Hacker News new | comments | show | ask | jobs | submit login
African Bus Routes Redrawn Using Cell-Phone Data (technologyreview.com)
60 points by tarunmitra 1515 days ago | hide | past | web | 33 comments | favorite

Why not "Ivory Coast Bus Routes Redrawn using Cell-Phone Data" ?

As someone living in an African country I find this headline beyond ridiculous. If this was in Lebanon would we be reading "Asian Bus Routes Redrawn using Cell-Phone Data"?

Same way with many news sources America is the USA.

Or Vilnius is Europe, yet St. Petersburg isn't.

Why? You mean St. Petersburg, Florida, or you mean that the Russian St. Petersburg is not part of Europe in the common American conception of it?

Please, tell us more about how article headlines ruin your day.

Because the headline is written according to the expected understanding of the American reader.

"Ivory Coast" is about as descriptive to an American audience as "Africa". I would be shocked if more than a tiny minority of Americans could tell you where the Ivory Coast was outside of "Africa", and I bet many wouldn't even know to place it in Africa, thinking maybe it's in India or SE Asia or something.

It's written like that because Americans don't know and don't really care about African geography.

>If this was in Lebanon would we be reading "Asian Bus Routes Redrawn using Cell-Phone Data"?

We'd probably see "Middle-East", the understood name for that region. We might even see "Lebanon" since its close to Israel and Americans like to feel involved in that scene.

>Because the headline is written according to the expected understanding of the American reader.

Not exactly a high bar it seems...

Can you tell me where every country in the world is?

Can you use Google? :) Not meaning to offend! As someone who grew up in an African nation, I will easily concede to some of the parents in this thread that it's often just easier to get people's attention if you use something recognizable in a news headline. That being said, it is odd to me that it's the MIT Technology Review doing it in an age when anyone can double-click a string of text in the browser, e.g. Ivory Coast, and use the right-click menu to search with Google. Or copy and paste into their favorite search engine and instantly find out where the place is located with a string of search results and a map embedded into the page...

Here's the real kicker: if I don't know where the country is, why would I care enough to google it? "Africa" in this situation gives me more info than the name of the country would.

You never look up things you don't recognize? I care to look things up every day, and before the web I used physical dictionaries and encyclopedias on a frequent basis. Some people may never feel that inquisitive, but quite a lot of us are, and I'd imagine most people on HN are like that.

There's a lot of articles posted to HN every day that I don't understand. If I researched them all, I'd have no time to actually do something productive. I skip them and move on unless something else about it catches my eye.

Most of them, yes. Most 12 year olds from my country are taught the names of all countries in the world and where to find them on a map and the names of the capital cities.

The I guess I'm less intelligent than most of the 12 year old kids in your country, because to be honest there's more relevant things I still have yet to learn before I get around to memorizing the name, location, and capital city of all 195(?) nations in the world and keep this information updated in my mind when, for example, a regime change in Central Africa means the name of the country is now different or what was one country is now two.

I should have your middle schoolers tutor me.

>I guess I'm less intelligent than most of the 12 year old kids in your country

I think you just hit the nail on the head.

It depends on the cultural consciousness, and whether that region has been subject to major news for recent generations. The Middle East and much of North-east Africa has has had the (mis)fortune of being in the US media for some time. For example, most US residents will know at least generally where Afghanistan is (some will know REALLY well), while I expect most don't really know where Luxembourg or Belarus are.

I wonder if this was done in, say Poland, would it be titled "European Bus Routes Redrawn Using Cell-Phone Data"?

Why is it no-one narrows down descriptions of Africa to below continent-level?

Or possibly "Government monitoring citizens movements by cell phone tracking" if done elsewhere.

Because realistically, "Africa" is most useful to most readers, even on HN.

Is anyone else slightly annoyed that there is no imagery to show this data on a map? That's all I was actually looking for when I clicked the link. Had a quick look for it and can't see anything. Can anyone else help with a link for the lazy?

Yes, the article is essentially disappointing without an image of a map showing the current bus routes versus traffic/cell data versus proposed. Are they trying to promote interest in this project by creating a little buzz?

Wake me when there is one.

The paper is being released later this week by the look of it.

They talk about 10% time reduction(probably average). Does anybody know more fine grained statistics ?

Hi, I'm from the IBM Research team who worked on this.

The reduction is in total system-wide journey times in passenger-minutes (travel time + expected waiting time). The model makes underlying assumptions on how people arrive at stops to do their trips, and how they choose paths through the network. The wait times are therefore related to frequency of services.

The statistic is also broken down by each route what happens when the changes (new routes or extensions) are implemented. For some routes, you see increase in ridership (on account of better connectivity), and on some you see decreased ridership (since the new routes offer more competitive paths through the network).

Thanks. Really interesting work.

Are you allowed to share some of the data ? If so:

Does the time includes time to walk to bus station or an estimate?

Is there some histogram of time saved per passenger ? or even something like X% of passengers save a big amount of time ?

Do you have a way to estimate time save by adding routes or alternative transport(like jitneys) before making the change?

Those are interesting, but i also think if some of this data was available(like for example big time saving for some part of the population), this would have better media coverage.

Yes travel time includes time to walk from the Cell tower (considered the "source" and "sink") to the stops.

We could generate passenger level stats for the sample we observed, but have presented it at the route level for now, since its targeted towards operators.

Yes, the improvements reported are from adding of new routes (the article title suggests that we "redrew" current routes). The new routes are generated from "frequent" patterns we see in the data. Haven't looked specifically at alternative modes, although there are several reasons to discourage jitneys.

Based on the research, how different is the proposed more-efficient bus routes from the already existing bus routes?

Or more to the point, how correct (or incorrect) were the urban planners in their choice of bus routes?

More information like how large the bus route system had grown since its inception -- for example did it grow organically over time in small growth spurts or was it in larger controlled route expansions/reductions -- would be useful as well.

Interesting as long as you don't ignore a whole demographic (e.g. very poor people without phones).

The population of Cote d'Ivoire is 20 million, Orange have released data from 5 million customers nationwide and they are one of 6 mobile phone operators. That suggests that mobile penetration is quite high in Cote d'Ivoire. Go to Abidjan and you'd see empirical evidence to support this theory too.

Cote d'Ivoire is a comparatively prosperous country in West Africa.

Welcome to the 21st century. Cell phones are a dirt-cheap commodity, not a luxury.

If you can't afford a second-hand feature phone, you likely can't afford to ride the bus.

I read yesterday that more people in India have access to a cell phone than to a toilet or running water. I assume they are very poor since they have to poop outside on the street, but they do have cell phones.

Even then, 40% of _households_ does not have a phone (http://www.nydailynews.com/life-style/health/indian-toilet-s...). That is in line with http://en.wikipedia.org/wiki/List_of_countries_by_number_of_..., which states there are about 2 mobile phones for every 3 Indians.

I think it is safe to bet that, among those using public transport, mobile phone ownership will not be higher. Also, of those with a mobile on those buses, mobile phone _usage_, likely, will be higher for the richer ones.

"Access to" and "owning it oneself" are not necessarily equivalent. If a household of seven has a single cell phone, seven people have "access to" a cell phone.

Mobile phones and internet access via mobile phones is a pretty high priority, even if the other items on the list are a toilet, a scooter, or a stove. The rationale is that the phone and internet enables you to earn more and get the other items.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact