Using Maps for Interlinking Geospatial Linked Data
1 like118 views
The document discusses the use of maps for interlinking geospatial linked data, emphasizing the importance of manually creating authoritative interlinks across datasets. The authors identify requirements for a user-friendly tool to search, display, and manage geospatial resources, and they developed a faceted browser for this purpose. An experiment was conducted to assess the tool's effectiveness with participants performing tasks related to linking resources in geospatial datasets.
Using Maps for Interlinking Geospatial Linked Data
- 1. Using Maps for Interlinking Geospatial Linked Data Dieter Roosens1, Kris McGlinn2, and Christophe Debruyne1,2 1 WISE Lab, Vrije Universiteit Brussel 2 ADAPT Centre, Trinity College Dublin Presented at COOPIS 2019
- 2. Preamble data.geohive.ie: Ordnance Survey Ireland (OSi) publishing their authoritative geospatial information starting from publicly available boundary and building data supporting two use cases: providing different representations of geometries and capture evolution of geometries. How can we facilitate others in engaging with (and create links to) their data? 10/23/19 christophe.debruyne@adaptcentre.ie 2
- 3. Context and Problem • The creation of links within and across datasets is key for the creation of a Linked Data Web. • SotA often focusses on the creation of interlinks in a (semi-)automatic manner. • Observation 1: A survey conducted by McKenna et al. (2018) showed the importance of manually creating authoritative interlinks. • Observation 2: Many datasets on the Linked Data Web have a geospatial or geographic component. • Research Question: Can maps be used to create and manage interlinks of geospatial data on the Linked Data web? 10/23/19 christophe.debruyne@adaptcentre.ie 3
- 4. Related work • The SotA on interlinking focuses mostly on the (semi-)automatic creation of interlinks. E.g., in a declarative manner or with ML techniques. Some have support for geospatial functions. • What about the use of maps? • YASGUI: SPARQL editor with support for displaying results on a map • STRABON: a triplestore with support for displaying results on a map • FACETE: a faceted browser with support for displaying results • GVIZ allows one to draw polygons on a map for the creation of a query • Some of the SotA mentioned in the paper were related to the EU FP7 GeoKnow project. • To the best of our knowledge, no tools supports the creation of interlinks using maps. 10/23/19 christophe.debruyne@adaptcentre.ie 4
- 5. Design and Development • Maps are mostly used to display results • Appropriate tooling for subject matter experts (non-technical) is lacking • We formulated the following requirements 1. Search for geospatial resources in a Linked Data dataset; 2. Display the results on a map; 3. Engage with the resources on the map for the creation of interlinks; 4. Manage the created interlinks; 5. Keep track of provenance information. • Requirements 2, 3, and 4 are key to our study • Requirement 1 provides a starting point • Requirement 5 allows for solutions to be integrated (in future work) https://github.com/dieterroosens/LinkedDataApplication 10/23/19 christophe.debruyne@adaptcentre.ie 5
