Posts tagged ‘curriculum’
The Story of MACOS: How getting curriculum development wrong cost the nation, and how we should do it better
Man: A Course of Study (MACOS) is one of the most ambitious US curriculum efforts I’ve ever heard about. The goal was to teach anthropology to 10 year olds. The effort was led by world-renowned educational psychologist Jerome Bruner, and included many developers, anthropologists, and educational psychologists (including Howard Gardner). It won awards from the American Education Research Association and from other education professional organization for its innovation and connection to research. At its height, MACOS was in thousands of schools, including whole school districts.
Today, MACOS isn’t taught anywhere. Funding for MACOS was debated in Congress in 1975, and the controversy led eventually to the de-funding of science education nationally.
Peter Dow’s 1991 book Schoolhouse Politics: Lessons from the Sputnik Era is a terrific book which should be required reading for everyone involved in computing education in K-12. Dow was the project manager for MACOS, and he’s candid in describing what they got wrong. It’s worthwhile understanding what happened so that we might avoid it in computing education. I just finished reading it, and here are some of the parts that I found particularly insightful.
First, Dow doesn’t dismiss the critics of MACOS. Rather, he recognizes that the tension is between learning objectives. What do we want for our children? What kind of society do we want to build?
I quickly learned that decisions about educational reform are driven far more by political considerations, such as the prevailing public mood, than they are by a systematic effort to improve instruction. Just as Soviet science supremacy had spawned a decade of curriculum reform led by some of our most creative research scientists during the late 1950s and 1960s, so now a new wave of political conservatism and religious fundamentalism in the early 1970s began to call into question the intrusion of university academics into the schools…Exposure to this debate caused me to recast the account to give more attention to educational politics. No discussion of school reform, it seems, can be separated from our vision of the society that the schools serve.
MACOS was based in the best of educational psychology at the time. Students engaged in inquiry with first-hand accounts, e.g., videos of Eskimos. The big mistake the developers made was they gave almost no thought to how it was going to get disseminated. Dow points out that MACOS was academic researchers intruding into K-12, without really understanding K-12. They didn’t plan for teacher professional development, and worse, didn’t build any mechanism for teachers to tell them how the materials should be changed to work in real classrooms. They were openly dismissive of the publishers who might get the materials into the world.
On teachers: There was ambivalence about teachers at ESI. On the one hand the Social Studies Program viewed its work as a panacea for teachers, a liberation from the drudgery of textbook materials and didactic lessons. On the other, professional educators were seen as dull-witted people who conversed in an incomprehensible “middle language” and were responsible for the uninspired state of American education.
On publishers: These two experienced and widely respected publishing executives listened politely while Bruner described our lofty education aspirations with characteristic eloquence, but the discussion soon turned to practical matters such as the procedures of state adoption committees, “tumbling test” requirements, per-pupil expenditures, readability formulas, and other restrictions that govern the basal textbook market. Spaulding and Kaplan tried valiantly to instruct us about the realities of the educational publishing world, but we dismissed their remarks as the musings of men who had been corrupted by commercialism. Did they not understand that our mission was to change education, not submit to the strictures that had made much of instruction so meaningless? Could not men so powerful in the publishing world commit some of their resources to support curriculum innovation? Had they no appreciation of the intellectual poverty of most social studies classrooms? I remember leaving that room depressed by the monumental conservatism of our visitors and more determined than ever to prove that there were ways to reach the schools with good materials. Our arrogance and naivete were not so easily cured.
By 1971, Dow realizes that the controversies around MACOS could easily have been avoided. They had made choices in their materials that highlighted the challenges of Eskimo life graphically, but the gory details weren’t really necessary to the learning objectives. They simply hadn’t thought enough about their users, which included the teachers, administrators, parents, and state education departments.
My favorite scene in the book is with Margaret Mead who tries to help Dow defend MACOS in Congress, but she’s frustrated by their arrogance and naivete.
Mead’s exasperation grew. “What do you tell the children that for?…I have been teaching anthropology for forty years,” she remarked, “and I have never had a controversy like this over what I have written.”
…
But Mead’s anger quickly returned. “No, no, you can’t tell the senators that! Don’t preach to them! You and I may believe that sort of thing, but that’s not what you say to these men. The trouble with you Cambridge intellectuals is that you have no political sense!”
Dow describes over two chapters the controversies around MACOS and the aftermath impacts on science education funding at NSF. But he also points out the problems with MACOS as a curriculum. Some of these are likely problems we’re facing in CS for All efforts.
For example, he talks about why MACOS was removed from Oregon schools, using the work of Lynda Falkenstein. (Read the below with an awareness of the Google-Gallup and EdWeek polls showing that administrators and principals are not supportive of CS in schools.)
She concluded that innovations that lacked the commitment of administrators able to provide long-term support and continuing teacher training beyond the initial implementation phase were bound to faster regardless of their quality. Even more than controversy, she found, the greatest barrier to successful innovation was the lack of continuity of support from the internal structure of the school system itself.
I highly recommend Schoolhouse Politics. It has me thinking about what it really takes to get any education reform to work and to scale. The book is light on evaluation evidence that MACOS worked. For example, I’m concerned that MACOS was so demanding that it may have been too much for underprepared students or teachers. I am totally convinced that it was innovative and brilliant. One of the best curriculum design efforts I’ve ever read about, in terms of building on theory and innovative design. I am also totally convinced that it wasn’t ready to scale — and the cost of that mistake was enormous. We need to avoid making those mistakes again.
A Generator for Parsons problems on LaTeX exams and quizzes
I just finished teaching my Introduction to Media Computation a few weeks ago to over 200 students. After Barb finished her dissertation on Parsons problems this semester, I decided that I should include Parsons problems on my last quiz, on the final exam study guide, and on the final exam. Parsons problems are a great fit for this assessment task. We know that Parsons problems are a more sensitive measure of learning than code writing problems, they’re just as effective as code writing or code fixing problems for learning (so good for a study guide), and they take less time than code writing or fixing.
Barb’s work used an interactive tool for providing adaptive Parsons problems. I needed to use paper for the quiz and final exam. There have been several Parsons problems paper-based implementation, and Barb guided me in developing mine.
But I realized that there’s a challenge to doing a bunch of Parsons problems like this. Scrambling code is pretty easy, but what happens when you find that you got something wrong? The quiz, study guide, and final exam were all going to iterate several times as we developed them and tested them with the teaching assistants. How do I make sure that I always kept aligned the scrambled code and the right answer?
I decided to build a gadget in LiveCode to do it.
I paste the correctly ordered code into the field on the left. When I press “Scramble,” a random ordering of the code appears (in a Verbatim LaTeX environment) along with the right answers, to be used in the LaTeX exam class. If you want to list a number of points to be associated with each correct line, you can put a number into the field above the solution field. If empty, no points will be explicitly allocated in the exam document.
I’d then paste both of those fields into my LaTeX source document. (I usually also pasted in the original source code in the correct order, so that I could fix the code and re-run the scramble when I inevitably found that I did something wrong.)
The wording of the problem was significant. Barb coached me on the best practice. You allow students to write just the line number, but encourage them to write the whole line because the latter is going to be less cognitive load for them.
Unscramble the code below that halves the frequency of the input sound.
Put the code in the right order on the lines below. You may write the line numbers of the scrambled code in the right order, or you can write the lines themselves (or both). (If you include both, we will grade the code itself if there’s a mismatch.)
The problem as the student sees it looks like this:
The exam class can also automatically generate a version of the exam with answers for used in grading. I didn’t solve any of the really hard problems in my script, like how do I deal with lines that could be put in any order. When I found that problem, I just edited the answer fields to list the acceptable options.
I am making the LiveCode source available here: http://bit.ly/scrambled-latex-src
LiveCode generates executables very easily. I have generated Windows, MacOS, and Linux executables and put them in a (20 Mb, all three versions) zip here: http://bit.ly/scrambled-latex
I used this generator probably 10-20 times in the last few weeks of the semester. I have been reflecting on this experience as an example of end-user programming. I’ll talk about that in the next blog post.
Teach two languages if you have to: Balancing ease of learning and learning objectives
My most recent CACM Blog post addresses a common question in computer science education: Should we teach two programming languages in a course to encourage abstraction, or just one? Does it hurt students to teach two? Does it help them to learn a second language earlier? My answer (in really short form) is “Just teach one, because it takes longer to learn one than you expect. If you teach two or more, students are going to struggle to develop deep understanding.”
But if your learning objective is for students to learn two (or more languages), teach two or more languages. You’re going to have to pay the piper sometime. Delaying is better, because it’s easier and more effective to transfer deep knowledge than to try to transfer surface-level representations.
