Literature review is a big part of a scientific underatkening. Thanks to the Internet, most papers are available on line. This resulted (at least for me) in a big number of papers that I read and stored for feature references… Welcome to the age of data mining. As a researcher I have yet another problem on my hands: after I solved the problem of finding information, how do I keep track of it?

This subject was touched upon in other posts (bookmarking, mind maps, references). Given that my major focus of attention (at this moment) is my disertation, I want to be able to have a solution for referencing the information I have found. This basically has some sub-problems: creating references, searching through references, adding them to my documents (thesis, publications, reports), and creating list of references in my documents. I’ll skip discussion of why I was not satisfied with the above mentioned tools. Instead I’ll tell you what I see myself using: Zotero.

Long story short, this is a common scenario: I found a paper using Google Scholar (acm/ieee/etc). Two clicks and my reference is created (with all the author, title, abstract, etc. information added properly). Another click and I’m actually reading the paper. One last click and the pdf is stored with my reference (for feature review). If I found something interesting I can add notes. I can tag the reference (for feature searches). I also can link other references as related. The bottom line: creation of references is a smooth, easy and quick process.

There are plugins for OpenOffice.org Writer and Microsoft Word that allows one to use Zotero reference in the documents. The process of adding references to the text and creating refere lists is a two click deal (also smooth and quick). Installation of both plugins (for firefox and for Writer) was very simple.

For more information take a Zotero tour.

Alexander Shyrokov.

As I’ve said in a previous post, IEEE’s Member Digital Library (MDL) really needs technical help. I continue to have problems accessing this service from work (!) because UNH’s subscription “interferes” with the IEEE system’s ability to recognize my login. So one minute I can access papers, the next I’m told that my subscription doesn’t cover the given content. Very frustrating. After calling IEEE on the phone, and being told to try accessing the system from another place, like home (!), I tried to email tech support. I got the following response from a “Member Service Associate,” that is from a person who is clearly not someone who has technical knowledge of what’s going on:

“Thank you for your email. If your company has a subscription to IEEE Xplore and your using the companys computer site then you may receive some interference when you login to the Member Digital Library under your membership username & password. We do apologize for any inconvenience.”

Well, inconvenience indeed. You’d think IEEE would try to have a technical person fix this problem, and not just have some customer service person apologize for it.

Andrew Kun

As I’ve written here before, I really cannot believe how poorly IEEE Xplore works. I have what the IEEE calls the IEEE Member Digital Library. This is a $35/month (!) service that allows you to download 25 (!!!) papers a month. So, I just tried to download a paper with the following result:

Seems pretty straightforward, right? IEEE Transactions, etc. However, I get this message:

“The content you requested is not included in your subscription.”

Hmmm… So I check to make sure my Member Digital Library is up to date, and I see that it expires April 30, 2008 (today’s date), so that should be fine. I call the IEEE and they confirm my suspicion that, since I’m logging in from work, and my institution (UNH) has a subscription, the IEEE system confuses me with UNH. SinceUNH’s subscription doesn’t allow access to this paper, I can’t access it.

Now, I did tell the very nice person on the phone (Virginia) that this is UNBELIEVABLE. This is the IEEE. And this is a $420/year service. And it’s work related, so I need to be able to access it from WORK. And most infuriatingly, it’s not even a consistent problem. I’ve downloaded plenty of papers from work using Xplore, while other times, like today, I was thwarted.

My grave concern was noted, and I received a promise that if I email Virginia the title of the article, I will receive a copy of it via email. I will also get credit for the papers I was not able to download this month on my next month’s allotment of papers (kind of like the minutes roll-over on a cell phone plan). Thanks Virginia. I hope you do tell management that they need to make fixing this service a priority.

Andrew Kun

IEEE’s Xplore service is an important source of papers for electrical and computer engineers. However, its search feature is frustratingly poor.

Here is an example of how poorly the search works. In advanced search, I initiated a query for “speech analysis by homomorphic prediction” and I specified that I want the query to look at all data fields (this is the default). I wanted to find a paper by Kopec et al. that appeared in the IEEE Transactions in Acoustic, Speech, and Signal Processing. IEEE Xplore informed me that “no results were found.” I copied and pasted my search into Google, and I got this link as the first result. Note that when you click on the link you get to the IEEE Xplore page that lists the paper as “Speech analysis homomorphic prediction” without the “by” in the title. The paper title includes the “by” however, it was transcribed incorrectly by IEEE, but this would have been difficult for me to guess correctly. Xplore can’t deal with this difference. Google can. Clearly, we have the technology. So, IEEE, please fix the search feature in Xplore. In the meantime, I guess I’ll log in to Xplore but search in Google.

Andrew Kun