Development of a public toilet search system using open data and chatbots
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
Title |
| |
Title of Series | ||
Number of Parts | 295 | |
Author | ||
Contributors | ||
License | CC Attribution 3.0 Germany: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
Identifiers | 10.5446/43310 (DOI) | |
Publisher | ||
Release Date | ||
Language |
Content Metadata
Subject Area | ||
Genre | ||
Abstract |
| |
Keywords |
00:00
Physical systemChatterbotPresentation of a group
00:22
Sign (mathematics)Machine codeAssociative propertyTranslation (relic)Programmer (hardware)TwitterPhysical systemFile formatAttribute grammarInformationMaxima and minimaChatterbotPrice indexVoronoi diagramDiagramPresentation of a groupNormal (geometry)Different (Kate Ryan album)Structural loadServer (computing)Doubling the cubeBuildingQuery languageAttribute grammarResultantPhysical systemInformationMetropolitan area networkAddress spaceSubject indexingWeb portalCoordinate systemOpen setUniform resource locatorFile archiverUtility softwareType theoryDiagramVapor barrierFreewareLink (knot theory)DatabaseOnline helpFlow separationForcing (mathematics)Thomas BayesOrbitAngleWordMultiplication signWebsiteGoodness of fitPrice indexProcess (computing)Total S.A.GoogolDemosceneFile formatRight angleEuler anglesPoint (geometry)Coefficient of determinationCountingPlastikkarteFerry CorstenLecture/Conference
07:28
DiagramVoronoi diagramMaxima and minimaPrice indexInformationTunisFunction (mathematics)SmartphoneUser interfacePhysical systemPlastikkarteMessage passingLine (geometry)Online chatComputer programElectronic visual displaySmith chartElectric currentChatterbotoutputAddress spaceConfiguration spaceTrigonometryLine (geometry)Uniform resource locatorInformationElectronic visual displayUser interfaceMechanism designPhysical systemPolygonAddress spaceCartesian coordinate systemComputer programmingPlastikkarteChatterbotoutputSubject indexingAttribute grammarFunctional (mathematics)Message passingResultantLink (knot theory)Variety (linguistics)Different (Kate Ryan album)SmartphoneConfiguration spaceString (computer science)Function (mathematics)MappingCopyright infringementMeasurementBitBit rateInternetworkingPlanningCommutatorData storage deviceAutomatic differentiationControl flowCore dumpRight angleTotal S.A.Row (database)WhiteboardWritingState of matterLocal ringSoftware repositoryMultiplication sign
14:21
User interfacePhysical systemDigital photographyInformationOperator (mathematics)Transportation theory (mathematics)Line (geometry)BuildingCoding theoryTelecommunicationMechanism designPoint cloudData managementChatterbotInternationalization and localizationService (economics)GoogolSmartphoneOnline chatLine (geometry)Physical system2 (number)ChatterbotComputer programmingAreaResultantSpreadsheetDatabaseCartesian coordinate systemProcess (computing)Interface (computing)Uniform resource locatorInternationalization and localizationMechanism designPoint cloudUser interfaceLevel (video gaming)Data managementMereologyTraffic reportingTransportation theory (mathematics)Metropolitan area networkVapor barrierObservational studyWritingPower (physics)Multiplication signRight angleSpeech synthesisPrisoner's dilemmaInformationGoogolMathematicsControl flowEstimatorAngleLecture/Conference
21:14
Control flowLecture/Conference
22:19
Focus (optics)Uniform resource locatorBit rateBuildingRoutingMultiplication signSmartphoneGradientPhysical systemDialectOffice suiteMappingCircleMessage passingArithmetic meanGame controllerRow (database)Error messageRight angleLine (geometry)Goodness of fitPoint (geometry)Negative numberRouter (computing)Lecture/Conference
Transcript: English(auto-generated)
00:07
Hello everyone, we are ready for our next but also the last presentation in the parallel session. I welcome Yoichi Kayama to present his work to us.