- 6. Design and Development: Searching for Resources We have chosen to adopt Faceted Browsing. Unfortunately, we were unable to adopt Facete (unable to compile and run both versions) and SemFacet (no support for geospatial information). A minimal viable faceted browser was developed for the purpose of the experiment Facets are used to create a SPARQL query. Each facet results in an additional triple pattern. The query also looks for an optional geometry and a label (with a preference for English labels, followed by default labels). 10/23/19 christophe.debruyne@adaptcentre.ie 6
- 7. Design and Development: Creating and Managing Interlinks Adoption of ConcurTaskTrees (CTTs) to design the steps in each tasks. Given a CTT, an algorithm provides information how an appropriate UI might look like (Mori et al, 2002). 10/23/19 christophe.debruyne@adaptcentre.ie 8
- 8. Experiment: Protocol 1. Read and sign an informed consent form; 2. Fill in a pre-questionnaire (prior knowledge); 3. Video with a quick introduction to RDF and Linked Data; 4. Slides with annotated screenshots and key points of the video were at the participants’ disposal; 5. Perform 8 tasks in a think aloud manner; 6. Fill in a post-questionnaire (PSSUQ) to assess the tool. 12 participants were recruited within the lead researcher’s network at home, work, and the VUB. Four participants were recruited within TCD. 10/23/19 christophe.debruyne@adaptcentre.ie 10
- 9. Experiment: Tasks Creating links from GeoHive to DBpedia and GeoNames and within GeoHive: 1. From County Dublin to same resource in DBpedia (URI provided) 2. From County Donegal to same resource in DBpedia (participants has to look for URI) 3. Find Wicklow Mountains in DBpedia and declare it to be within the right county. 4. Find the four counties that border County Cork and assert four geo:sf-touches relations 5. Look for the highest mountain, find its corresponding resource in DBpedia, and create the interlink 6. Look for a specific shop in County Dublin and declare it to be within GeoNames’ representation of that County 7. Look and correct any mistakes in the interlinks (one was introduced) 8. Download the interlinks 10/23/19 christophe.debruyne@adaptcentre.ie 11
- 10. Results: Prior Knowledge Participant OWL RDF TTL GEO Total 1 0 0 0 0 0 2 0 0 0 0 0 3 0 0 0 1 1 4 0 0 0 0 0 5 0 0 0 0 0 6 0 0 0 0 0 7 0 0 0 0 0 8 0 0 0 0 0 9 0 0 0 0 0 10 2 3 3 1 9 11 2 3 3 2 10 12 3 3 3 2 11 13 2 1 1 1 5 14 0 0 0 0 0 15 0 0 0 0 0 16 0 0 0 0 0 10/23/19 christophe.debruyne@adaptcentre.ie 12
- 11. Results: Task Performance 10/23/19 christophe.debruyne@adaptcentre.ie 13 Participant Task1 Task2 Task3 Task4 Task5 Task6 Task7 Task8 Percentage Task1 Task2 Task3 Task4 Task5 Task6 Task7 Task8 Percentage 1 1.00 0.00 1.00 1.00 0.00 0.00 1.00 1.00 62.50 1.00 0.50 1.00 1.00 0.50 0.00 1.00 1.00 75.00 2 1.00 0.00 1.00 1.00 0.00 0.00 1.00 1.00 62.50 1.00 0.50 1.00 1.00 0.50 0.00 1.00 1.00 75.00 3 1.00 0.00 0.00 1.00 1.00 0.00 1.00 1.00 62.50 1.00 0.50 0.00 1.00 1.00 0.50 1.00 1.00 75.00 4 1.00 0.00 1.00 0.75 1.00 0.00 1.00 1.00 71.88 1.00 0.50 1.00 0.75 1.00 0.00 1.00 1.00 78.13 5 1.00 0.00 1.00 0.25 1.00 0.00 1.00 1.00 65.63 1.00 0.00 1.00 0.25 1.00 0.00 1.00 