Check out VADLO.
When I speak with faculty in the humanities, one topic that comes up pretty often is word processor support for foreign languages. Some struggle to get diacritical marks correct. Others write in languages that are read from right to left, and when this isn't supported they have to spell out words backwards as they write. (As if writing isn't hard enough to start out with, right?)
My background is in biology, so I don't have a whole lot of experience with this stuff, but faculty have told me about a couple of pieces of software that they've found useful.
Nisus Writer
Nisus Writer Pro supports Latin, Cyrillic, Greek, Chinese, Japanese and Korean scripts and enables you to enter right-to-left text as in Arabic, Hebrew, and Persian. Sorry Windows users, but it's Mac only. And it's pronounced nice-us.
Keyman
Keyman Pro allows you to enter text in other languages by remapping the character keys on your keyboard for the language you want to write in. I gave it a quick test drive and got it working quickly and easily. It sets you up with a menu that pops up from the Windows task bar, and that makes switching between languages a snap. Windows only.
Here are a few blogs over at the library that you might want to keep your eye on:
Don't have time to check a bunch of blogs every day? Then install a feedreader!
I posted a little while back about the ToughBook, a laptop from Panasonic designed for extreme conditions. Dell now offers a laptop like this, too. It's called the Latitude XFR D630, and it meets the same military specifications as the ToughBook.
I spend a lot of time speaking with faculty about their IT needs, and one topic that comes up over and over is web design. Everybody, it seems, wants a better web site, and the reasons are obvious. The web is the first place we turn to for information nowadays. It's where your colleagues go to find out what you've been up to. It's where prospective grad students and postdocs go to look for somebody to work with. And it's where funding agencies go to learn what you've been doing with their money and to decide whether to give you any more.
For those at the School of Medicine, it's starting to get a lot easier to design slick, modern web sites. The folks there have developed a set of templates for web pages which will provide a consistent look for web sites across departments and organizations. These templates will also make it easier to design web pages. Easier means faster, which means less expensive. To date the the templates have saved about twenty hours of design time per site.
The next improvement in the works is the implementation of a web content management system, which will allow Yale personnel who haven’t been trained in web design to manage the content of their web sites. That means that when you want to use your web site to tell the world about your latest and greatest research, you won't have to go through a web design pro to do it. You'll just edit your web site as easily as you edit a Word document. The School of Medicine considered thirteen WCM products, and in January they selected SDL Tridion. The migration of sites built with the templates into SDL Tridion is underway, and support of end users is ramping up.
There's one final thing I haven’t been able to find out. The materials that I’ve read about the templates and SDL Tridion mention web sites for schools, departments, and organizations, but they don’t mention web sites for individual researchers, so I'm not sure if you'll be able to use them for your research group.
You'll find more details here and even more in this PDF, which includes a timetable for the implementation of SDL Tridion.
Perhaps this has happened to you once or twice. You're giving a talk, working your way through your beautiful PowerPoint, when somebody asks a question, and to answer the question you need to draw something. What do you do? Well if you're lucky the projector screen isn't covering the entire white board, and you have maybe two feet off to the side to draw something. And if you're even luckier, there's a marker for you to draw with. It's good enough for you to get the question answered, but there must to be a better way, right?
One solution is to use a tablet PC for your presentation. If you're using a tablet and you need to draw something, you just open a new document, whip out the stylus, and draw away on that wide open white background. Your creation and inspiration would fly right through the projector and onto the screen for all to see. Or you could even draw right on top of your PowerPoint presentation. That way you wouldn't be constrained by what you put in your PowerPoint ahead of time. You could add to the talk on the fly in response to your listeners' interests.
There are researchers at Yale who work in the lab, and there are researchers at Yale who work the library. And then there are researchers at Yale who work in the jungle. Or in the desert. Or on a mountain. These researchers have special technology needs. Their laptops, for instance, face dangers greater than a knocked over cup of coffee. One laptop built for abuse is Panasonic's Toughbook. It meets military standards for withstanding impacts, low temperatures, high temperatures, liquids, and dust.
And just how tough is the Toughbook? Well take a look at this:
As I've posted before, Yale has a new site license for Matlab. Students get Matlab for free, and everybody else has to pay $50.00. But, the Matlab that students get for free is a little less Matlab than the Matlab that everyone else gets when they pay. The difference is in the toolboxes.
These are the toolboxes included in the student version:
- Simulink
- Bioinformatics Toolbox
- Control System Toolbox
- Curve Fitting Toolbox
- Data Acquisition Toolbox
- Image Processing Toolbox
- Image Control Toolbox
- Optimization Toolbox
- Signal Processing Blockset
- Signal Processing Toolbox
- SimMechanics
- Simscape
- Stateflow
- Statistics Toolbox
And these are the toolboxes that everybody else gets:
- Simulink
- Bioinformatics Toolbox
- Control System Toolbox
- Curve Fitting Toolbox
- Data Acquisition Toolbox
- Signal Processing Blockset
- Image Processing Toolbox
- Instrument Control Toolbox
- Optimization Toolbox
- Signal Processing Toolbox
- SimMechanics
- Stateflow
- Statistics Toolbox
- Simscape
- Database Toolbox
- Datafeed Toolbox
- Distributed Computing Toolbox
- Excel Link
- Filter Design Toolbox
- Financial Toolbox
- Fixed-Point Toolbox
- Fuzzy Logic Toolbox
- GARCH Toolbox
- Gauges Blockset
- Genetic Algorithm and Direct Search Toolbox
- Image Acquisition Toolbox
- Link for Code Composer Studio
- Mapping Toolbox
- MATLAB Builder for Excel
- MATLAB Builder for Java
- MATLAB Compiler
- MATLAB Report Generator
- Neural Network Toolbox
- Partial Differential Equation Toolbox
- Real-Time Workshop
- RF Toolbox
- Robust Control Toolbox
- SimBiology
- Simulink Control Design
- Simulink Parameter Estimation
- Simulink Response Optimization
- Simulink Verification and Validation
- Spline Toolbox
- System Identification Toolbox
- Target for TI C6000
- Video and Image Processing Blockset
- Virtual Reality Toolbox
- Wavelet Toolbox
- xPC Target
When you have to get up to speed on a new topic, there are a few different ways to go about it. You can do a literature search, and that's a pretty good way because it's reasonably comprehensive, but it won't tell you anything about which papers have been influential and which have not. Another way is to let a recent paper serve as a jumping-off point by using its reference list as an entry way into the literature. You just go out and read every paper that the first paper cites. This is a pretty good way, too, because it stands to reason that the most frequently cited papers are the most influential.
A newer way to do this is through a web site called CiteULike, and here's how it works. You create an account in CiteULike, go to your favorite database, and find a paper you like. You save the paper to your CiteULike library by clicking on a special bookmark. Once you've done that, you can find other CiteULike users who have saved the same paper to their libraries, and then--this is the important part--you can see the other papers that those users have saved. This means that you can leverage the literature searches that have been carried out by people with similar interests. (If this sounds like del.icio.us for academics, it's because it pretty much is.) CiteULike helps you navigate that forest of literature you have to master by revealing the footsteps of those who have gone before.
Yesterday I learned about a Yale IT tool that I thought I'd pass it along even though it's not specifically for research, because somehow I've managed to work at Yale for almost seven years before discovering it. It's the Yale University Portal, and it's an easy route to the Yale online world. It leads to Yale's calendar and announcements; it shows the headlines from the Yale Daily News, Yale Global, and the New York Times; and it lets you search for people and find Yale buildings on a map. If you login it shows your status on the central backup and gives you a digest of your email. That's not too bad for one-stop shopping.
Speaking of maps, have you checked out the main online Yale map? I think it's pretty cool the way it slides and zooms. I use it a lot to figure out where I'm going before I head out to talk to faculty about how they use technology in their research. I've had a lot of luck with the map's search facility. Whether I enter a building name, a 3-letter building code, or a street address, it always seems to bring back what I'm looking for.
And one last thing non-research thing. If you ride the Yale shuttle, the real-time shuttle tracker is indispensable.
I'd guess you use EndNote to format reference lists, right? I think most researchers do. What most researchers probably don't know is that there's a web version of this popular software. It's a separate product from the version you use on your desktop, and Yale has a site license for it, so it's absolutely free. You can read all about it here.
I took it for a short test drive, and it doesn't seem to have all the features you might be used to in the desktop version, but it does have the advantage of being, well, on the web. That means that you can access your reference library whenever you can access the web. It also means you can easily share your reference library. All you have to do is provide the email addresses of the people you'd like to share with. This could be a good way to collaborate on writing a document. You'd set up a group of shared references, and when people edit the document, they would add the necessary references to the group. You can even leverage all the work you've done building your reference library in your desktop EndNote, because you can upload references to the web version. I just tried that out, and it was easy. (You can also download from EndNote web to EndNote desktop.)
I can't recommend EndNote web without qualification because I tested it only briefly, but it's a new tool, and it's web-based, and it's free to anybody at Yale, so I thought I'd let you know.
Last week I posted about The Vault, a centralized file service at Yale. There's another place at Yale where you can put your files, and this one is completely free. It's called Pantheon, and nearly everyone at Yale already has an account. (The exceptions are anybody who started at the School of Medicine a long time ago and anybody at the School of Management.) You get 500 MB of space, and you can mount it on your Windows or Mac desktop and access it like a local drive. The file server is kept in the data center, so it's secure, and everything is backed up, allowing you to recover deleted files and old versions of file. You can read all the details here.
Yale limits the size of email attachments to 20 MB, and sometimes that might not let you do what you need to do. In those cases you need the file transfer facility, an easy-peasey way to transfer files up to 200 MB. Through a web interface you upload a file and provide the email address of the person you want to receive the file. The facility then emails a link to that person, and to download the file all that person has to do is click on the link. You can send files to people outside Yale, people outside Yale can send files to you, and it's more secure than email.
When I meet with faculty to discuss their technology needs, the first thing most of them mention is file storage. Sometimes they need somewhere to put their files, sometimes they need a place to back them up, and sometimes they need a way to share them. Most don't seem to know that Yale has a central location to put files for sharing. It's called The Vault (formerly Med1.) It's secure, backed up, and accessible off campus via VPN. It appears as a drive on your Windows or Mac desktop, and it's possible to have multiple volumes attached to different billing sources with different sets of permissions. The price is $10.00 per month for the first 150 MB and $3 per month for each additional 150 MB. (That's subject to change. Look here for current pricing.)
Friday's post was about this week's Matlab workshop, and somebody emailed me to ask how much you need to know to get something out of the workshop. Do you need to be an experienced Matlab user? Do you need to have programmed in Matlab? Do you need a background in parallel computing? No, no, and no. For the morning session you don't have to know anything about Matlab at all, and for the afternoon sesssion it'll help to know about Matlab, but you don't need to be a programmer or a parallel computing ninja.
Yale recently purchased a new site license for Matlab, and the folks from The MathWorks are coming to campus to put on a workshop and answer your questions. The workshop is free and open to everyone. You can register in advance if you like.
Friday, April 25, 2008
140 Prospect Street
Room 101 (StatLab)
Free
8:30 – 9:00 Registration
9:00 – 10:30 Taking Control of your Code
Steve Eddins, Ph.D., Software Developer and Manager, The MathWorks
Coauthor of Digital Image Processing Using MATLAB
Engineers and scientists often have to write a lot of software to get
their work done. Many are too familiar with problems such as these:
- Last year’s results can no longer be reproduced because the code changed.
- Collaborators use different versions of the code, resulting in persistent confusion.
- Critical projects rely on “brittle” software that is difficult to modify without breaking.
Fortunately there are basic tools and methods that provide tremendous improvement in software development efficiency and quality. These methods can be learned and applied successfully without formal software engineering training. The audience will learn to use these techniques to control software changes, maintain and improve software, and deal with bad code.
10:30 – 10:45 Break
10:45 – 11:15 Intro to Simulink
Dave Forstot, Application Engineer, The MathWorks
11:15 – 11:45 Q&A
11:45 - 12: 30 Break
12:30 – 1:00 Registration
1:00 – 2:30 Solving Large Technical Problems Using MATLAB
Dave Forstot, Application Engineer, The MathWorks
- Dealing with large data sets
- Implicit multi-threaded computations
- Parallel programming
- Interactive task parallel applications
- Interactive data parallel applications
- Interactive applications to batch applications
- Tips/Tricks on parallel coding in MATLAB
2:30 – 3:00 Q&A
Pubmed is amazing, of course, and so is Google, but have you thought about the next level of internet search? It's probably going to be built on Resource Description Framework. RDF aspires to allow any nugget of knowledge to be expressed in a machine-readable format. It uses a syntax of triplets in the form of subject-predicate-object, something like this: "John Fitzpatrick" (subject) "works at" (predicate) "Yale" (object). Or something like this: ""The Sun Also Rises" (subject) "was written by" (predicate) "Ernest Hemingway" (object).
You might think of this as a way of expressing a piece of information that lives in a table. The rows of the table are all subjects, the columns are predicates, and the entries are objects. It's easy to see that you could express all the information in the table with subject-predicate-object triplets.
Ok, so what? Well, the key is that a piece of knowledge expressed in this way can be put in a machine-readable format with XML. Once the knowledge is in a machine-readable format, it's pretty easy for machines to aggregate the knowledge. That means that when you search the web, you search not for keywords but for relationships. You search for facts instead of terms.
I think this could have a big impact on research. Think of the summary slide at the end of a research talk. It usually has four or five bullets outlining the principal points of the talk. The bullet points are probably phrased as triplets already, or close to it. Imagine if all those bullet points were out there on the web and you had a search engine that could pull in the ones you need to know. Your searches would be more efficient, you'd find out what you need to know faster, and you'd be a better researcher. Stay tuned.
Just in case you missed it, back in February the Faculty of Arts and Sciences at Harvard voted to allow access to their scholarly works through an online repository. (Here's the resolution, and here's the press release.) This means that everything the Harvard faculty produce will be available free to everyone everywhere. Almost. The resolution contains an opt-out clause that allows faculty to request that the policy be waived on a particular article basis. In the past such loopholes have been crippling; the NIH Pubilc Access used to be voluntary, and participation was only 4% then. We won't find out the effect of the loophole for three years, when the policy is due for review.
The National Institutes of Health have changed their policy regarding public access. As of last week, all peer-reviewed publicatons resulting from work funded fully or in part by the NIH must be submitted to PubMed Central within a year so that the publication is available free to everyone. This is a change to the previous policy, implemented in 2005, under which submission to PubMed Central was voluntary. The law was changed after a 2006 progress report found that voluntary compliance was only 4%. The NIH web site has a helpful FAQ on the policy and a list of journals which submit your paper to PubMed Central for you so that you don't have to do any extra work to be in compliance. If you're interested in the evolution of this policy, you'll find a helpful timeline here.
