skip navigation
search


innovation-2 (1)

A small, innovative group like ours is continually forced to make difficult decisions about what projects to take on next.  We just haven’t got the time to try all the things we can think up.  As a result, we’ve had to develop our own set of notions about what’s cool — what adds value rather than glitz, what helps our users the most, what will pass twin tests of novelty and utility.  Over the years we’ve developed an aesthetic. We think that the best innovations are both fundamental and (mostly) invisible to the user.  They’re just the way things should be. Make no mistake — we like fancy, flashy tech stuff just as much as anybody– but we’re very aware that a lot of it is just geek-fashion that, however pretty it may be, doesn’t offer much to our audience.

Indentation is an interesting case in point.  Over the years, a lot of you have told us that indentation is one of the best features of our CFR edition. It makes our edition much more readable than the unindented versions in print and on the Web.  Now, indentation is hardly innovative — it’s  been around since at least 1482, when it appeared in an incunabulum of Heinrich Knoblochtzer.  It’s not like we were the first people to think of it.

However… even though indentation itself is not particularly innovative, designing computer programs that automate the process over collections the size of the CFR is really, really hard (and, to be honest, our work on it is far from perfect).  The indentations don’t appear in the text files that we get from the Government Printing Office.  The only clues we get are the placement of enumerators (text addresses like the (a), (1), (iii), and (A) in 8 CFR 103.3 (a)(1)(iii)(A)). In theory, they follow a pattern (small letter, digit, roman numeral, capital letter) as the levels become more granular, but in reality the pattern can be different in different places in the CFR, and even within a particular section.  And, as you can see in the section linked above, it repeats when there are more than 4 levels.  To make things even more interesting, the Federal agencies who write the regs in CFR occasionally make mistakes — it is not unheard-of for a small roman-numeral “(i)” to be followed by a “(j)” because someone wasn’t paying close attention when editing.  The software that sorts all of this out — and then tries to audit the results for correctness — is very difficult to design; we’re currently on our third attempt, and as you can see in the linked example, we have a ways to go.  Writing computer programs that attempt this task is a pretty good way to make your head hurt (DAMHIKT).

But here’s the thing about our donors: you’re the kind of people who think indentation is cool, too.  Recently, a bunch of you said so in your responses to one of our surveys — and we were delighted to hear it. Interestingly, many of the same people said they placed a fairly low priority on  “development of innovative features.”  We loved that, because we think innovation should be so ready-to-hand as to be invisible.  And the fact that you guys find it so is very cool.

 

Mar 122015

IRS logo
No, we’re not talking about a tax-deductible donation to the LII (this time–but, feel free to go ahead and donate if you’d like).

We’re here instead to introduce a new feature you might not have noticed in 26 U.S.C., aka the tax code.

This is going to be a little bit interactive at first, so please play along.

Please click here to look at a section of the tax code that’s near and dear to us, Section 501: “Exemption from tax on corporations, certain trusts, etc.”  Note the light blue tabs at the top of the page:

  • US Code
  • Notes
  • IRS Rulings
  • Authorities (CFR)

What’s new is the IRS Rulings tab.  Please click on it.

Voila!

So, where’s the magic?  To appreciate the practicality of this feature, you should follow our link from that page to the IRS’s own collection of these letters (or by just clicking here).  What we’ve done is organize this large collection of guidance from the IRS in a way that is meant to be useful to tax lawyers, tax preparers, and others who are interested in it.  We’ve gathered together all the Written Determinations that cite to a given section of the tax code and put them with that section of the tax code.  There is no other publisher, free or commercial, who has done this.  (The IRS’s collection, by contrast, is a list sortable only by the number, the release date, the rather unhelpfully generic “subject” or the ponderous “Uniform Issue List Code.”

This is a project we undertook at the separate suggestion of two different friends of the LII–including a donor like you and Cornell University’s own, in-house tax guru.  In fact, one reason to spotlight this feature is for its value as a case study in where our projects come from.  They come from YOU.   If you see a gap in the way government or other websites or legal publishers are providing information–especially one that can be solved with the clever application of computer science–we are always happy to hear from you.  Email our Director here.

Another reason to talk about this feature is it demonstrates the overlap between data, computer science, and legal informatics.  That is the world in which operate, and we like to explain it to our friends like you whenever and however we can.  Good examples like this are a great opportunity.

