Saturday, February 25, 2012

Microsoft Academic Search Update

Back in December I blogged about Microsoft Academic Search. My conclusion was that though it offered some innovative ways of presenting data, it was still in an early stage of development compared to the offerings from Thomson (Web of Science/Researcher ID), Elsevier (Scopus), and Google (Scholar and Citations).

I recently received an e-mail from the project lead for Microsoft Academic Search. They have merged my various profiles. The number of citations is now 2109 which is similar to the number in Scopus. There are still, however, a lot of publications in the profile that aren't mine. It also says I have collaborated with 147 co-authors. I'd be surprised if the number was greater than 37. Based on Google that is roughly the number. The large number must be because of the extraneous publications. People like me are problematic for disambiguation algorithms of this type as we move around a lot.

Though they didn't point this out to me, it seems that there is a way to correct the record. I submitted corrections to my record. It is a similar approach to the Scopus Author Profile. You can identify which articles are not yours and then submit the info to Microsoft. Similarly, you can request author merges.

Microsoft have also opened their API to users. That's beyond my capabilities I think but very interesting.

So, this is potentially a very interesting service, but still really in the "beta" stage. But then, all author profile services have their issues.

No comments:

Post a Comment