1.00 65.63 6 1.00 1.00 1.00 1.00 1.00 0.00 1.00 1.00 87.50 1.00 1.00 1.00 1.00 1.00 0.00 1.00 1.00 87.50 7 0.50 1.00 1.00 0.75 1.00 0.00 1.00 1.00 78.13 0.50 1.00 1.00 0.75 1.00 0.00 1.00 1.00 78.13 8 1.00 1.00 0.00 0.75 1.00 0.00 1.00 1.00 71.88 1.00 1.00 0.00 0.75 1.00 0.00 1.00 1.00 71.88 9 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 75.00 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 75.00 10 0.50 0.00 1.00 1.00 1.00 0.00 1.00 1.00 68.75 0.50 0.50 1.00 1.00 1.00 0.50 1.00 1.00 81.25 11 1.00 0.00 0.00 1.00 0.00 0.00 1.00 1.00 50.00 1.00 0.50 0.00 1.00 0.50 0.50 1.00 1.00 68.75 12 1.00 1.00 1.00 0.00 1.00 1.00 1.00 1.00 87.50 1.00 1.00 1.00 0.00 1.00 1.00 1.00 1.00 87.50 13 1.00 0.00 1.00 1.00 1.00 0.00 0.00 1.00 62.50 1.00 0.50 1.00 1.00 1.00 0.00 0.00 1.00 68.75 14 1.00 0.00 1.00 1.00 0.00 0.00 1.00 1.00 62.50 1.00 0.50 1.00 1.00 0.50 0.50 1.00 1.00 81.25 15 1.00 0.00 1.00 1.00 1.00 1.00 1.00 1.00 87.50 1.00 0.50 1.00 1.00 1.00 1.00 1.00 1.00 93.75 16 1.00 0.00 0.00 0.50 1.00 0.00 1.00 1.00 56.25 1.00 0.50 0.00 0.50 1.00 0.50 1.00 1.00 68.75 AVG: 0.94 0.25 0.75 0.81 0.75 0.13 0.94 1.00 69.53 0.94 0.56 0.75 0.81 0.88 0.28 0.94 1.00 76.95 w.r.t. Gold Standard w.r.t. Silver Standard • Most seemed to have problems with content negotiation. They looked for DBpedia resources via a search engine, but copies the page’s URI instead of that of the resource. • Only two found the correct resource in GeoNames, and one participant knew it was tricky. Others often chose the URI corresponding to Dublin City. • Some misread the tasks and continued creating links with DBpedia instead of within (task4) or GeoNames (task6) • What if we gave ”partial credit” with a silver standard?
- 12. Results: Task Performance 10/23/19 christophe.debruyne@adaptcentre.ie 14 Participant Task1 Task2 Task3 Task4 Task5 Task6 Task7 Task8 Percentage Task1 Task2 Task3 Task4 Task5 Task6 Task7 Task8 Percentage 1 1.00 0.00 1.00 1.00 0.00 0.00 1.00 1.00 62.50 1.00 0.50 1.00 1.00 0.50 0.00 1.00 1.00 75.00 2 1.00 0.00 1.00 1.00 0.00 0.00 1.00 1.00 62.50 1.00 0.50 1.00 1.00 0.50 0.00 1.00 1.00 75.00 3 1.00 0.00 0.00 1.00 1.00 0.00 1.00 1.00 62.50 1.00 0.50 0.00 1.00 1.00 0.50 1.00 1.00 75.00 4 1.00 0.00 1.00 0.75 1.00 0.00 1.00 1.00 71.88 1.00 0.50 1.00 0.75 1.00 0.00 1.00 1.00 78.13 5 1.00 0.00 1.00 0.25 1.00 0.00 1.00 1.00 65.63 1.00 0.00 1.00 0.25 1.00 0.00 1.00 1.00 65.63 6 1.00 1.00 1.00 1.00 1.00 0.00 1.00 1.00 87.50 1.00 1.00 1.00 1.00 1.00 0.00 1.00 1.00 87.50 7 0.50 1.00 1.00 0.75 1.00 0.00 1.00 1.00 78.13 0.50 1.00 1.00 0.75 1.00 0.00 1.00 1.00 78.13 8 1.00 1.00 0.00 0.75 1.00 0.00 1.00 1.00 71.88 1.00 1.00 0.00 0.75 1.00 0.00 1.00 1.00 71.88 9 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 75.00 1.00 0.00 1.00 1.00 1.00 0.00 1.00 1.00 75.00 10 0.50 0.00 1.00 1.00 1.00 0.00 1.00 1.00 68.75 0.50 0.50 1.00 1.00 1.00 0.50 1.00 1.00 81.25 11 1.00 0.00 0.00 1.00 0.00 0.00 1.00 1.00 50.00 1.00 0.50 0.00 1.00 0.50 0.50 1.00 1.00 68.75 12 1.00 1.00 1.00 0.00 1.00 1.00 1.00 1.00 87.50 1.00 1.00 1.00 0.00 1.00 1.00 1.00 1.00 87.50 13 1.00 0.00 1.00 1.00 1.00 0.00 0.00 1.00 62.50 1.00 0.50 1.00 1.00 1.00 0.00 0.00 1.00 68.75 14 1.00 0.00 1.00 1.00 0.00 0.00 1.00 1.00 62.50 1.00 0.50 1.00 1.00 0.50 0.50 1.00 1.00 81.25 15 1.00 0.00 1.00 1.00 1.00 1.00 1.00 1.00 87.50 1.00 0.50 1.00 1.00 1.00 1.00 1.00 1.00 93.75 16 1.00 0.00 0.00 0.50 1.00 0.00 1.00 1.00 56.25 1.00 0.50 0.00 0.50 1.00 0.50 1.00 1.00 68.75 AVG: 0.94 0.25 0.75 0.81 0.75 0.13 0.94 1.00 69.53 0.94 0.56 0.75 0.81 0.88 0.28 0.94 1.00 76.95 w.r.t. Gold Standard w.r.t. Silver Standard
