Foss4 g2019toilet search_chatbot
1 like230 views
This document summarizes the development of a public toilet search system using open data and chatbots in Tokyo, Japan. The system allows users to search for the nearest public toilet by sending their location or address to a chatbot on the LINE messaging app. The chatbot searches a Voronoi diagram built using open data on public toilet locations from Tokyo and two cities. Search results include the toilet name and a photo link. While evaluated positively, challenges remain in adding search filters, transportation directions, and internationalizing the system and data.
![We can find the nearest toilet by searching
the Voronoi diagram.
I installed the Voronoi diagram on GeoServer
If you use GetFeatureInfo on the Voronoi diagram using the location,
you can get the index information of the toilet near that location
{"type":"FeatureCollection",
"features":[{"type":"Feature","id":"boronoi.443",
"geometry":{"type":"MultiPolygon","coordinates":
[[[[139.52236293,35.686222],[139.5217894,35.68823733],
[139.52265736,35.69349495],[139.52832257,35.70079911],
[139.52863594,35.70075584],[139.53256036,35.70003536],
[139.53793244,35.69468782],[139.53787073,35.6946377],
[139.52385133,35.68644029],[139.52236293,35.686222]]]]},
"geometry_name":"the_geom",
"properties":{"kid":2877,"tbname":"park_barrier_free_wc","id":270,
"name":"野川公園 地上 6号トイレ"}}],
"totalFeatures":"unknown","numberReturned":1,
"timeStamp":"2019-08-17T06:12:48.523Z",
"crs":{"type":"name","properties":{"name":"urn:ogc:def:crs:EPSG::4326"}}}
GetfeatureInfo (query with location)
Query result( index of toilet data record)](https://image.slidesharecdn.com/foss4g2019toiletsearchchatbot-191017080545/85/Foss4-g2019toilet-search_chatbot-16-320.jpg)
![The toilet index information contains the
name of the seat to be referenced and the
index of the data in it
Different sheets have different attribute data
structures
Use this information to get information about
public toilets that users can see
You can use this method to return search
results even if the data structure is different
{"type":"FeatureCollection",
"features":[{"type":"Feature","id":"boronoi.443",
"geometry":{"type":"MultiPolygon","coordinates":
[[[[139.52236293,35.686222],[139.5217894,35.68823733],
[139.52265736,35.69349495],[139.52832257,35.70079911],
[139.52863594,35.70075584],[139.53256036,35.70003536],
[139.53793244,35.69468782],[139.53787073,35.6946377],
[139.52385133,35.68644029],[139.52236293,35.686222]]]]},
"geometry_name":"the_geom",
"properties":{"kid":2877,"tbname":"park_barrier_free_wc","id":270,
"name":"野川公園 地上 6号トイレ"}}],
"totalFeatures":"unknown","numberReturned":1,
"timeStamp":"2019-08-17T06:12:48.523Z",
"crs":{"type":"name","properties":{"name":"urn:ogc:def:crs:EPSG::4326"}}}](https://image.slidesharecdn.com/foss4g2019toiletsearchchatbot-191017080545/85/Foss4-g2019toilet-search_chatbot-17-320.jpg)
More Related Content
Similar to Foss4 g2019toilet search_chatbot (20)
More from Yoichi Kayama (20)
Foss4 g2019toilet search_chatbot
- 1. Development of a public toilet search system using open data and chatbots Yoichi Kayama OSGeo Foundation Japan Chapter Aero Asahi Corporation
- 2. Self introduction ? OSGeo.JP member ? OSGeo.org Charter member ? A member of GIS Association of Japan (Also a member of FOSS4G SIG) ? A Coordinator of QGIS Japanese translation team ? GIS Programer and Researcher at Aeroasahi Corporation Twitter ID @pokopen ? A member of ITDART ? A member of Code for Japan
- 3. The Tokyo Metropolitan government publishes a lot of their data
- 4. Since 2017, Tokyo Metropolitan government has been hosting a contest to compete in how to use open data every year.