The issue is like the question of recursion-first or iterative-control-structures-first. (See this earlier blog post.) If your students don’t have to learn iterative control structures, then teach recursion-only. Recursion is easier and more flexible. But if you have to teach both, teach iteration first. Yes, iteration is hard, and learning iteration-first makes recursion harder to learn later, but if you have to do it, iteration-first is the better order.
There’s a lot we know about making computing easier to learn. But sometimes, we just can’t use it, because there are external forces that require certain learning objectives.
I correct, continue, and explore tangents on this blog post here: https://computinged.wordpress.com/2018/06/15/are-you-talking-to-me-interaction-between-teachers-and-researchers-around-evidence-truth-and-decision-making/
Stanford is NOT switching from Java to JavaScript: I was mistaken
Last April, I wrote a blog post saying that Stanford was abandoning Java for JavaScript in their intro course (see post here). The post was initiated by an article in the Stanford Daily. The post caused quite an uproar, way more than I expected. More than one Stanford faculty member reached out to me about it. In particular, Marty Stepp told me that I was definitely wrong, that Stanford would mostly be teaching Java in a year. I promised that if I was wrong a year later, I would write another post correcting my first post.
It’s been a year, and I was wrong. Stanford is NOT abandoning Java for JavaScript.
I’m glad I was wrong, but it has nothing to do with Java or JavaScript.
I heard about the possible switch to JavaScript several months before from a Stanford faculty member. When I saw the Stanford Daily article, I thought it was okay to talk about it. Marty told me at the time that I was wrong, and that the article was ill informed. Still another Stanford faculty member wrote me about the tensions over this issue.
A lesson I learned from Mike Lach and others involved in the NGSS roll out is that all curricular decisions are political decisions. A framework might be based on scientific expertise, but what is actually taught is about choice and vision — different opinions of how we interpret where we are now and what we want in the future. If you haven’t heard about the politics of curricular choices before, I highly recommend Schoolhouse Politics.
I am not at Stanford, so I don’t know how curricular decisions have been made and were made here. I based my post on talking with some Stanford faculty and reading the Stanford Daily article. I predicted that the forces pushing for JavaScript would end up changing the curriculum. They didn’t (or haven’t so far). The Stanford lecturers are excellent, and they are the ones actually teaching those classes. I’m glad that they get to continue teaching the classes the way that they think is most valuable.
Below is what Marty wrote me about the courses at Stanford, and a link to the Stanford course offerings, showing that Stanford is still primarily a Java house:
This calendar year our CS1 Java course is still quite clearly the dominant course. Nick Parlante is also teaching two smaller experimental offerings of a Python class in our winter and spring quarters. There may be another experimental JavaScript and/or Python course on the books for fall, but it certainly will not be the main class; the CS1 in Java will continue to be so throughout all of the next academic year. Currently no plan is under way to change that, though we certainly are open to evolving our courses in the long term like any other school would be. I would like to note that the state of intro at Stanford is exactly as was described to you by myself and others 10 months ago.
Keeping the Machinery in Computing Education: Back to the Future in the Definition of CS
I’ve been excited to see this paper finally come out in CACM. Richard Connor, Quintin Cutts, and Judy Robertson are leaders in the Scotland CAS effort. Their new curriculum re-emphasizes the “computer” in computer science and computational thinking. I have bold-faced my favorite sentence in the quote below. I like how this emphasis reflects the original definition of computer science: “Computer science is the study of computers and all the phenomena surrounding them.”
We do not think there can be “computer science” without a computer. Some efforts at deep thinking about computing education seem to sidestep the fact that there is technology at the core of this subject, and an important technology at that. Computer science practitioners are concerned with making and using these powerful, general-purpose engines. To achieve this, computational thinking is essential, however, so is a deep understanding of machines and languages, and how these are used to create artifacts. In our opinion, efforts to make computer science entirely about “computational thinking” in the absence of “computers” are mistaken.
As academics, we were invited to help develop a new curriculum for computer science in Scottish schools covering ages 3–15. We proposed a single coherent discipline of computer science running from this early start through to tertiary education and beyond, similar to disciplines such as mathematics. Pupils take time to develop deep principles in those disciplines, and with appropriate support the majority of pupils make good progress. From our background in CS education research, we saw an opportunity for all children to learn valuable foundations in computing as well, no matter how far they progressed ultimately.
Source: Keeping the Machinery in Computing Education | November 2017 | Communications of the ACM
Survey to inform the next round of Computing Curricula
We have to teach where the students are: Response to “How We Teach Should Be Independent Of Who We Are Teaching”
Valerie Barr has great insights into computing education, especially with regards to diversity (e.g., see the blog post last CS Ed Week about alternative ways to view data about diversity in computing). I like what she has to say in her most recent Blog@CACM blog post, but I think the title is somewhat misleading.