- 13. PSSUQ: Perceived Usability • 19 Questions • System’s usefulness questions 1 to 8 • Information quality questions 9 to 15 • Interface quality questions 16 to 18 • Overall 1 to 19 • Inverted Likert scale from 1 completely agree (positive) to 7 completely disagree (negative). There was room for comments. 10/23/19 christophe.debruyne@adaptcentre.ie 15
- 14. PSSUQ: Perceived Usability 10/23/19 christophe.debruyne@adaptcentre.ie 16 Most values are close to 3, which is not terrific but leans towards a favorable impression. Only exception is question 9.
- 15. PSSUQ: Perceived Usability 10/23/19 christophe.debruyne@adaptcentre.ie 17 Question 9: “The system gave error messages that clearly told me how to fix problems.” 3 participants did not provide an answer.
- 16. Results 10/23/19 christophe.debruyne@adaptcentre.ie 19 • Task performance vs. system usefulness with p = 0.0061 • Task performance vs. overall with p = 0.019 • The data thus seem to indicate that as task performance goes up, the perceived system’s usefulness and overall usability are more appreciated.
- 17. Conclusions and Future Work • Linked Data interlinking requires tools that subject matter experts can use. The SotA, however, often looks at the (semi-)automatic creation of links. In this paper, we wanted to investigate whether maps can be useful for linking data with a geospatial component. • We proposed an approach and, while limited, data seem to indicate the approach is viable. Data indicates that Linked Data principles seem to cause problems. • Future work includes the support for GML and other geospatial vocabularies (implementation) and the inclusion of the tool in a wider governance scenario (research). 10/23/19 christophe.debruyne@adaptcentre.ie 20
- 18. References • Debruyne, C., Meehan, A., Clinton, E., McNerney, L., Nautiyal, A., Lavin, P., O’Sullivan, D.: Ireland’s authoritative geospatial linked data. In: d’Amato, C., Fer-nández, M., Tamma, V.A.M., Lécué, F., Cudré-Mauroux, P., Sequeda, J.F., Lange, C., He-flin, J. (eds.) The Semantic Web - ISWC 2017 - 16th International Semantic Web Con-ference, Vienna, Austria, October 21-25, 2017, Proceedings, Part II. Lecture Notes in Computer Science, vol. 10588, pp. 66–74. Springer (2017). • McKenna, L., Debruyne, C., O’Sullivan, D.: Understanding the position of information professionals with regards to linked data: A survey of libraries, archives and museums. In: Chen, J., Gonçalves, M.A., Allen, J.M., Fox, E.A., Kan, M., Petras, V. (eds.) Proceedings of the 18th ACM/IEEE on Joint Conference on Digital Libraries, JCDL 2018, Fort Worth, TX, USA, June 03-07, 2018. pp. 7–16. ACM (2018). • Mori, G., Paternò, F., Santoro, C.: CTTE: support for developing and analyzing task models for interactive system design. IEEE Trans. Software Eng. 28(8), 797–813 (2002). 10/23/19 christophe.debruyne@adaptcentre.ie 21