- 5. At 2017,Several themes were presented by the organizers at the contest. There was a theme to make it easier to find public toilets
- 6. I thought that a toilet search system could be created using the FOSS4G tool if toilet location data was published
- 7. There were multiple toilet location data in the public data archive in Tokyo Metropolitan government
- 8. I found 4 toilet data in the Tokyo archive Toilet data for parks and public facilities owned by Tokyo metropolitan government. Data on public toilets owned by Shinagawa City and Taito City(Special wards in Tokyo)
- 9. Each data had coordinates. But those data had different attribute data formats
- 10. Items of public toilet information which in the public facilities and parks owned by the Tokyo Metropolitan Government Name of facility floor Toilet name latitude longitude Coordinate system Existence of taxiway Position navigation using voice Door type Wheelchair access turn in a wheelchair Backrest handrail Equipment for ostomate Large bed Facilities for changing infants' diapers Baby chair Picture URL Remarks
- 11. There are links to the photos of the toilet in the data
- 12. Items of public toilet information which in the public facilities and parks owned by the Shinagawa City Name of facility Name of facility （In English) Number of men's toilets Number of wemen's toilets Number of unisex toilets Number of barrier-free toilets Baby bed Equipment for ostomate Open time Close Time latitude longitude Picture URL Description Description (English)
- 13. Items of public toilet information which in the public facilities and parks owned by the Taito City Major classification Small classification name Address phone number Toilet information latitude longitude
- 14. I created index data for searching toilets and attribute data attached to it separately Public toilet index With location data Toilets attribute data of Tokyo Metropolitan Gov Toilets attribute data of Shinagawa city Toilets attribute data of Taito city
- 15. I created a Voronoi diagram using the toilet location as an index to search for public toilets
- 16. We can find the nearest toilet by searching the Voronoi diagram. I installed the Voronoi diagram on GeoServer If you use GetFeatureInfo on the Voronoi diagram using the location, you can get the index information of the toilet near that location {"type":"FeatureCollection", "features":[{"type":"Feature","id":"boronoi.443", "geometry":{"type":"MultiPolygon","coordinates": [[[[139.52236293,35.686222],[139.5217894,35.68823733], [139.52265736,35.69349495],[139.52832257,35.70079911], [139.52863594,35.70075584],[139.53256036,35.70003536], [139.53793244,35.69468782],[139.53787073,35.6946377], [139.52385133,35.68644029],[139.52236293,35.686222]]]]}, "geometry_name":"the_geom", "properties":{"kid":2877,"tbname":"park_barrier_free_wc","id":270, "name":"野川公園 地上 6号トイレ"}}], "totalFeatures":"unknown","numberReturned":1, "timeStamp":"2019-08-17T06:12:48.523Z", "crs":{"type":"name","properties":{"name":"urn:ogc:def:crs:EPSG::4326"}}} GetfeatureInfo (query with location) Query result( index of toilet data record)
- 17. The toilet index information contains the name of the seat to be referenced and the index of the data in it Different sheets have different attribute data structures Use this information to get information about public toilets that users can see You can use this method to return search results even if the data structure is different {"type":"FeatureCollection", "features":[{"type":"Feature","id":"boronoi.443", "geometry":{"type":"MultiPolygon","coordinates": [[[[139.52236293,35.686222],[139.5217894,35.68823733], [139.52265736,35.69349495],[139.52832257,35.70079911], [139.52863594,35.70075584],[139.53256036,35.70003536], [139.53793244,35.69468782],[139.53787073,35.6946377], [139.52385133,35.68644029],[139.52236293,35.686222]]]]}, "geometry_name":"the_geom", "properties":{"kid":2877,"tbname":"park_barrier_free_wc","id":270, "name":"野川公園 地上 6号トイレ"}}], "totalFeatures":"unknown","numberReturned":1, "timeStamp":"2019-08-17T06:12:48.523Z", "crs":{"type":"name","properties":{"name":"urn:ogc:def:crs:EPSG::4326"}}}
- 18. Public toilet search mechanism I created Location info from GNSS Address String Geocoding Search to Voronoi polygon using location information Get a Public toilet information index Public toilet information Google Spread Sheet (Tokyo Metropolitan Government) Public toilet information Google Spread Sheet (Shinagawa city) Public toilet information Google Spread Sheet (Taito City) Result Toilet information
- 19. I had to create a user interface for this mechanism Location info from GNSS Address String Geocoding Search to Voronoi polygon using location information Get a Public toilet information index Public toilet information Google Spread Sheet (Tokyo Metropolitan Government) Public toilet information Google Spread Sheet (Shinagawa city) Public toilet information Google Spread Sheet (Taito City) Result Toilet information
- 20. What is required of the user interface of this system The ability to use this function on a smartphone Usage is not complicated
- 21. I thought it would be nice to create a toilet search system using the message application for smartphones that many people use
- 22. There is a chat system called LINE with many users in Japan
- 23. Many users use LINE on smartphones
- 24. We can post text, location information, etc, in LINE chat. location Text picture etc
- 25. In LINE, we can create a program called chatbot and let the program respond to users Chatbot Program Send information Response Send information Response
- 26. I thought that if I sent location information to a chatbot on LINE, I could search for nearby toilets using that location and display the results. Search to Voronoi polygon using location information Get a Public toilet information index Public toilet information Google Spread Sheet (Tokyo Metropolitan Government) Public toilet information Google Spread Sheet (Shinagawa city) Public toilet information Google Spread Sheet (Taito City) Result Toilet information Input location Display result chatbot
- 27. I have prepared two ways to enter location information
- 28. The first location information input method is to send the current location information acquired using the smartphone's GNSS to the chatbot.
- 29. The second location input method is to send the address text to the chatbot Enter the address string Entered address Searched toilet information
- 30. System configuration Search to Voronoi polygon using location information Get a Public toilet information index Public toilet information Google Spread Sheet (Tokyo Metropolitan Government) Public toilet information Google Spread Sheet (Shinagawa city) Public toilet information Google Spread Sheet (Taito City) Result Toilet information Input location From GNSS chatbot Input location Address String
- 31. Toilet data displayed as search results.
- 32. Click the link of photo
- 33. Public toilet search system with chat system(LINE) as user interface
- 34. What was the result of the contest? I'm not in the photo of the award
- 35. What was the result of the contest? However, my system was highly evaluated by the judges for the location information search system using the chat system.
- 36. Problems and challenges ? Conditional search is not implemented ? The use of transportation means should be included in the search for nearby toilets ? Not internationalized（Most of the data is only in Japanese） ? There are few toilets in the data of Tokyo and the data of two cities （This year, data including toilet data of railway operators was released）
- 37. New Application using LINE ? System development using LINE is increasing in Japan ? I have created an information reporting system when a disaster occurs using the LINE ? Building a system using LINE can reduce the coding effort for the user interface and communication
- 38. Conclusion ? If we have data with location information and FOSS4G tool, it is easy to create a search tool ? We can easily create a data management mechanism using cloud services such as Google Spread Sheet. ? We can easily create a user interface with Line chatbots. In particular, we can easily create a user interface for a system using a smartphone with a LINE chatbot. ? It is difficult for us to internationalize the system unless usage data is internationalized
- 39. Thank you ! Question ?