“How we teach should be independent of who we are teaching” is clearly not true. No one would argue for teaching Linux kernel developing via all day long bootcamps in C to middle school students. Few people use CS Unplugged with machine learning graduate students. What Valerie is explicitly addressing in her blog post is an issue called essentialism.
As we continue efforts to diversify computing, we cannot afford to paint any group in a monochromatic way. We have to embrace the richness of today’s student population by making what we teach meaningful and relevant to them. There are women who want to geek out about hard-core tech, and there are men who care deeply about computing for the social good. There are students of all genders and ethnic and racial backgrounds who will be happy with an old-fashioned lecture, and those who will thrive on active learning with examples drawn from a range of cultures and application areas. Many students will be motivated by knowing how the techniques and subject matter they’re learning fit into their future workplace or life goals.
Source: How We Teach Should Be Independent Of Who We Are Teaching | blog@CACM | Communications of the ACM
Here’s a definition of essentialism (from the Geek Feminism Wiki):
The concept of Essentialism states that there are innate, essential differences between men and women. That is, we are born with certain traits. This is often used as an explanation for why there are so few women in science and technology.
In contrast, the critical issue is who is in your classroom, what do they know, and what are their motivations. As How People Learn describes it:
There is a good deal of evidence that learning is enhanced when teachers pay attention to the knowledge and beliefs that learners bring to a learning task, use this knowledge as a starting point for new instruction, and monitor students’ changing conceptions as instruction proceeds.
This is hard to do. We can’t redesign every class for each new student population. What I think Valerie is admonishing us to do is to actually check and not assume certain interests and motivations because of the demographics of the students. When we were developing Media Computation, we did focus groups with students to get their feedback on our developing designs. We surveyed the students to get a sense of what they were interested in and what motivated them. Great work like Unlocking the Clubhouse suggested our starting point, but we did not assume that the majority-female class would have stereotypical responses. We checked with our student population, and we provided different kinds of media interactions to attract different kinds of students within our population.
It would be best if we could provide educational opportunities that meet each student’s needs individually. Short of that, we can design for the students who enter our classrooms, not for the stereotypes that we might expect.
Have We Reached a Consensus on a National CS Curriculum? I hope not
Alfred Thompson raises an important question here. I agree with him — we haven’t reached consensus. We also will never have a national CS curriculum in the United States, because we have a distributed education model. It’s a state decision. I do fear that there may be a de facto standard now.
But the bigger concern is at a higher level of abstraction: How should we make curricular decisions in CS (or anywhere else)? I hope that we make our decisions based on empirical evidence. I don’t see that we have the empirical evidence that any of the below classes ought to be the dominant model.
Oh boy are things up in the air in the HS CS curriculum these days. While we have some great advice from the CSTA (CSTA K-12 Computer Science Standards) the implementation of those standards are still left up to individual schools/districts/states. Still it is easy to come to the conclusion from watching social media and some conferences that there is a consensus on a high school Computer Science curriculum. Today I got the following from a friend.
Is it an incorrect read or has a national consensus for CS in HS’s been achieved with a sequence of :
–ECS (Exploring Computer Science) Curriculum
–CS Principles/BJC Curriculum (Beauty and Joy of Computing)
–AP CS (JAVA [for now])
via Computer Science Teacher: Have We Reached a Consensus on a National CS Curriculum?.
Mehran Sahami wins ACM Presidential Award for the CS2013 Curriculum Revision
This is really well-deserved. Mehran worked amazingly hard to pull a wide range of stakeholders together for the CS2013 curriculum. The ACM Presidential Award is discretionary — they only give it out if someone really deserves it. Glad to see Mehran getting this recognition!
- For outstanding leadership of, and commitment to, the three-year ACM/IEEE-CS effort to produce CS2013 a comprehensive revision of the curricular guidelines for undergraduate programs in computer science
Mehran Sahami of Stanford University, recipient of the ACM Presidential Award for leading the revision of an innovative computer science curriculum that reflects the application of computing tools in a wide variety of disciplines. Sahami led the effort by ACM and the IEEE Computer Society to develop guidelines for undergraduate degree programs that redefine essential computing topics and set the standards for computer science education worldwide for the next decade. The report includes examples of flexible courses and curricula models for a broad range of higher education institutions worldwide.