Clean data from the IRS made this, frankly, relatively easy to do.  By publishing these letter rulings in xml with consistent metadata tags and uniform citations to the US Code, the IRS made it easy for us (or anyone else) to do what we did.  Compare that to the output of some other federal agencies–such as these pdfs of decisions of the Administrative Appeals Office of U.S. Citizenship and Immigration Services.  (We’re working on those next.)

From issues you might find familiar (such as “flat” pdfs with handwritten marginalia and other challenges for optical character recognition software) to others you likely will not (like the challenges of temporal instability at the subsection level of the CFR), most government-published collections of potentially useful data look more like the USCIS’s than the IRS’s.  In short, this is fertile ground for both helping the government to improve the way it makes its work product available and for applying advanced methods of processing and analysis to improve the usability of the existing piles (and piles and piles) of government-created data out there.  We do both.

Just like we’ve written about the indentation in our CFR, one way to measure the quality of a feature is by measuring its “invisibility”–can users like you find it and use it without ever appreciating just how much effort might be going on behind the scenes to bring you this “simple” little bit of functionality?  While we strive to bring you innovation that looks and feels light-weight, it’s important to our mission and the future of online legal publishing that every once in a while we stop and say “Voila!”

That’s a small glimpse into the world in which we operate.  Meanwhile, in the world in which we ALL operate, Tax Day is just around the corner.

 

 

photo2She’s baaaack!  Former LII employee Sylvia Kwakye recently returned to Ithaca, and we are lucky to add her to our staff as a Text Systems Developer. “I practice the dark arts of data mining and natural language processing to transform dry legal texts into attractive, easier to relate to but accurate reflections of themselves,” she explains.

Sylvia worked for the LII years ago while studying for her PhD in Biological and Environmental Engineering at Cornell.

“I took Computer Science 501 as part of my Computer Science minor,” she says. “In this class we were asked to do a project, and given a list of Cornell groups that had problems to solve. I was always interested in doing something real and useful, so I chose the LII. I was interested enough to ask to continue working when the class was over.”

As LII Director Tom Bruce recalls it, “Sylvia thought it would be fun to work on the U.S. Code with us.” Since the U.S. Code project was complex and difficult to create, it’s clear that Sylvia finds her “fun” in taking on complicated problems.

Once again, she’s working with the U.S. Code–now as a full-time member of the LII staff. One of Sylvia’s noteworthy projects is adding definitions of key terms. For example, there are 47 different uses of the word “water” in the U.S. Code. Sylvia is developing a system that will display the correct definition each time the word appears.

Sylvia comes from Ghana, but spent most of her first eight years in the U.S.  Her father studied at the University of Chicago. When the family returned to Ghana, he worked for several agencies and the University of Ghana. “We moved around, but mostly lived in Accra, the capital. We spent school vacations in the villages with our grandparents.”

Stories of the University of Chicago led Sylvia to apply there. “Initially, I wanted to be a doctor, and I got into medical school in Ghana,” she recalls. “But Dad loved liberal arts education, and said, go to a liberal arts school and then do medical school afterwards if that’s what you still want.” Stories of the University of Chicago led Sylvia to apply there.  She also applied to Swarthmore, which offered her a scholarship.  Ultimately, she chose Swarthmore.

Engineering seemed like a good pre-med degree. However, computational biology was becoming better known, “and I just happened to be in a position to understand both biology and engineering,” says Sylvia.

After graduating from Swarthmore in 1998 with a BS in Engineering, she worked as a research engineer with the Computational Biology Group at the Dupont Experimental Station in Wilmington, Delaware. There, she says, “I got interested in writing and using software to make life easier.”

“We were trying to understand what protein switches turn genes on and off as a plant grew,” she says. “A microarray is a tray with 96 wells , so you could do 96 DNA tests at a time. So, you quickly end up with thousands of data points to analyze. That’s where all the computer science courses I had taken for fun at Swarthmore College came to the rescue. I wrote software tools to reduce months of analysis work to a matter of hours. From then it’s been routine for me to learn programming to help me get work done.”

For her PhD work at Cornell, she developed a system for rapidly detecting pathogens for use in low-resource communities: that is, a small, portable box that would test for the dengue fever virus, among other things.

After receiving her PhD in 2010, Sylvia teamed up with a graduate of Cornell’s Johnson Graduate School of Management and a professor at Case Western Reserve University to develop her detection system into a commercially viable product. Unfortunately, they ran out of resources. “Because we were such a small team, major life changing events had quite an impact on our abilities to carry on,” Sylvia says. Those events included the death of her father, which required her to spend time in Ghana settling the estate.

Meanwhile, her husband Stefan continued studying for his PhD at Cornell, while his mother came from Trinidad to help with their two children, now aged nine and eight. Stefan has since completed his degree, works in the renewable energy industry, has a consulting job, and also writes software for fun.

When Sylvia returned to Ithaca, the LII was delighted to bring her back onboard.  In addition to her primary responsibilities engineering improvements to the website, she’s also helping Sara Frug, LII’s Associate Director for Technology, to mentor Master of Engineering students who are working on projects for the LII. “I was just like them,” Sylvia recalls. “I had the same anxiety about getting it right!”

Talking with Sylvia, however, you don’t see anxiety about getting things right; only the enjoyment she takes in solving complicated problems.

STEPHEN_YALE-LOEHRWhile gifts of money are crucial to our operating model, the vote of confidence they represent is just as important to us.   Donations to the LII come in many forms, all of them helping us realize our mission in different ways.

As we note in our story about IRS Written Determinations, some people give us great ideas to explore.  We are always interested in hearing about problems like, for example, unavailable or poorly-organized government data we can help organize into a useful tool for researchers, academics, businesses, attorneys, or just “the general public.”  Send us your ideas here.

But right now, we’d like to focus on a  third kind of donor: our donors of content.  Whether it is a retired computer scientist writing about the Digital Millennium Copyright Act or a practicing attorney updating our overview of bankruptcy law, folks from many different walks of life contribute content, primarily through our Wex legal reference feature.  (You can join them by signing up here.)

Recently, law professor and long-time friend of the LII Stephen Yale-Loehr upped the ante by getting his students involved.  He made writing for Wex part of the coursework in his immigration law seminar.   We caught up with Steve to talk about why he’s donating effort–both his own and his students’–to the LII.

You are both a law professor and a practitioner of immigration and asylum law. How do those two roles complement each other?

I think the two help each other. Practicing law allows me to bring real world stories to my immigration law classes, which my students always find interesting.  And having to keep up on the latest immigration law developments for my classes helps my immigration clients.

You’ve been on the Law School faculty since 1991.  Has the advent of the Internet had a bigger impact on teaching law or practicing it?  Why?

The Internet has had a big impact on both.  For example, most of my assignments for my immigration classes now are links to articles or cases on the Internet, rather than a hard copy.  I also use YouTube video clips to illustrate certain points.  That was not possible before the Internet.

The Internet has dramatically changed immigration practice.  I do almost all of my research online now rather than through books.  With the advent of the Internet we can represent clients anywhere in the world.  For example, over half of our immigration clients at Miller Mayer live overseas.  We never may never meet them, but we communicate effectively with them via email and receive and send them documents electronically.

This year, you asked your students to publish summaries of important Supreme Court immigration decisions in Wex.  Did you have a particular pedagogical goal in mind, or did you just want to see the content added to Wex?

Both.  I wanted to see if my students could summarize Supreme Court immigration opinions, which can be pretty dense and complicated, in a way that would be accessible to lay people.  Also, some of the Supreme Court’s key immigration decisions were decided 50 or 100 years ago, but remain important today.  While Wex has started summarizing major Supreme Court decisions, a lot of older opinions have not been summarized yet.  People should know about key immigration decisions, no matter when they were decided.

Do you think your students approached that particular assignment differently because their work was going to be (and is now) viewable by the general public?

Like any teacher, he called on one of his students to answer this question.  Jessica Flores, a member of the Class of 2015, replied:

“I approached the writing to the immigration case summary differently to some extent.  Since I knew my audience was going to be the general public, as opposed to other law students or lawyers, I wanted to make sure I carefully explained the case and the legal concepts of the case in a simplified way.  I wanted to make sure that anyone without any legal or immigration knowledge/background would be able to understand the case.  I know that legal cases can sometimes be difficult even for law students to understand so I tried to explain the case as clearly as possible.  I liked that I could hyperlink legal concepts in my case summary to other LII posts because I knew that other summaries would assist the general public in understanding my case summary.  For example, I hyperlinked the Fourth Amendment in my case summary.”

How much editing did you do of the student pieces before LII published them in Wex?

Very little.  I was pleasantly surprised at how well the students did.  This is in part because of the excellent template and instructions that LII developed.