00:31
Hello This is our last presentation of this conference. So my name is Yoichi Kayama and I am a member of OSG or JP
00:41
and many are a member of GIS in Japan and a contributor of QGIS. So this Tokyo Metropolitan Government published a lot of data and
01:01
they have many data and to make a role using open data government and public people. Since 2017 they had been hosting contest to compute to how to use open data every year.
01:28
So 2017 several payments were present and there was a payment to make it easier to find public toilets
01:41
So I think I thought that the toilet such system could be created using a phosphor Z tools if there are toilet location data was published. So I search Tokyo Metropolitan Government
02:03
open data portal site maybe using a secant and and zero is three of three data, I found four toilet data set in Tokyo archive and the Metropolitan Government has their
02:23
utility building toilet and the city of Shinagawa and Taito and and Suginami, I you forgot right here has their toilet
02:42
information. Each data had coordinated so I can make a search system, but those data had different attribute as homelands. This is a homeland of
03:03
Metropolitan Government data and there was name and floor, toilet name and latitude, longitude and coordinate system. Maybe
03:24
This one. Okay. Oh, thank you and Their character or character is in Japanese, maybe you can't understand. The door type and the
03:44
and some information about barrier free information about toilet. And there are links to the picture of a toilet.
04:01
So I arrange this data in database and search system, making a search system using special information. Using a very good idea. But this is a another data at Shinagawa City.
04:29
There are very difficult So It took Metropolitan Government's data has many columns, but Shinagawa City has simple data.
04:52
And there is Taito City. Taito is nearby Akabara or Asakusa. Asakusa includes Asakusa.
05:03
So many foreign people will together. And there are many simple format, latitude and longitude address and a name of toilet. So
05:29
different format of data, how to use and normalization is
05:43
very difficult. So I made a separated attached to data at a Google spread, loading a Google spread, different seat loading. And to make index data
06:04
was it has latitude and longitude and seat name and index. So we can search toilet in the
06:20
in this index and reference different seat. I made index data using QGIS at a border diagram using a toilet location and
06:44
search for public toilets. So search a border diagram to GeoSaba. It is a server of using, we can use this server, double Ms, double Fs. So
07:08
if we have had some special location data query this GeoSaba, get future info,
07:22
query with location. And return is result is coordinates of polygon and index to toilet data, seat name and index.
07:47
The toilet index information contains the name of the seat referenced and the index of the data in it. So this is a result of get future info and
08:02
the different seats has different attribute data structure, but we get such a data using JSON. So it is easy to display it. So such mechanism I created is
08:24
address string or location info and search abroad and there are different seat and get a result. I had to create a user interface for this mechanism.
08:42
Required the ability to use this function on a smart home and usage is not complicated. I thought it would be nice to create a toilet search system using the message application for smart phones that many people use.
09:05
In Japan, cold line messaging system many users, many people using. But it is not popular in Europe or
09:21
States. Many users use the line on smart home and and they post text or a location data. And we can post text or location information and picture or movie or
09:42
voice in the line. In line, we can create a program called the chatbot. Chatbot is auto-replaining program in line. So if I can, if I
10:05
this program return many information text or HTML or movie or picture can return.
10:24
So that actually if I sent location information to a chatbot online search nearby toilets using that location and display result. So making I made this system
10:42
at PHP program. So this program published as published as a chatbot and so we can use this program input to line and
11:01
output to line this mechanism. So I have prepared two ways to enter location information. At first it's using GNSS. So using smart home and we can
11:23
use the GNSS location data. In line, click this. This is a location information and the display
11:43
maps and specified where. Then call using this latitude and latitude using and return here is a toilet information.
12:04
And the second method is address strings. Now it is used only Japanese, sorry.
12:23
Using these strings and geocoding so return such literate information we can. System configuration is using
12:40
from GNSS or your text address text and such to border polygon and get public toilet information and return. This is a very simple
13:01
data displayed such a format. Sorry, this is all Japanese. And there is a many text in
13:24
and link to toilet picture. So some people use in toilet
13:40
using a wheelchair or using this such equipment is lying baby bed. So in Japan, Japanese public toilet have
14:00
as many variety of equipment with. So some people want a toilet has baby bed or can use a wheelchair.
14:22
Once, no. So we can search in this system such a toilet with or without such equipment. So this is a
14:42
system using here, using result and we can see at a map what was the result of the contest. This is a
15:02
awards people. I'm not in. But judges comment in they comment on my system. This is a very good system, but not
15:25
good using not a good map. So if your map is excellent, I have awarded.
15:43
So problem and so I can, this system can search toilet using only special information, location.
16:03
There are no ability of a conditional search is not to implement and use of transportation means not implemented. And this is very
16:20
most important things not internationalized. So if you want to Tokyo, you maybe if you want to use this system faster, learning Japanese. So and there are a few future rights, the data of Tokyo and the data of two cities.
16:48
But this year, the Tokyo Metropolitan Government published data including the toilet data, railways operator, that data is a very
17:04
huge data. So if I implement using this data is more useful, making more useful system. And this, that system is my first system using line chatbot. So I'm, I study how to make a chatbot and, and so
17:34
now I make a new application using line chat, chatbot. The system, the part
17:45
to, this is when disaster happened, making a report from the disaster area system. Many people post
18:09
how about your area using line and this result is in Google spreadsheet and there are,
18:24
and other people can see how about there and where is disaster area. And so yesterday, Japan had a tissue area to a very big, huge line and had some disaster. So yesterday,
18:56
many people are using this system. So conclusion, we have data with the location
19:05
and it is easy to create searching tools. There are many useful tools in hospital and we can easily create a data management mechanism using cloud service,
19:21
such as Google spreadsheet, making own database is some jobs need. And we can easily create a user interface with line chatbot. The chatbot has a simple interface, but if I made a special
19:47
interface program and many user must install it is very hard. Line is many people using in Japan
20:06
and it is difficult for us to internationalize the system. Our data is not internationalized. So if we met internationalized system in Japan, such data internationalized
20:31
or make data internationalized must have our work. Thank you.
20:45
And I am not good at listening and speaking. If you have a question, please, slowly and simply. Thank you. I just noticed that he was finished without any warning almost to the second.
21:10
So no wonder Japan is so much ahead of us. So if you have any questions, we have five minutes and then we'll have a coffee break down the hall. So just give your best shot in a simple
21:21
English for Yoichi. I have the first question for you. Meanwhile, how many toilets are in Tokyo?
21:41
How many? Not exactly, but like tens, hundreds, thousands. Oh, this data has about a thousand, but maybe more than 10,000.
22:07
Have I heard correctly 10,000 toilets in Tokyo? Public toilets, right? Public. How many people live in Tokyo?
22:29
Do we have other questions? I have one question for you again.
22:52
How do you know how accurate is the location when you send it in the message?
23:00
Yeah, so you send a message with the location. How do you trust that the location is accurate? I mean, how exact is the location? I mean, can you, let's say, control or record the accuracy or the error, the horizontal error? Because we know that with a smartphone, sometimes
23:24
you can be hundreds or maybe even kilometers away if the sensor is not good enough. Or maybe if you are among high buildings, well, Tokyo is quite a good example. You'd be like in mountains there, right? So how do you record the error of the location? In Japan,
23:48
GNSS data, smartphones, location data, GNSS is only using a smartphone, GNSS data, using Wi-Fi data. So in buildings, so
24:14
sometimes very good using Wi-Fi point data is usually. But sometimes I have a Wi-Fi router
24:34
with mobile. It has illegal position data. So if I'm using
24:49
a, my mobile router is on and fairly, I acquired the location data.
25:07
My office, so it is not good position data. But posting GNSS data is in line.
25:25
There is a display maps. So we can correct about if using maps. Thank you so much, Oichi. And toilets maybe look something like not of interest. But this is,
25:46
should be on the emergency system. So you should know exactly where the toilet is. You don't have to make a mistake. All right. Thank you, Oichi.