I don’t believe it: Early STEM Education Will Lead to More Women in IT
I don’t believe the main propositions of the article below. Not all STEM education will lead to more women discovering an interest in IT. Putting computing as a mandatory subject in all schools will not necessarily improve motivation and engagement in CS, and it’s a long stretch to say that that will lead to more people in IT jobs.
I addressed the quote below, by Ashley Gavin, in my Blog@CACM post for this month: The Danger of Requiring CS in US K-12 Schools.
“You make it an option, the girl is not going to take it. You have to make it mandatory and start it at a young age,” says Ashley Gavin, curriculum director at Girls Who Code, a nonprofit working to expose more girls to computer science at a young age that has drawn support from leading tech firms such as Google, Microsoft and Intel.
“It’s important to start early because, most of the fields that people go into, they have exposure before they get to college. We all study English before we get to college, we all study history and … social studies before we get to college,” Gavin says. “No one has any idea what computer science is. By the time you get to college, you develop fear of things you don’t know. Therefore early exposure is really important.”
via Early STEM Education Will Lead to More Women in IT – CIO.com.
The ACM/IEEE 2013 CS Curriculum is released (in the nick of time!)
Posted by Mehran Sahami to the SIGCSE members list. Congratulations to the team for finishing it in time.
Dear Colleagues,
We are delighted to announce the release of the ACM/IEEE-CS Computer Science
Curricula 2013 (CS2013) Final Report. The report is available at the CS2013
website (http://cs2013.org) or directly at:
http://cs2013.org/final-draft/CS2013-final-report.pdf
(The report will also soon be posted at the ACM website as well as at
doi.org.)
The CS2013 Final Report contains guidance for undergraduate programs in
computer science, including a revised Body of Knowledge, over 80 course
exemplars (showing how the CS2013 Body of
Knowledge may be covered in a variety of actual fielded courses), and 5 full
curricular exemplars from a variety of educational institutions. The report
also contains discussions of characteristics of CS graduates, design
dimensions in introductory courses, and institutional challenges in CS
programs, among other topics. The report has been endorsed by both the ACM
and IEEE-Computer Society. We hope you find it useful.
CITING THE CS2013 FINAL REPORT
To cite the CS2013 report, please use the canonical citation provided below
in ACM format and BibTex.
ACM format:
ACM/IEEE-CS Joint Task Force on Computing Curricula. 2013. Computer Science
Curricula 2013. ACM Press and IEEE Computer Society Press. DOI:
http://dx.doi.org/10.1145/2534860
BibTex:
@techreport{CS2013,
title = {Computer Science Curricula 2013},
author = {ACM/IEEE-CS Joint Task Force on Computing Curricula},
month = {December},
year = {2013},
institution = {ACM Press and IEEE Computer Society Press},
url = {http://dx.doi.org/10.1145/2534860},
doi = {10.1145/2534860}
}
Warm regards,
Mehran Sahami and Steve Roach
Co-Chairs, CS2013 Steering Committee
CS2013 Steering Committee
ACM Delegation
Mehran Sahami, Chair (Stanford University)
Andrea Danyluk (Williams College)
Sally Fincher (University of Kent)
Kathleen Fisher (Tufts University)
Dan Grossman (University of Washington)
Beth Hawthorne (Union County College)
Randy Katz (UC Berkeley)
Rich LeBlanc (Seattle University)
Dave Reed (Creighton University)
IEEE-CS Delegation
Steve Roach, Chair (Exelis Inc.)
Ernesto Cuadros-Vargas (Univ. Catolica San Pablo, Peru)
Ronald Dodge (US Military Academy)
Robert France (Colorado State University)
Amruth Kumar (Ramapo College of New Jersey)
Brian Robinson (ABB Corporation)
Remzi Seker (Embry-Riddle Aeronautical Univ.)
Alfred Thompson (Microsoft)
Why Flipping Classrooms Might Not Make Much Difference
This paper is getting a lot of discussion here at Georgia Tech:
In preliminary research, professors at Harvey Mudd College haven’t found that students learn more or more easily in so-called flipped courses than in traditional classes, USA Today reports. In flipped courses, students watch professors’ lectures online before coming to class, then spend the class period in discussions or activities that reinforce and advance the lecture material.
Earlier this year, the National Science Foundation gave four professors at the college in Claremont, Calif., a three-year grant for $199,544 to study flipped classrooms. That research isn’t complete yet, but the professors already tried flipping their own classes last year and found “no statistical difference” in student outcomes.
The reason why it’s generating a lot of discussion is because we know that it can make a difference to flip a classroom. Jason Day and Jim Foley here at Georgia Tech did a careful and rigorous evaluation of a flipped classroom seven years ago (see IEEE paper on their study). We all know this study and take pride in it — it was really well done. It can work. The Harvey Mudd study also shows that it can be done in a way that it doesn’t work.
That’s really the story for all educational technology. It can work, but it’s not guaranteed to work. It’s always possible to implement any educational technology (or any educational intervention at all) in such a way that it doesn’t work.
CS Teacher Repositories: CS OER Portal, Ensemble, CSTA, CAS, and…
I just received this via email:
We would like to inform you that we have added recently many new resources to the Computer Science Open Educational Resources Portal (CS OER Portal) (http://iiscs.wssu.edu/drupal/csoer ). For those of you who have not visited it, the Portal hosts a rich collection of links to open teaching/learning materials targeted specifically the area of Computer Science. It provides multiple ways for locating resources, including search with filtering the results by CS categories, material type, level, media format, etc., as well as browsing by institutional (OpenCourseWare) collections, by CS categories, or by topics as recommended by the ACM/IEEE Computer Science Curriculum. The browsing functionality is supplemented with recommendations for similar courses/resources.
My first thought was, “Is this competition for Ensemble, the big NSF-sponsored digital library of CS curricular materials?”
If we’re specifically thinking just about computing in schools (K-12 in the US), we should also consider the CSTA Source Web Repository and the Resources section of the Computing at Schools website (which is pretty big and growing almost daily).
Specifically for a particular tool or approach, there’s the Greenfoot Greenroom, ScratchEd for Scratch Teachers and other Educators, the Alice teacher’s site, the TeaParty site (for the Alice + MediaComp website), and of course, the Media Computation site. I’m sure that there are many others — for particular books (like this one introducing Python with Objects), for particular curricular approaches (like Exploring Computer Science and CSUnplugged), and even for particular methods (I reference the Kinesthetic Learning Activities site in my TA preparation class).
It’s really great that there are so many repositories, so many resources to help CS teachers, and so many people developing and sharing resources. I get concerned when I’m in a meeting where we’re talking about how to help CS teachers, and someone suggests (and it really happens in many of the meetings I attend), “If we only had a repository where teachers could find resources to help them…” No, I really don’t think that more repositories is going to solve any problems at this point.
Strong vision drives growth in CS course at Princeton
The course at Princeton does sound really cool, but I’m not convinced that the course content/curriculum is driving the growth. As we found with MediaComp, the curriculum seems to have little to do with enrollment in a course. I wonder what comparable courses (say, at Harvard or Yale) look like in terms of enrollment.
I strongly agree with the argument that they’re making below for the importance of computational literacy.
Sedgewick said he is pleased that the course leads many students to a greater interest in computer science, but he feels strongly that computers are so integral to modern society that a basic understanding of the field should be a part of any education. In the introduction to their textbook, “Introduction to Programming in Java,” Sedgewick and Wayne say that in the modern world computer science cannot be left to specialists.
“The basis for education in the last millennium was reading, writing and arithmetic; now it is reading, writing and computing,” they write.
Writing programs using ordinary language: Implications for computing education
Once upon a time, all computer scientists understood how floating point numbers were represented in binary. Numerical methods was an important part of every computing curriculum. I know few undergraduate programs that require numerical methods today.
Results like the below make me think about what else we teach that will one day become passé, irrelevant, or automatized. The second result is particularly striking. If descriptions from programming competitions can lead to automatic program generation, what does that imply about what we’re testing in programming competitions — and why?
The researchers’ recent papers demonstrate both approaches. In work presented in June at the annual Conference of the North American Chapter of the Association for Computational Linguistics, Barzilay and graduate student Nate Kushman used examples harvested from the Web to train a computer system to convert natural-language descriptions into so-called “regular expressions”: combinations of symbols that enable file searches that are far more flexible than the standard search functions available in desktop software.
In a paper being presented at the Association for Computational Linguistics’ annual conference in August, Barzilay and another of her graduate students, Tao Lei, team up with professor of electrical engineering and computer science Martin Rinard and his graduate student Fan Long to describe a system that automatically learned how to handle data stored in different file formats, based on specifications prepared for a popular programming competition.
via Writing programs using ordinary language – MIT News Office.
Computing Ed Research - Guzdial's Take 
Recent Comments