Friday 8 June 2012

Who does Microsoft Academic Research think you are?

A while ago, I posted some thoughts about Microsoft Academic Research. Today, I was sent a link by a colleague about one of the features on the site - the Co-Author "Academic Degrees of Separation" path. She sent me the path between us - we're five steps apart. The only problem was, it wasn't me! The picture was me and the name was me but all the other details were wrong!

Here's the entry:

The worrying thing for me, is that I did not create this entry and yet they have somehow affiliated both my picture and one of my genuine URLs - my Southampton Computational Modelling Group homepage - with a tiny handful of wrong publications and an institution where I have never worked or visited. Perhaps more worrying, I cannot work out how or why. I can only guess that they have some kind of data-mining algorithm that has gone badly wrong. Very badly wrong!

Here's the real me:

As you can see, I have a totally different set of metrics, keywords and interests to the spurious me. My colleague is not really to blame, though. Search my name, and up pops the wrong entry complete with my actual photo before the real me:

The sooner they roll out ORCID, the better! For now, I guess I need to work out how to get the wrong Richard Edwards removed and/or merged with the right one.

For those interested in the Co-Author path, here it is for the real me to the fictitious me:

Of course, any one of these links could be through a totally incorrect entry. This whole scenario just reinforces my views of Microsoft Academic Research: they have some nice tools but until they can sort out the underlying data, it remains essentially useless and possibly worse than useless.

If you have a remotely common name, even if you have no interest in a profile on Microsoft Academic Research, go and look yourself up now! You never know what someone else might be finding out about "you".

UPDATE. I was able to log on and remove my picture and URL from that entry, although it looks like it will take a week to take effect. Hopefully, it will work...

4 comments:

  1. Maybe he's from an alternate universe where the cabbages ride BMW's instead of Harleys. . . .

    ReplyDelete
  2. Well, if nothing else, it proves my point about why I use "cabbagesofdoom" and not my actual name as my online identity. Too many "Richard Edwards" out there - only one cabbagesofdoom. (For now...)

    ReplyDelete
  3. There must be something completely wrong with the data-mining algorithm. Two papers are missing, the other two are cited as published in Protein Science though they are in Proteins: Structure, Function and Bioinformatics. There is even one unexistant publication which title is just a part of the title of another paper!
    Any reference system as unreliable as this one could do more harm than good. It definitely needs some manual curation, maybe a more social network-driven approach...


    http://academic.research.microsoft.com/Author/7164651/nicolas-palopoli

    ReplyDelete
  4. Google certainly seem to be winning this battle. Google Scholar correctly and painlessly found my last paper pretty much as soon as it was out and emailed me to say so.

    ReplyDelete

Thanks for leaving a comment! (Unless you're a spammer, in which case please stop - I am only going to delete it. You are just wasting your time and mine.)