22

I would like to do some analytics on CDR data, preferably mobile, but anything will do.
Ideally I would like to have between 100MB to 10GB of data.

I have been trying to look for this type of data on the web, so far with no luck.

Apparently a lot of interesting studies have been going around academia but I couldn't find anything to download.

Basically I want to experiment with (ideally large) CDRs dataset. As the CDR layout may vary (e.g. may contain price elements; time of call; duration; position of caller etc.), the analytics I'll do will depend on the data I'll find.

Any help is greatly appreciated.

Improve this question
10
  • 1
    If academics are publishing papers, it wouldn't hurt to ask them for their data sets. You can also ask them how much of their research is publically funded.
    – philshem ♦
    Mar 5 '14 at 15:10
  • 1
    I don't want to appear negative but I do not have a good experience in asking academics. I was looking for an insurance claim dataset a while ago and I asked help to a prof. in Barcelona who published a lot using the dataset (coming from an insurer with HQ in Barcelona). No answer. It is true that she made a life long career through that dataset...
    – Enzo
    Mar 5 '14 at 19:13
  • 1
    Enzo : as there's info about the position of the caller, it's possible that the researcher may not be able to release the data to you. It all depends on what sort of agreements they made to aquire the data. The fact that they've not responded to you could be another problem, as it means that we've got a situation where the evidence used in research isn't available. You might be able to get access to the data to validate their analysis, you'd likely have to enter into an agreement that wouldn't give you free use of the data.
    – Joe
    Mar 5 '14 at 19:49
  • 1
    @Enzo - I know it can be frustrating, but think of yourself as a data journalist digging for the story. If not one researcher, try another. It's possible to mask data so in theory a CDR can be anonymous and therefore shared.
    – philshem ♦
    Mar 6 '14 at 8:11
  • 1
    You missed this: telecomitalia.com/tit/en/bigdatachallenge/contest/dataset.html Perhaps it will open again in the future.
    – philshem ♦
    Nov 3 '14 at 14:26
8

Try with d4d challenge from orange Senegal:

http://www.d4d.orange.com/en/home

"2.5B anonymized records of 5 million mobile phones"

Improve this answer
2
  • Great news! I was aware of the initial challenge (but I got to know about it when it was too late). At least now I'm in time to apply. Thanks!
    – Enzo
    May 13 '14 at 12:49
  • 1
    Updated link? d4d.orange.com/en/Accueil
    – philshem ♦
    Mar 18 '16 at 18:37
6

The Flowminder.org project has made available anonymized West African CDRs related to the Ebola crisis. You can read more at the WorldPop page.

Here we provide version 1 Flowminder (www.flowminder.org) human mobility models for West Africa, built on WorldPop population data, to support ongoing efforts to control the ebola outbreak. Before downloading any data, please read the documention carefully as it provides details on the datasets and models provided through the links below.

enter image description here

Worldpop datasets are licensed under the Creative Commons Attribution 4.0 International License

Improve this answer
1
  • Great post. I'll look into this project. It is really interesting.
    – Enzo
    Nov 5 '14 at 11:33
4

This answer was posted by @peterm, and I think it could be useful for you project.

Call Detail Record Generator

Our call detail records (CDR) generator is developed with GEDIS Studio and is available on-line. To use it, you need to register for an account at http://www.data-generator.com and import into your private workspace the Telecom project available from the shared workspace.

Improve this answer
1
4

I wrote one Java utility to generate the data.

Utility to generate Telecom Call Detail/Data Records Generator

Improve this answer
1
  • Great utility. I've starred the project on github and I'll test it. Thanks!
    – Enzo
    Mar 18 '16 at 10:34
4

While looking for Dataset for call data in the telecom domain, I bumped into the following STRUCTURED Data samples spanning call logs/comapliants related features -

  1. https://data.dublinked.ie/dataset/b40eabb2-52db-42b1-aafb-3cfa06bac5eb/resource/e83c8cfb-e73a-4e33-8c2e-3f95cc3bba46/download/dcccustomerservicerequests010111-170713.xls

  2. http://www.iainpardoe.com/teaching/dsc433/data/Churn.xls

  3. https://data.gov.uk/dataset/telecommunications-market-quarterly-data-tables/datapackage.zip

  4. https://data.montgomerycountymd.gov/api/views/v2ei-xfce/rows.csv?accessType=DOWNLOAD

  5. http://www.sgi.com/tech/mlc/db/churn.data

I have more links, but my "reputation" at this sub-domain doesn't allow me to post more links.

Improve this answer
2
  • 1
    Thanks! Could you try to paste some additional link as reply to this comment?
    – Enzo
    Sep 25 '16 at 18:06
  • 1
    Added ! Courtesy - the upvotes :) ! Please note that the 3rd link (.zip) is more concerned with call logs & the related financial Data.
    – user1412066
    Sep 26 '16 at 5:49
3

While I had completely missed the TelecomItalia 2014 big data challenge kindly mentioned by @philshem (many thanks again to bring this to my attention), I just spotted that the dataset created for the challenge has just now been classed as Open Data and made available to the general public.

If you want to download the data, you need to get to: https://dandelion.eu

Then click on Product to select Open Data

The site requires a registration but the download of the data is free.

Thanks to all the people that have / are supporting this search!

Improve this answer
2

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.