AeRO Fifth National Forum

Yesterday I attended the AeRO 5th National Forum held at the Australian Research Resources Centre. It’s the first time this event was held in Perth, it’s also the first time I heard about this event. It’s a great opportunity for me to see the bigger picture and be more aware of what the eResearch community is trying to achieve, and the challenges they face.

I enjoyed the presentations about the various projects on building sustainable research communities, particularly HuNI and CATAMI. The common theme is to improve interoperability, standardise the way different entities in a field store, describe, annotate, and provide access to datasets to researchers in the field. Making it easier for researchers to search, discover, analyse this data and turn it into new knowledge.

I’m also fascinated by the Murchison Widefield Array project. Radio astronomy is totally beyond me, but hearing about the scale of the project, the data being transmitted wow’ed me. The current infrastructure around the world cannot cope with this bandwidth hungry field, therefore it’s difficult for researchers to collaborate with experts in other countries. I thought, is it actually cheaper to transport people than data?

It is apparent that research data management is important to researchers, and that’s where librarians fit into this big picture. It seems the participants recognise the importance of librarians in their projects as well, that’s great. I have good understanding of RDM, can’t wait to get more hands-on in this area.

Probably not related to eResearch, when I hear about all these projects, I kept thinking what would be possible if all library data are open, linked, and accessible via API. I wonder what the community would do with it? What will happen if all library management systems speak to each other?

Finally, can I just say, nerds are cool!

Fitbit flex

It’s a well-known fact that sitting down all day is bad for you. My day job is exactly that, and when I concentrate on a task I can sit in front of my computer for a very long time. I drive to work, so the distance I walk each day is pretty much between the car park and the library building. I tend to eat at the lunch room because I want to get back to my desk as soon as I finished eating, I’m a very slow eater, it’s impossible for me to walk outside, find a spot, eat, and walk back within 30 minutes. Very bad…

Age is probably catching up, I often get sore neck, shoulders, and back. I participated in the Global Corporate Challenge once, but failed miserably. I joined the gym at work, I made the pilate class every week for a couple of months, then I dropped out because my instructor went on leave and I didn’t like her replacement, I haven’t been back since.

About a week ago I bought myself a Fitbit flex since everybody is raving about it, I’ve got to admit I was skeptical at first of this little device. Is it just a hype? Does the sleep tracking function actually work? Is it going to keep me motivated in the long run? Is it comfortable to wear 24/7? Anyway, I thought I’ve tried everything, I might as well try this one. Umm… it’s cheaper than my gym membership that I’m still paying.

I’m happy to report that although I couldn’t reach my goal of 10,000 steps most weekdays, it’s easier on the weekend with the grocery shopping, dog walking, and house stalking activities. I like my Fitbit because all I need to do is walk and it does the rest, every time I open the iPhone app it syncs with my Flex, and when I reach the 10k steps goal it vibrates, it’s good feeling. Last week I parked at the farthest bay in the car park, I took the stairs instead of lift, I went for a walk around the campus in the afternoon. That’s not enough to reach my daily goal but a lot better than not moving at all, so it does remind me to keep moving, and being a data geek, I like having data about myself to play with.

I’m thinking of other things I can do this week to meet my goal during the week. Perhaps I should park at the student car park instead of the staff car park, go for a walk around the campus in the morning as well as afternoon, I also want to go back to the gym, hopefully at least once a week after work. Will see how I go this week!

Damn you ampersand!

By Leo Reynold via Flickr http://www.flickr.com/photos/lwr/32256861/

I’ve been working on an Excel version of the Google scripts written by Stephen Flynn, which automates the searching of the Sherpa/Romeo database, you can view his blog post and video here. I’m very glad that I found his article, it inspired me to explore this automated workflow again**.

I want to create an Excel version because I want something that anyone in my library can use with a click of a button, and we won’t be limited by Google the number of calls we can make each day. Excel spreadsheets also integrate with our workflow better than XML and Google spreadsheets. Additionally, I want to get specific restrictions for preprint and postprint archiving, as well as all the other conditions for each title, so all required information is brought down to the spreadsheet, we don’t have to run the search again when a title requires double checking.

I’ve been agonising about that piece of VB code, while it works for most of the rows in my test spreadsheet, I can’t seem to get the information for a bunch of journal titles. Coincidentally they’re all of one publisher, I don’t think publisher would make any difference because the Sherpa API returns XML data. I thought there must be something wrong with my code but I cannot determine what’s wrong. Why does it work for every other title I tested but a bunch of titles by a particular publisher? I examined the XML results 2000 times, I rewrote my code in at least 4 different ways and the result is still the same. So I bit the bullet and asked a question on Stack Overflow, mentally prepared for the harsh critics, or even worse, no response at all.

Someone replied and said there’s an invalid character in the XML document. I was like, WTF? I didn’t notice that?! All this time I was thinking there’s something wrong with my code, but it’s actually something wrong with the Sherpa data….

So there we go, my codes are probably fine. Looks like it’s the & that appears in the copyright link text tag for all titles of one publisher, that’s why it returned a result but it won’t parse at all, I’ll report the problem to Sherpa. I’m learning something new every day!

Edit: I learned how to bypass the validation from the person who answered my question, and my codes are now working fine. woohoo! 

Ahh… the wonderful world of codes 🙂

** My manager suggested the use of Sherpa API a couple of years ago, but at the time the XML reference in Excel was not available to my machine, therefore it wasn’t possible at the time.

(Cod* OR Program*) AND librarian*. Any results?

I’ve been in coding mode for about a week tackling 2 post-it notes that have been put off due to other higher priority things my procrastination.

The first task was to establish an automated process for registering DOIs for a journal we host in the institutional repository. The cons of having a proprietary system is that your vendor will not give anything away when you’re trying to do something they offer as additional service, so pretty much I have to start from scratch. I’m happy that I knocked out this one without too much struggle. I created a DOI generator with Excel VBA. Then I wrote an XSLT stylesheet to transform the OAI XML to CrossRef XML. I can’t remember the last time I hand coded an entire document, it would have been when I worked as a web developer many moons ago, I was a PHP whiz. Time has certainly changed, XML, XSLT and Xpath are so much easier to understand.

The other task I’m currently working on is to export the data from citation databases like Scopus, clean up the data, semi-automate the copyright checking process, and batch load metadata and full text documents into the institutional repository. I followed the logic of the article, Using XSLT and Google Scripts to Streamline Populating an Institutional Repository. I imported the references to EndNote and exported the reference list as XML, the data is very clean, but it lacks affiliation information, which is crucial to us. So I started exploring the Excel export which gives me the “author with affiliation” field as one loooong string… I spent the afternoon trying to figure out how I can populate the spreadsheets properly and pick up all the correct affiliations and search the staff directory for author’s faculty/school/research centre information (which is usually missing from the database export) with minimal human intervention.

Anyway, I’m not going into further details here, most importantly I haven’t figured it out yet 😛 I thought I’d put it out there, I’m just wondering if there’s someone out there doing similar things. In the past few days, I feel a bit isolated, I can’t talk to my colleagues and bounce ideas because nobody here knows how to code. I can contact colleagues from other libraries but I have yet to come across another librarian who codes, I’m sure I’m not the only one trying to do this.

If you are or know of anyone who are coding for your library, perhaps we should connect, have a chat about what we’re working on, what programming languages or tools we know or want to learn. We can also meet regularly to share ideas, learn from each other, and see if there’s opportunity to collaborate?

Even if you don’t know how to code but want to acquire this skill, let’s connect too. I mean, if you can construct the above Boolean search, you already know how to code 😉