Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
El 19/04/2010 16:56, Jonas Gabriel escribió: Every person creates personal routes based on his knowledge of his living area.[...] Then a service could produce routes by using members of this alternative routing graph forest connected by a traditional routing service [...] I would like to hear some of your comments.Does it make any sense? Yeah, it makes sense and would be doable by lowering the weights of the graph arcs in the routing algorithm for every uploaded route, or add a new arc for every uploaded route, with a lowered weight. However, besides from being a cool research project... what problem does it *solve*? Why would *I* be interested in uploading routes to such a service? Cheers, -- Iván Sánchez Ortega i...@sanchezortega.es ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
Jonas, Jonas Gabriel wrote: I am in search of a Master Thesis' project idea in the field of Location Based Applications/Services. Last night I came across OpenRouteService.org http://www.openrouteservice.org/ and had an epiphany about a routing service based on social contribution.Let me clear out my idea: Peter Miller has done a talk at the 2008 State of the Map conference (in Limerick) about something in that direction; he explained that different people may favour different routes for very different reasons and that it could be a good idea to ask: What route would people like me use? Personally I would probably not make it as difficult as you; I would simply allow individual weights for bits of the routing graph. I.e. you would not upload a route; instead you would have the system compute the route it thinks is best, and then you'd say: Nah, I don't like that because it takes me along a busy road. I'll add a 50% penalty to that road and see where it takes me now. Oh well it doesn't use that bit of footway there but it really should, so I'll give that a bonus, etc. Anyone could then use the routing engine either with the normal weights, or with a combination of weights of different users. But still this might be too big for a Master's thesis. Bye Frederik ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
Frederik Ramm schrieb: Personally I would probably not make it as difficult as you; I would simply allow individual weights for bits of the routing graph. I.e. you would not upload a route; instead you would have the system compute the route it thinks is best, and then you'd say: Nah, I don't like that because it takes me along a busy road. I'll add a 50% penalty to that road and see where it takes me now. Oh well it doesn't use that bit of footway there but it really should, so I'll give that a bonus, etc. This is done using vias in google maps right now. Just drag'n'drop from the calculated route.. Peter ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
El 20/04/2010 13:02, Peter Körner escribió: Nah, I don't like that because it takes me along a busy road. I'll add a 50% penalty to that road and see where it takes me now. This is done using vias in google maps right now. Just drag'n'drop from the calculated route.. No, you cannot avoid stuff by dragging in GMaps. You can, however, use OpenRouteService for that :-) -- Iván Sánchez Ortega i...@sanchezortega.es ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
@Ivan: I agree. The more people using my favourite route, the slower I will get forward, so why should I share? Answer: not at all if I know which informations my actions do supply to the service. But: tell the user to make a perfect route for her purpose if she supplies the service with detail information such as car type, personal driving style, comfort preference, urgency, and, of course, starting time and personally expected arrival time. And don't forget about the feedback next time she logs in again. (think carefully about the validity of the result) last.fm names this 'skobbling' and gives you recommendations, groups user types and creates correlations to form groups by similarity of preferences. The idea is not bad, but I don't expect it to grow successful, especially because the project's scope is short-term. You need a deep breath to get it all sexy and charming enough for average people to use. Without a wide userbase such a service won't get any value. Creating the toolset services like this could be based upon looks promising, though. Thomas Original-Nachricht Datum: Tue, 20 Apr 2010 11:37:02 +0200 Von: Iván Sánchez Ortega i...@sanchezortega.es An: dev@openstreetmap.org Betreff: Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments El 19/04/2010 16:56, Jonas Gabriel escribió: Every person creates personal routes based on his knowledge of his living area.[...] Then a service could produce routes by using members of this alternative routing graph forest connected by a traditional routing service [...] I would like to hear some of your comments.Does it make any sense? Yeah, it makes sense and would be doable by lowering the weights of the graph arcs in the routing algorithm for every uploaded route, or add a new arc for every uploaded route, with a lowered weight. However, besides from being a cool research project... what problem does it *solve*? Why would *I* be interested in uploading routes to such a service? Cheers, -- Iván Sánchez Ortega i...@sanchezortega.es ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev -- http://www.openstreetmap.org/?lat=47.172lon=7.4395zoom=15layers=B000FTFTTmlat=47.16696mlon=7.43509 GRATIS für alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01 ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
