bioportal.bioontology.org/
Lightweight Implementation of the Unified Foundational Ontology: https://nemo-ufes.github.io/gufo/
Smart Applications REFerence Ontology, and extensions: https://saref.etsi.org/

Some additional links for tooling

Within the lab sessions you will set up your tools/environment for linked data. Here we provide some additional links that are not necessary, but at some point might become handy.

PLDN Triplestore (free account for storing triples and setting up a SPARQL endpoint by sending an e-mail to Erwin): https://data.pldn.nl/ (although it might look a bit Dutchy, when you login it is a full English environment)
Templated tool to transform CSV into LD is LD Wizard (might be useful either as first step or when you want avoid code): https://github.com/netwerk-digitaal-erfgoed/LDWizard-Core
- The tool has multiple implementations, but the implementation at Kadaster contains a direct upload feature to the PLDN triplestore mentioned above. By that it is not only transformation, but also storing triples and spinning up a SPARQL endpoint (a manual activity). The tool itself is English, including documentation, code: https://labs.kadaster.nl/demonstrators/geodatawizard/#1
- However, the implementation at Kadaster contains a Dutch intro/tutorial: https://labs.kadaster.nl/demonstrators/geodatawizard/
There is also a server based version of OpenRefine, free to use: http://openrefine.server.pldn.nl
Example SPARQL endpoint GUI: https://data.labs.kadaster.nl/kadaster/kg/sparql/default
Example of SPARQL queries put together in a story: https://labs.kadaster.nl/stories/
Visualization of Ontologies can be done with Weaver. https://utwente.wvr.io/
This commercial tool is provided by Weaver; For a login just send an email to Erwin. For an example see this: https://kadaster.wvr.io/kg-kadaster?tab=home&branch=main

Milestones and deliverables

M0: Case analysis and refinement
M1: Tripliflication of non-LD dataset
- Deliverable: Report. Triplified dataset with existing vocabularies/ontologies.
- “Live”: Published triples on the web.
M2: Ontology engineering
- Deliverable: Report. New ontology (or ontology profile) with instances that can be result from triplification or manually addition.
- “Live”: Published View of ontology
M3: Integrated analysis
- Deliverable: Report on linking (& ontology matching), querying and analyzing datasets from M1 and M2.
- “Live”: Dashboard (website) containing some SPARQL query visualizations.
M4: Project presentation
- Deliverable: Reports merged (M1-M3).
- Presentations on 13/04 13:45 - 17:30 (see Rooster).