This seems like a fairly simple model that could be replicated in other classrooms–and not just law classrooms–around the country.  Do you see any potential pitfalls to LII expanding it into other seminars writing on other areas?

None at all.  LII has developed a good template and instructions to make it easy for any professor to give this assignment to his or her students.  It is also a good way to see if students really understand key cases!

To see the finished summaries from Steve’s students, click here.

 

thank-you-newsletter (1)Friends:

Because of you, our year-end fundraising campaign was quite successful, raising well over $70,000.  That’s a 6% increase from the previous year — an impressive number when you consider that the average for nonprofits nationally is only 4%. Thanks to all of you, both for your generous contributions and for the vote of confidence that they represent.  As part of the community that recognizes that open access to law helps all sorts of people solve important problems, you’re helping 28 million people around the world find and understand the law — and to use that understanding to solve problems, either for themselves or for others.  Thank you.  We’re immensely grateful for your belief in the LII’s mission, and for your investment in our work.

Your contributions buy time and talent — the two things that we never have enough of.  One of the advantages of being affiliated with a large university is that the students are unbelievably smart and skilled; your contributions allow us to employ people who will, in a year or two, be working for the likes of Google, Facebook, Oracle, and for an array of high-end law firms.  They turn out amazing work in developing new features, in writing new material for the site, and — most importantly — in undertaking collaborative work that needs both legal and technical expertise.  Because they’re students, your dollars go a very long way (and, incidentally, you’re helping them buy groceries).

This year, we’re planning significant improvements to our US Code collection, primarily in the form of linkage to interpretive information like the IRS written determinations we added to the tax code last year.  Right now, we’re working on the interpretive letters that the USCIS issues in response to questions about immigration law, on finishing our grand project of linking all of the words in the CFR and US Code that are defined by statute to their respective statutory definitions, and on making the interpretive material in our 5,000-article WEX legal encyclopedia directly accessible from the laws they explain.  We expect that WEX itself is going to expand significantly during the next year — we are actively recruiting volunteers to help us create more expert commentary and explanation, and many of you are already helping out.  Contributions of effort are really important, too — and I’m hoping that many of you in our community will join us in making WEX the best place for people who want to understand the law.

Once again, thanks for helping us out.  LII donors are a unique bunch, and all of us are delighted to have your support.  As always, I’m eager to hear from you — so don’t hesitate to drop me a line with your comments, suggestions, and criticisms.

 

 

1458.jpgWe recently coordinated a very successful speaking engagement at Cornell Law School by long-time LII supporter Ed Walters.  Ed is the CEO of Fastcase, which describes itself as “the leading next-generation legal research service that puts a comprehensive national law library and smarter and more powerful searching, sorting, and visualization tools at your fingertips.”  Fastcase, among other things, pioneered citation analysis, data visualization, mobile apps, and eBooks for legal research.

But Ed wasn’t at Cornell Law to talk about Fastcase.  In his “free” time, Ed is a pioneer in the study of what he calls “robot law.”  He’s teaching the subject this semester as an adjunct professor at the Georgetown University Law Center.

The LII invited him to come speak at Cornell because, like the LII itself, this topic implicates expertise in computer science, information science, and law.  In fact, LII Director Tom Bruce was sure to invite his colleagues from IS/CS and their students, and they showed up in large numbers.

“Robotics embodies the physical transformation of the Industrial Revolution with the cultural upheaval of the Internet Revolution, and it has the potential to be bigger and faster than either,” Ed told the gathered audience.  “The challenge of the next 20 years will be to make sure our law and society are ready for self-driving cars, surgical robots, pervasive surveillance, and drone warfare.”

The lecture started with this same historical perspective–placing the increased use of robots in a continuum beginning with the Industrial Revolution and continuing through the advent of the internet into today.  Ed challenged the audience–as he challenges his students at Georgetown–not only to think of the vast applications for autonomous, sentient or quasi-sentient machines in all walks of life, but also to consider the legal and ethical ramifications of how society incorporates, regulates, and even punishes these machines.

Ed’s message was largely upbeat.  He believes that, like first machines and then computers, robots will increase human productivity without utterly replacing the human labor force.  But he was quick to emphasize the potential consequences if courts, legislatures, and society as a whole fail to make law that adequately addresses and accommodates the difference between human intelligence and the “emerging” intelligence we see in computers like IBM’s famous jeopardy-playing computer Watson.   A key component of his message was that these decisions cannot be delayed until some distant future:  “This isn’t science fiction; it’s science present.”

Feedback from around campus was entirely positive.  Eduardo Penalver, the Alan R. Tessler Dean of the Cornell Law School told us afterward that Ed “is thinking at — and working at — the cutting edge of law and technology.  We were fortunate to have him bring his insights to Cornell Law School.”

We were pleased to bring Ed to Cornell, and equally pleased to bring an audience of engineers, computer scientists, and lawyers alike to hear his thought-provoking talk.

http://merkosoncampus.com/wp-content/uploads/2011/03/ripple.jpgLast year, you made it possible for us to help nearly 27 million people from over 200 countries find and understand the law.  But that is just the tip of the iceberg.  Many members of that enormous-but-anonymous audience — including a lot of you — use our materials to help people who have never looked at the LII website and probably never will.  And it is out there, on the ground with the helpers that we help, that your contributions are changing lives.

Elsewhere in this newsletter, you can read about Deb Fisher, who uses the LII in her work with Tax-Aide, a program of the AARP Foundation.  Tax-Aide helps 2.6 million low- and middle-income Americans, most of them elderly, file their income taxes every year. Deb is one of ten people who develop training materials for Tax-Aide volunteers.

There are many more stories like Deb’s. A couple of weeks ago, I spent a day “in the numbers”, looking at what we know about how our site is used.  Here are a few of the people and organizations that are helped when you donate to the LII:

That’s only a part of it. And even so, it’s not the 3-bullet-point, telegraphic list that a clever professional would use in a newsletter like this.  The scope and diversity are important and impressive.  When you give to the LII, the impact is transmitted through those who use the site to help others.  Your contributions, directly and indirectly, help millions.

Those big numbers — the 27 million in our audience, or the 2.6 million who are helped by Tax-Aide volunteers like Deb — are impressive as hell,  but they lose detail.  They’re like an aerial photograph of a crowd.  We can see that there are a lot of people, and that they tend to cluster around particular needs and issues.  Every once in a while, with the help of some of you who are closer to the ground, we can see an individual face or two.  In Deb Fisher’s experience, it’s a face that is glowing with gratitude.

It’s a remarkable act of faith on your part.  You put contributions in our hands, trusting that we’ll build something people can use to help themselves or help others in ways that neither you nor we will ever know in detail.  We try to return that trust by using your money wisely.  We describe that process in more detail elsewhere, and I would invite you to look into the details and mail me personally with any questions or comments you might have.

There are other, more dramatic stories to tell and perhaps I’ll do that in some future newsletter. But I personally prefer the story of a few people like you making a hundred little things possible that have meaning for real people — millions of them, all over the world.

Thank you.

Displaying Deb & Warren.jpgDeb Fisher spends a lot of time in the LII’s U.S. Code Title 26 – also known as the Internal Revenue Code. That’s because she’s a retired civil engineer who volunteers for the AARP’s Tax-Aide program.  Tax-Aide provides tax preparation for low-to-middle income people.  Most, but not all, are seniors. It serves 2.6 million taxpayers annually at more than 5,000 sites nationwide. Nationally, the organization has 35,000 volunteers, and only 12 paid staff.   It’s a great example of how the free legal information at the LII helps those who help others.

Deb and her husband Warren work for Tax-Aide because they like to solve puzzles. “I saw an article in the newspaper about the training,” Deb says. “I had just retired and was missing the numbers part of my brain. I got absolutely hooked.” She would come home and tell Warren about the intriguing puzzles she had solved, so he volunteered too.

“In some volunteer jobs, you understand how important they are, but go a long time before seeing results,” explains Deb. “With Tax-Aide, at the end of the hour you have someone just glowing with thanks—it’s almost a drug, being so appreciated.” The anti-drug happens “when you have people who unexpectedly end up owing money and have to make a monthly payment plan. That’s emotionally hard.”

Twenty years after helping with her first tax return as one of Tax-Aide’s thousands of volunteers, Deb is one of 10 who serve on the National Tax Training Committee (NTTC), working to develop the materials used for training the Tax-Aide volunteers.

The NTTC volunteers collaborate to develop Tax-Aide training in a specific area, as well as assisting the IRS to develop the official training materials. Deb explains: “The ten of us produce the AARP supplements. We also produce special programs for the instructors in case the class needs a particular problem to supplement their training.”

Her husband Warren is a district coordinator, running what is called a Super Site. “There are five tax sites in our county,” Deb explains. “The fifth one, the Super Site, is at a local mall. It’s open six days a week and two evenings, with 12 to 15 volunteer preparers. They do 4,000 tax returns each year. Warren makes sure the sites run well, and that the instructors cover everything the volunteers need to know. He also takes one day a week to do tax returns.”

Deb and Warren trained as civil engineers, then worked for the forest service. “Campgrounds need power, water, roads, bathrooms, and offices,” says Deb. “I mostly worked on roads and trails. We did trail bridges, a lot of road and trail design. Warren was a ski-lift engineer—so he had to go skiing.”

The couple lived in Anchorage, Arizona, New Hampshire, Kentucky, Atlanta, San Francisco, and Quincy, CA. “When we were married, we made a list of places we wanted to live,” Deb explains. “We had wonderful careers, with good people who were all working toward the same thing.” They chose to retire early—and have had 20 years of fun!

The Fishers now live in Coeur d’Alene, Idaho, “We had bought land there when we were first married, and we would come and camp on it every Labor Day,” recalls Deb. “It was 80 degrees, sunny, with wonderful lakes. We decided that had to be heaven. But when we moved up here we discovered that winter is cold and gray. The Rockies lie to the east of us, and the clouds pile up here.” On the other hand, she says, the Tax-Aide work is perfect for gray days.

Now, however, Deb helps with training during the tax season, and then spends the rest of the year working with the NTTC to create the training. Working at home, she spends three or four days a week on Tax-Aide. “We certainly take off on vacation now and then,” she notes. In their spare time, Warren likes golfing; Deb likes to line dance—and now she’s a dance instructor. She also enjoys gardening, and, like Ithaca gardeners, has a running battle with nibbling deer. “I’d like to make a deal with them,” she says. “Eat my plants in the daytime when I can see you!” Perhaps the Fishers can solve that puzzle, too.

Deb concludes,“ When someone has a volunteer job they really love and the spouse doesn’t, it can create some conflict, but when we’re both involved, it’s great.”

Nov 042014

Brian Laurence HughesLII supporters are very generous people, and we like to make sure that they get credit for their generosity. Of course, we use computers for that—in fact, a complex of computers that run several different, highly secure software systems. Meet the man who glues them together.

In 1999, Brian Hughes was the first full-time programmer to join the founding team of Peter Martin and Tom Bruce. Around 2002, Brian wrote the first, very simple donor-tracking program. These days, with added concerns about security and identity theft, those systems are handled by systems built (and operated by) trustworthy vendors such as Verisign, Elavon and Salesforce—but the LII still needs to integrate them, and that’s where Brian comes in. Although he works on other projects for the LII, such as transforming our up-to-the-minute data feed from the Supreme Court into the pages you read on the site, Brian finds working with donor information much more satisfying. “It’s been something that I’m completely responsible for,” he says.

Brian’s father was a programmer for Sylvania, DEC, and MIT, before working for the Peace Corps and USAID, bringing the Hughes family along, first to France, then to Morocco, finally to Niger. In those countries, young Brian went to French-speaking schools, and learned Moroccan Arabic. When the Hughes family returned to Massachusetts, Brian spent a year in public school before going to Harvard, where he studied linguistics. His parents and two of his three sisters also attended Harvard, while the third sister became a lawyer.

Brian didn’t start out as a programmer. “Library work appealed to me and it’s where I worked for a long time,” he says. “I was always a library assistant, not a professional—I didn’t want the aggravation of being a professional.” (His wife, Cathy Conroy, is a library professional, so Brian hears all about the aggravation that job can bring.)  Then, as now, Brian was strongly oriented toward customer service and to helping people find and understand information.

After college, Brian’s first job was at the Northeastern Law School Library. “I worked at the circulation desk,” he says. “I helped the students find books the professor wanted them to Shepardize” (For the non-lawyers among you, to “Shepardize” is to use the Shepard’s Citation Service–now a part of Lexis Nexis– to check legal citations to ensure the law cited remains current and accurate.

Brian then returned to Harvard, first in their geology library, then as an international law library assistant in the law library. Finally, he joined a computer support team. “I liked that job,” he recalls. “E-mail started near the end of my time there. I got a computer and fooled around with programming. I even wrote a program. You know what happens when you lean your finger on a key, right? You get 7,000 periods ……………………….. on your screen.”  Brian wrote a program to set the repeat at a reasonable rate.

With his growing computer skills, he maintained a database for personnel records. “I enjoyed that job,” Brian says. Since the background of much of LII website consists of complicated databases, one might see that as a sign of Brian’s future career.

Around that time, LII co-founder and current director Tom Bruce was consulting for Harvard Law Library director Terry Martin. They created LEDA, an institutional repository for archiving and distributing legal scholarship—and hired Brian to write the code for it. “Then Tom said—come to the LII,” Brian recalls.

Now, of course, people graduate from college with a degree in computer science. Brian, however, is self-taught. “I’ve always been learning as I go,” he says. For the computer geeks among you, Brian mostly uses Perl, PHP, Python, and MySQL. “They’re good for working on big wads of text,” Brian explains—for example, the U.S.  Code or the Supreme Court decisions.

Brian Hughes speaks or reads several languages, including French and Arabic, and he’s teaching himself Latin. Is this why he’s so good at learning programming languages?

“They call it a programming language, and programming books invoke human language. I think they’re nuts,” Brian says. “Computer languages are just a set of instructions.” In human language, he explains, “I don’t have to say things exactly right, and we still understand each other. We mentally correct as we read, but computers don’t do that.” Computers are literal, Brian explains. Leave out a comma, and the program breaks. Then you have to find that spot and add the comma. As with proofreading a written document (like this one), Brian notes, “it takes immense concentration NOT to see what should be there.” Brian’s work includes hunting for misplaced commas or equals signs, along with reading complex documentation for the various products that he glues together to create the systems that process your gifts.

Brian works from his home in Andover, Massachusetts—the first LII staff member to work remotely, and now the only one. The Harold Parker State Forest is in his back yard, which is perfect for Brian. “I’ve always liked the outdoor stuff,” he says. “When I came back from Niger and was in high school, it was a cheerful time—I was with my age group and speaking American. I had a bike, got books, and rode around identifying things. I can still remember identifying my first titmouse.”

LII Director Tom Bruce says, “Brian is a constant source of astonishment. I mean, this is a guy who labels his spice jars with the scientific names of the plants. But the thing that always makes my jaw drop is his music collection, especially modern stuff. He’s got Scriabin and Stravinsky, John Adams and John Harbison, and I’d bet money that there’s some Esa-Pekka Salonen on the shelf too.”

Now Brian’s back yard is full of birdfeeders. While programming the donor systems, Brian can listen to Esa-Pekka Salonen and watch the titmice forage for seeds.

BOOM WinnersOctober 1st was a big day at Cornell. Bill Gates was on campus for the dedication of Bill & Melinda Gates Hall, the new home of Cornell’s Faculty of Computing and Information Science. In his remarks, Gates spoke of the “the importance of interdisciplinary collaboration in advancing computer science”. That’s an idea we like a lot – it’s what made LII possible from the start.

The Gates Hall dedication also gave us an opportunity to catch up a bit with two students we’d worked with last year. Cornell Masters of Engineering  (and LII project team) alumnae Deepthi Rajagopalan and Neha Kulkarni had been invited to return to campus to attend the dedication ceremony and present their project. Their team, won the Googliest project award at the BOOM science fair and the faculty-selected departmental M.Eng. project award last year. The project involved the use of advanced natural-language processing techniques to identify definitions in the Code of Federal regulations and determine their scope. It was gratifying to see them receive further distinction for their work.

The project was in many ways a model of interdisciplinary collaboration between engineering students, who researched the performance of several techniques for extracting the definitions, and law students, who produced gold-standard data for the engineering students to use for training and evaluating their software. The underlying purpose is to help people who need to read and understand regulations know which terms in the text they’re reading have been explicitly defined and get access the definitions for those terms.

After a frantic but fruitful search for a year-old project poster, we got a chance to catch up a bit on how we’ve ended up building the CFR definition feature for the web site (for example, at http://www.law.cornell.edu/cfr/text/9/1.1). Deepthi (now at Oracle in San Francisco) and Neha (now at Nomura in New York) hadn’t forgotten the challenges of working with legal text – convoluted sentences, paragraph nesting, enumeration – and had a wealth of experience to share.

Deepthi and Neha are two of more than 30 M.Eng. students in computer science who have worked with LII over the years. This year, M.Eng. students are working on mapping financial concepts and explanatory materials to financial regulations in the CFR, and MPS students are working on the visual presentation of law related to hydrofracking.