a survey of open accessibility data
DESCRIPTION
Presentation for w4a 2014TRANSCRIPT
A Survey of Open Accessibility Data
Chaohai Ding, Mike Wald, Gary Wills
{cd8e10, mw, gbw}@ecs.soton.ac.uk
Web and Internet Science Group, ECS
University of Southampton, UK
Mash-up Real World Places with Accessibility Data
W4A '14, April 07-09, 2014, Seoul, Republic of Korea
Outline
• Motivation
• Open Accessibility Data
• Lessons Learned
• Conclusion
http://sydycster.blogspot.pt/2013/01/tanggung-jawab-mahasiswa.html
Motivation
Travelling is listed as one of top difficulties for people with limitations, impairments or disabilities.
• Pre-trip (route planning with accessibility data)
• On-trip (personal assistance, real time data and accessibility data of surrounding environment)
• After-trip (Evaluation, review or complains)
http://www.wlv.ac.uk/files/images/knowledge4/mind-the-gap.jpg
Data for Accessible Travelling
• Open Transport Data (stations, public transport data and real time traffic data);
• Other Real Time Data (weather, events or roadwork);
• Open Accessibility Data:
– Open data that can improve the accessibility experience;
– Data related to accessibility information (toilet, lifts, ramp, wheelchair space in bus or trains etc.);
– Data can help people with special needs.
http://innovations.tamuc.edu/wp-content/uploads/2011/10/Accessibility.jpg
A Survey of Open Accessibility Data
National Rail AccessTogether
Wheelchair Access (wheelmap.org)
Categories Yes No Limited Unknown
Transfer 1445 (0.59%) 272 (0.11%) 158 (0.06%) 244094 (99.24%)
Food 1038 (1.82%) 367 (0.64%) 430 (0.75%) 55135 (96.78%)
Leisure 100 (1.7%) 13 (0.22%) 27 (0.46%) 5744 (97.62%)
Bank 351 (2.25%) 43 (0.28%) 33 (0.21%) 15204 (97.27%)
Education 132 (1.12%) 15 (0.13%) 32 (0.27%) 11597 (98.48%)
Shopping 767 (2.22%) 118 (0.34%) 123 (0.36%) 33552 (97.08%)
Sport 19 (1.01%) 0 13 (0.69%) 1851 ((98.30%))
Tourism 88 (0.76%) 45 (0.39%) 15 (0.13%) 11459 (98.72%)
Accommodation 66 (0.94%) 35 (0.50%) 13 (0.18%) 6940 (98.30%)
Misc 484 (2.28%) 78 (0.37%) 35 (0.16%) 20648 (97.19%)
Government 26(1.34%) 3 (0.15%) 2 (0.10%) 1910 (98.40%)
Health 171 (2.39%) 12 (0.17%) 17 (0.24%) 6946 (97.20%)
All nodes 4687 (1.11%) 1001 (0.24%) 898 (0.21%) 415080 (98.44%)
Restaurants-UK (Factual)
Wheelchair Accessible
Yes No Blank
Restaurants-UK 8904 (4.23%) 1786 (0.85%) 199923 (94.92%)
Tube Station Step Free Guide (TfL)
Items Yes No Unknown
Blue Badge Car Park Spaces
147 (40.61%) 35 (9.67%) 180 (49.72%)
Taxi Ranks Outside Station
13 (3.59%) 169 (46.69%) 180 (49.72%)
Specific Entrance 16 (4.42%) 141 (38.95%) 205 (56.63%)
Accessible Toilet 54 (14.92%) 128 (35.36%) 180 (49.72%)
Access Via Lift 98 (27.07%) 84 (23.20%) 180 (49.72%)
Limited Capacity Lift 8 (2.21%) 174 (48.07%) 180 (49.72%)
Tube Station Step Free
Accessibility Data (National Rail)
Items Yes No Unknown
Ramp for train access
1726 (66.36%)
842 (32.37%)
33 (1.27%)
Step free access coverage
1347 (51.79%)
1220 (46.91%)
34 (1.31%)
Wheelchairs available
374 (14.38%)
2195 (84.39%)
32 (1.23%)
National key toilets 168
(6.46%) 623
(23.95%) 1810
(69.59%)
Accessible public telephones
152 (5.84%)
320 (12.30%)
2129 (81.85%)
Accessible ticket machines
995 (38.25%)
285 (10.96%)
1321 (50.79%)
Special Needs (National Rail)
Items Yes No Unknown
Staff help available 1199
(46.10%) 1370
(52.67%) 32
(1.23%)
Customer help points 1726
(66.36%) 843
(32.41%) 32
(1.23%)
Car Park 1548
(59.52%) 910
(34.99%) 143
(5.50%)
Public Wi-Fi 200
(7.69%) 0
2401 (92.31%)
Toilets 792
(30.45%) 1778
(68.36%) 31
(1.19%)
Baby changing facilities
365 (14.03%)
426 (16.38%)
1810 (69.59%)
National Rail
AccessTogether
• 58 specified attributes related to accessibility;
• Attributes are too complex and specific for both data reusing and crowdsourcing;
• Only 46 entities in UK, while 1147 entities in US;
• Reference to other applications (Foursquare and Factual).
Lessons Learned
• Open accessibility data is available and from multiple resources;
• Various accessibility related attributes, but no standard guideline to specify these attributes;
• Government published high quality data but might be out of date;
• The classification of accessibility data could simply focus on some specific disabilities, such as mobility (wheelchair accessible), or much more complex if it refers to special needs or all type of disabilities;
• Crowdsourcing is powerful but not enough for gathering accessibility data due to the data quality.
Interlinking Accessibility Data
Linked Open Accessibility Data Repository for Accessible Travelling and Living
• Data Integration challenges: entities matching, data conflict, dimensions deduction for matching;
• Ontology challenges: no existing ontology fits the needs for data integration (top-to-down or down-to-top ontology engineering?);
• Reasoning rules for ontology: the accessibility metrics for reasoning;
• No standard guideline for providing accessibility data to mash-up places.
Challenges
• A survey of open accessibility data in UK;
• Lessons learned;
• Applying Semantic Web and Linked Data to address information barriers for accessibility;
• Enrich and improve existing accessibility related datasets;
• Linked open accessibility data repository for users or developers to publish, consume or evaluate accessibility data.
Conclusion