It sounds good. Anyone who knows about routing knows that it takes a crazy long time to be sure of the perfect route. This could be a way for the program to say I know the best way to get most of the way, because Jane told me she goes this way.. Adding a load of sample routes in one area, and with a good weighting on provided routes, it would be interesting to see how the program faired to other routing algorithms that. Are you going to look at motoring routing? Walking routing might be just as interesting (and more easy to get people's routes added). I walk around university a lot, sometimes two routes look exactly the same length but I take the route through a park because it's nicer and makes the journey feel shorter. Sometimes I go one way because it just has a few steps, returning I go another way because it's downhill. On 20 April 2010 10:00, Thomas Meller thomas.mel...@gmx.net wrote: @Ivan: I agree. The more people using my favourite route, the slower I will get forward, so why should I share? Answer: not at all if I know which informations my actions do supply to the service. But: tell the user to make a perfect route for her purpose if she supplies the service with detail information such as car type, personal driving style, comfort preference, urgency, and, of course, starting time and personally expected arrival time. And don't forget about the feedback next time she logs in again. (think carefully about the validity of the result) last.fm names this 'skobbling' and gives you recommendations, groups user types and creates correlations to form groups by similarity of preferences. The idea is not bad, but I don't expect it to grow successful, especially because the project's scope is short-term. You need a deep breath to get it all sexy and charming enough for average people to use. Without a wide userbase such a service won't get any value. Creating the toolset services like this could be based upon looks promising, though. Thomas Original-Nachricht Datum: Tue, 20 Apr 2010 11:37:02 +0200 Von: Iván Sánchez Ortega i...@sanchezortega.es An: dev@openstreetmap.org Betreff: Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments El 19/04/2010 16:56, Jonas Gabriel escribió: Every person creates personal routes based on his knowledge of his living area.[...] Then a service could produce routes by using members of this alternative routing graph forest connected by a traditional routing service [...] I would like to hear some of your comments.Does it make any sense? Yeah, it makes sense and would be doable by lowering the weights of the graph arcs in the routing algorithm for every uploaded route, or add a new arc for every uploaded route, with a lowered weight. However, besides from being a cool research project... what problem does it *solve*? Why would *I* be interested in uploading routes to such a service? Cheers, -- Iván Sánchez Ortega i...@sanchezortega.es ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev -- http://www.openstreetmap.org/?lat=47.172lon=7.4395zoom=15layers=B000FTFTTmlat=47.16696mlon=7.43509 GRATIS für alle GMX-Mitglieder: Die maxdome Movie-FLAT! Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome01 ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev -- Gregory o...@livingwithdragons.com http://www.livingwithdragons.com ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
[OSM-dev] Open Social Routing Idea: Request for preliminary comments
Dear community, I am in search of a Master Thesis' project idea in the field of Location Based Applications/Services. Last night I came across OpenRouteService.orghttp://www.openrouteservice.org/and had an epiphany about a routing service based on social contribution.Let me clear out my idea: Every person creates personal routes based on his knowledge of his living area.For example Mr John Doe uses RouteX to get from LocationA to LocationB because between 15:00 and 21:00 traffic is low. This kind of routes could be uploaded at OpenStreetMap and piece-by-piece create an alternative routing graph forest (probably by using a tagging system) based on social contribution. Then a service could produce routes by using members of this alternative routing graph forest connected by a traditional routing service such as OpenRouteService.org http://www.openrouteservice.org/. An example: Let say Jane Does wants to go from LocationA to LocationD at 16:00. The system could instruct her to follow John Doe's RouteX from LocationA to LocationB, then use OpenRouteService.org to go from LocationB to LocationC and finaly use John Smith's RouteZ from LocationC to LocationD. I would like to hear some of your comments.Does it make any sense? Do you think is feasible as a Master Thesis' project, or is it too big, too small ? ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
Jonas Gabriel wrote: Every person creates personal routes based on his knowledge of his living area.For example Mr John Doe uses RouteX to get from LocationA to LocationB because between 15:00 and 21:00 traffic is low. This kind of routes could be uploaded at OpenStreetMap and piece-by-piece create an alternative routing graph forest (probably by using a tagging system) based on social contribution. Then a service could produce routes by using members of this alternative routing graph forest connected by a traditional routing service such as OpenRouteService.org http://www.openrouteservice.org/. An example: Let say Jane Does wants to go from LocationA to LocationD at 16:00. The system could instruct her to follow John Doe's RouteX from LocationA to LocationB, then use OpenRouteService.org to go from LocationB to LocationC and finaly use John Smith's RouteZ from LocationC to LocationD. My main concern would be the loading of personal and arbitrary routes into OSM. OSM routes tend to be verifiable routes that are marked 'on the ground', such as a bus route (from the route numbers on bus stops) or a cycle route (from the signs that show the way). They are there to help people find it again. Bill's route to work does not fit this style. If you load the routes elsewhere and use OSM as the background it might be interesting, and I can see that it might produce a powerful routing engine based on local knowledge and time-based traffic patterns, if you can get enough people to contribute. Cheers, Chris ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
On Mon, Apr 19, 2010 at 5:12 PM, Chris Hill o...@raggedred.net wrote: My main concern would be the loading of personal and arbitrary routes into OSM. +1 but if we find city tour from travel agencies, is that less arbitrary than personal routes ? Creating route relations in OSM is for me a more general issue, especially when you see how it is done today with relations and segmented roads. Pieren ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Open Social Routing Idea: Request for preliminary comments
On Mon, Apr 19, 2010 at 5:12 PM, Chris Hill o...@raggedred.net wrote: If you load the routes elsewhere and use OSM as the background it might be interesting, and I can see that it might produce a powerful routing engine based on local knowledge and time-based traffic patterns, if you can get enough people to contribute. I think this would be an interesting project if done this way. Of course then there's a question of What if your route includes public transportation?- or maybe that's simply outside the scope? - Serge ___ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev