IT’s Academic

Managing Content on the Princeton Web

2009-11-11T21:23:46Z

During the past decade, Princeton’s web site has grown from a relatively simple tool supporting producers and consumers of information about the institution, its programs, and its people to what is today a complex, mission-critical appliance for teaching, research, administration, and collaboration.

Such complex web sites publish and sustain every day vast amounts of time sensitive information. To manage the mountain of content, Princeton has turned to Content Management Systems that offer an integrated set of powerful features for creating, storing, versioning, and publishing everything from news articles and brochures though audio, video, and images.

]]> For the last three years, Henry Umansky has managed the University’s Web Programming team. At the November 11 Lunch ‘n Learn seminar, he reviewed the concepts and philosophies underlying Content Management Systems and debated their pros and cons.

Large web efforts should begin, he recommends, by assembling key stakeholders, by defining the site’s short- and long-term goals, by querying users about their needs, likes, and dislikes, and by creating a comprehensive information architecture.

Umansky showed a video from the popular show, the IT crowd, to emphasize why IT architecture is so important. Above all, users want to be able to gain access to the information they want quickly and intuitively.

The information architecture tools used to gather up and organize the content for a web site include data sheets, user stories, concept maps, bullseye diagrams, and navigation maps. Data sheets gather up general and specific information about the target audience. User stories reflect how individuals use or want to use the sites. Some developers and designers brainstorm with concept maps, essentially an annotated set of lines that interconnect the key areas within the domain. Bullseye diagrams help to limit scope creep by prioritizing tasks. Navigation maps define and group all of the landing pages, a key step towards understanding the full navigation schema.

“Content is King,” Says Umansky. Complex sites keep users coming back by keeping content current. Useful content requires a careful review process, freshness dating, and tailoring the writing for the intended audience. Throughout, developers must listen to and get to know their audience. A successful site, will therefore involve usability testing, customer satisfaction surveys, focus groups, and even market research.

CMS tools aid the effort by organizing and easing access to content with metadata, indices, and consistent storage and retrieval, security, workflow, versioning, and archiving. They aid productivity by providing automated, reusable templates, by separating content from design, and by simplifying the editing and approval of content.

Princeton relies upon the Roxen CMS for its web efforts, but Umansky notes that the University actually maintains many other CMSs. WebSpace, SharePoint, and DSpace encourage academic and administrative collaboration, Blackboard provides access to course related material, and Movable Type supports blogging and wikis.

A podcast and the presentation are available.

Princeton's Partnership with Google Books

2009-11-04T21:28:20Z

The Google Books Library Project is a collaborative effort between Google and more than 20 academic libraries and publishers to scan and make searchable major research collections. When books are out of copyright and in the public domain, the public can now use Google Book Search to view bibliographic information, to read and search the texts, and even download them.

Google Books is a product resulting from the combination of the Google Library Partnerships (29 libraries) and the Google Publishers Partnership (many thousands)

The library project began in 2005 with Harvard, the New York Public Library, Stanford University, the University of Michigan, and Oxford University.Princeton joined the project in 2006. Many other institutions of higher education and several publishers have now joined the endeavor.

]]> When Princeton joined, University Librarian Karin Trainer remarked:

“Generations of Princeton librarians have devoted themselves to building a remarkable collection of books in thousands of subjects and dozens of languages. Having the portion of that collection not covered by copyright available online will make it easier for Princeton students and faculty to do research, and joining the Google partnership allows us to share our collection with researchers worldwide, a step very much in keeping with the University’s unofficial motto of Princeton in the nation’s service and in the service of all nations.”

In the November 4 Lunch ‘n Learn, three members of the University Library described recent developments in the Library’s Partnership Project with Google, provided an update on the Google Settlement from the Library’s perspective.

As Head of the Donald E. Stokes Library for Public and International Affairs and Population Research, Nancy Pressman-Levy is regularly involved in teaching the use of Library electronic resources to all Library patrons. She began the talk by demonstrating some of the recent new features in Google Book Search, which permits you to search the full text of the books that Google is scanning as part of this project.

She has assembled an online guide to complement the talk.

When books are out of copyright, you can see books in Full View. You can see a more limited view if publishers or the author have given the right to view works in copyright. For some works, you have a Snippet View that provides some context around your search, you can create your own libraries, you can zoom in and out of scanned text, and you can clip portions of text or images. She also showed how you can link to such searches directly from the Main Catalog and other Library systems.

Richard J. Schulz is Associate University Librarian for Technical Services. His staff oversees the project in which Google is scanning many of the Library’s books. He briefly reviewed the operational side of the project. Princeton has provided a scope list of the books in our location that are in the public domain. Google returns a list of the books they want to scan. Prior to shipping, University staff assess the condition of the books. Some are too fragile, but most can be sent. Each book is carefully logged such that it can be returned quickly if a patron requests it. Accessibility to the collection is never lost.

Princeton and Google have copious guidelines regarding the condition of books, and some books in the scanning process require special attention, but in the end, 99.6% of the books that are sent are scanned.

In February, Karin Trainer estimated in the Daily Princetonian 2007 that the project would involve six years and the full text of approximately one million books. More than half of the books sent are now available in full text. Another third are available in Snippet mode. We are now looking now at special collections books to determine whether some subset can be sent.

Marvin Bielawski is Deputy University Librarian and Head of the Library Systems Office. He has been involved in negotiating the contract with Google and the projected settlement amendment.

He noted that project critics have pointed out flaws in Google Book Search. Google is aware of these issues and has assembled a huge team to work towards more accurate metadata, better handling of linking multi-volume works, and to improve the quality of OCR and image capture.

The Authors Guild of the United States and the Association of American Publishers sued Google in a class action suit. The main concern was that Google ought not to have scanned the in-copyright materials from partner libraries. Google has maintained throughout that their approach fell under fair use. A revised settlement agreement is due in court this November 9. Bielawski said that a provision of the settlement would create a new independent entity, a Book Rights Registry staffed by representatives from the Authors Guild and Association of American Publishers. They will distribute the royalties that will come from Google’s income related to the in-copyright material.

Google in turn will be able to create a commercial model, and an institutional model for access. Anyone on the internet will be able to buy and download one of these books. The institutional model will be subscription based for cultural and educational institutions. Full text of much of the in-copyright material will be available in the subscription.

If the Settlement is approved, Princeton will be able to subscribe, thus unlocking a tremendous amount of content for our users. And, if we sign on as a Full Partner in the Settlement, we will have the option of sending our in-copyright material to Google for scanning and adding to the corpus.

A podcast and a presentation are available.

Digital Telephony at Princeton

2009-10-28T16:22:29Z

Voice over Internet Protocol (VoIP) refers to a family of hardware and software technologies that deliver voice communications over the internet rather than the public, switched telephone network. To make it work, technologists have to convert traditional analog signals into a digital format and then translate that signal into IP packets for transmission over a private or public network.

At the October 28 Lunch ‘n Learn, Dave Wirth, Manager of Operations within the Office of Information Technology’s (OIT) Networking/Telephony group, reviewed the technology, its present implementation at Princeton, and the University’s plans for the future.

]]> VoIP systems have actually been in use at Princeton, mostly by members of OIT’s staff, for seven years. The experimental equipment, primarily technologies that interfaced with our existing telephone switch, permitted us to learn a great deal, says Wirth, but as we upgrade our current switch, all of that equipment is now being replaced. The new VoIP trial will be taking place at 701 Carnegie Center, the new home for OIT on Canal Point Road.

There are many myths about VoIP, emphasizes Wirth. Such services are not totally free, VoIP lines cannot survive electrical blackouts, and dial-up Internet connectivity will not generally provide satisfactory quality.

Princeton is relying on an advanced yet simpler VoIP interface, SIP (Session Initiated Protocol). Says Wirth, SIP is modular and scalable, easier to design and update, and internet friendly. From the users’ perspective, one of the most attractive benefits of SIP is the ability to make a call based on an e-mail or IM address rather than always needing a telephone number.

There are advantages to tradition telephony. The conventional handsets are less expensive, and our present telephone systems are remarkably reliable; by contrast VoIP systems rely on hardware platforms that have not benefitted from the years of experience of the telephone companies. We will now have to worry about echoes, latency, and jitters. As a result, alarms and 911 calls may continue to rely on traditional copper circuits for quite some time.

With old systems, you call a number. With new systems, you call a person. Philosophically, that’s a big difference, emphasizes Wirth. IP systems will offer more attractive pricing, far more features, and for most users better and richer ways of communicating. With VoIP, you can purchase phones from a wide variety of vendors, and you can combine voice, data, and images on the same network. The result over time will be a range of interesting and compelling applications for personal and departmental communication and collaboration.

Primarily to ensure the best voice quality and for security reasons, the possibility of viruses and denial of service attacks, Princeton has chosen to keep its data and VoIP networks separate. In addition to being able to test various new applications for eventual use on campus, the VoIP system will permit staff in the new building to retain their 258 campus extensions.

So what’s in it for users? There are intriguing benefits, including conferencing, call forwarding, and auto redial features that normally cost more money. Among the new features, there are distinctive ring tones, and everyone will have a display in their phone to view CallerID and to view a phone log of recent calls that supports Click-to-Call, the ability to click on a name or number and connect. Among other features, staff will gain the ability to set up ad hoc conferences and to set up global and personal directories.

When will these benefits extend to the rest of campus? The infrastructure is demanding and so, new construction like the Chemistry building will be the first to benefit, says Wirth. The rest of us, unfortunately, will have to wait.

A podcast and the presentation are available.

Blogs, Lulz and Tweets: Social Media Comes to Princeton

2009-10-14T21:01:26Z

Why has the use of Facebook and other social networking sites exploded? Perhaps, suggest John Jameson and Shani Hilton of Princeton’s Office of Communications, because it is now possible to interact socially with very large numbers of people in ways that are no more difficult than sending out a simple e-mail.

Most users need not worry about the coding or the construction of their pages. They can simply concern themselves with what they should share, and not share.

The technologies are changing rapidly (MySpace, for example, has lost 20% of their users in just two months), bringing enormous opportunities, challenges, and some significant policy headaches.

]]> At the October 15 Lunch ‘n Learn seminar, Shani Hilton and John Jameson showed examples of social networking from the University’s institutional presence as well as faculty activity within the social Web.

There are, for example, numerous faculty blogs, in many ways the entry point to social media. In these, the social interaction is limited to comments about each post and popularity depends primarily upon each author’s ability to post regularly, to be interesting or provocative, and it certainly helps to respond usefully to comments. Almost all of the blogs offer the ability to search, to subscribe, and to comment.

Ed Felton’s blog, Freedom to Tinker, from the Center for Information Technology Policy, is notable because, in addition to tapping his own celebrity status to entice a large audience, he offers to more than half a dozen academics the opportunity to post there regularly, essentially sharing the burden of creating fresh content. The result is an interactive and engaging academic journal.

Perhaps the most read faculty blog is Paul Krugman’s Blog at the New York Times. By partnering with the mainstream media, faculty gain a pre-built audience and a staff to help maintain the traffic. It is a tempting approach which nonetheless sacrifices some independence in format if not in thought.

University departments also have blogs. Obviously, you are reading one right now. Hilton and Jameson also showed off EQN, the School of Engineering and Applied Science’s blog, essentially an extension of the news section of their web site.

Departments interested in starting a blog should go to blogs.princeton.edu which will assist with the setting up a Moveable Type weblog.

A few faculty are also using Twitter. Melissa Harris Lacewell, a professor of politics and African American Studies and a frequent contributor on nightly news stations also shares prolific “tweets” to more than 7,800 followers. The Office of Communications offers up tweets to communicate informally about events of relevance to the community.

Princeton’s YouTube channel has 950 subscribers. It informally describes campus and academic life, and offers some fun videos including a fly-over of the campus.

Princeton’s Facebook site now has 9,000 fans. Career Services and even Lunch ‘n Learn have their own Facebook pages.

As it turns out, these kinds of sites are enormously popular. More traffic comes to Princeton.edu from Facebook, Wikipedia, College Confidential, FARK, and StumbleUpon than any mainstream media news site. As you might imagine, controlling Princeton’s image on these social media sites is not always possible.

“I’m not convinced,” offered one member of the audience. “There’s so much more noise than signal…. And I question if there’s any benefit in this.”

Lauren Robinson Brown, Assistant Vice President of Communications, replied that the University Trustees are extremely interested in uses of social media primarily because the students are there. In addition, the trustees, as parents, have personally experienced the power of these tools. For Princeton, the tools provide accessibility and more flexibility. They open up the University to more of the world, and often in a manner that is less formal than it would appear on either the home page or in more traditional news sources. A story on Digg.com might reach 80,000 readers in an hour, traffic that we never had before.

Hilton and Jameson also explored the role social media can take in a strategic communications plan. For departments and individuals considering establishing or expanding a presence on the social web, Hilton and Jameson review the considerations that should be addressed before taking the plunge. They stress the importance of careful planning, especially when sites represent a department or the University as a whole. Ask the right questions to save time and to increase the department’s impact. Give careful consideration to your audience and what they expect, set policies regarding interaction carefully in order to regulate the amount of maintenance that such sites will require, and correctly incorporate University web policies to avoid potential legal pitfalls.

Shani Hilton and John Jameson chair the Social Media SPIN committee, which consists of campus communicators leading departmental and institutional social media initiatives. In that role, they are drafting best practices for the use of social media at Princeton.

Hilton and Jameson work in the Office of Communications, to which they bring nine years of professional experience and 826 Facebook friends.

A podcast and the presentation are available.

Reassembling the Wall Paintings of Thera

2009-10-05T19:12:22Z

Imagine trying to assemble a jigsaw puzzle without knowing the number of pieces or even what the final image might look like.

The archaeological site of Akrotiri on the small, volcanic island of Thera (modern-day Santorini, Greece) has yielded an unparalleled trove of artifacts and information from the prehistoric Aegean. The ancient trading civilization was destroyed by a volcanic eruption, which buried the remains of a flourishing Late Bronze Age (c. 1630 B.C.) settlement in ash. Among the most significant finds are numerous wall paintings, ranging from every day scenes and coming-of-age rituals to abstract motifs. However, these paintings are recovered as thousands of plaster fragments, and reassembling them consumes a substantial portion of the effort expended at Akrotiri.

]]> At the September 30 Lunch ‘n Learn seminar, Szymon Rusinkiewicz summarized a project that has been on-going in Computer Science at Princeton for nearly three years in collaboration with archaeologists and conservators in Greece and researchers at University College London, KU Leuven, and elsewhere. A team has developed a system that uses 3-D and 2-D digitization hardware, together with computer-based matching techniques, to assist archaeologists and conservators in documenting and reassembling the wall paintings.

Apart from the obvious fact that the frescoes were shattered and then buried by the volcanic eruption, they are very well preserved. Most were created while the plaster was still wet. More flaking of the pigment would have occurred had the frescoes had been painted after the plaster had dried.

The difficulty of reassembling the frescoes varies greatly. The conservators have had a relatively easy time reassembling frescoes that contain designs, but many involve solid colored backgrounds. Hence, there are fragments from as far back as the 1970s that still not reconstructed. The storeroom on site contains crates almost without end, with tens of thousands of fragments.

The pieces are organized by date and location. To assemble a fresco, the conservators usually gather together the crates whose pieces might contribute to that fresco. Nonetheless, pieces from one wall of one room may today be spread among many different crates. Once the easy pattern matches are found, the conservators devote significant time to finding additional matches. Hence, the scale of the problem is very large.

The researchers are scanning in each piece, including detail about shape and contextual cues, with the aim of proposing possible matches. Although mature technologies exist for acquiring images, geometry, and surface normals of small objects, they remain cumbersome and time-consuming for non-experts to employ on a large scale. The newly developed system addresses the scalability, usability, and quality challenges of large-scale 3-D and 2-D digitization, by incorporating new algorithms to align 3-D scans automatically, to register 2-D scans to 3-D geometry, and to compute surface normals from 2-D scans.

To test the system, the conservators constructed a fresco with same characteristics as the ones being excavated, and then professionally destroyed it. The Princeton team then attempted to reconstruct it without any clue about what it ought to look like. They learned that among many variables, the thickness of the fragments and the level of erosion were especially useful cues. And, as it turns out, computers are very good at combining multiple sources of information. Other important features? The front of the frescoes was generally flat, but the back surface had variations and contextual clues that proved useful in finding candidate matches. Door frame pieces were easy to locate because such fragments had thicker portions fitting into the frame. These represent the edges of the puzzle. Strings were originally used to make indentations that served as a guide for painting. They left very faint impressions, a strong cue for matching.

At the talk, Rusinkiewicz explained how the team used the scanners to record the geometries of the fragments. A novel 3-D matching algorithm efficiently searches for matching fragments using the scanned geometric models. In essence, computers can use brute force searching to look for candidate matches by looking for complementary geometries.

Is the effort useful for other applications? Yes, suggests Rusinkiewicz, if you are attempting to reassemble objects that have been broken.

Szymon Rusinkiewicz is an associate professor of Computer Science at Princeton University. His work focuses on acquisition and analysis of the 3D shape and appearance of real-world objects, including the design of capture devices, data structures for efficient representation, and applications (most notably to cultural heritage objects and human skin). He also investigates algorithms for processing complex datasets of shape and reflectance, including registration, matching, completion, symmetry analysis, and sampling. His research interests also include illustrative depiction through line-drawings and non-photorealistic shading models.

A podcast of the talk is available.

Is Academia finally ready for Videoconferencing?

2009-07-14T14:34:42Z

The current recession has persuaded institutions of higher education to look in new places for significant savings. And so, rather than flying cross country for a conference, imagine being able to take part in sessions, or even delivering a paper, right from your office or from a specialized videoconferencing facility on campus.

Professor John Nash and Professor Robert Socolow, for example, have given several keynote addresses via videoconferencing. Says Professor Socolow: “The Princeton Environmental Institute (PEI) and the Center for Human Values sponsored a video conference lecture last spring for the popular Ethics and Climate Change Lecture Series. Robyn Eckersley of the University of Australia at Melbourne presented a virtual lecture entitled: “The Ethics of Carbon Trading” to an audience which was very receptive to the videoconference.” The lecture and more information about the series are available at the ECC website.

]]> Socolow continues: “Videoconferencing gives me the convenience of being more present globally, while at the same time keeping my carbon footprint in check. Kris Kauker of OIT/Media Services has been indispensable to us in helping us work with Web-Ex and other formats to help us use videoconferencing so we can easily connect to faraway places such as New Zealand, Australia, Norway, Portugal, and so on. It’s also a wonderful option for places close by and saves an enormous amount of time and money that travel would require. Kris is a fantastic resource and always goes the extra mile to make things very easy for us on this end. The connections are seamless. It’s as though we are in the same room with those on the other end of the world.”

And, here at Princeton, a working group at Princeton’s Benheim-Thoman Center for Research on Child Wellbeing has been using videoconferencing for nearly five years. A group of faculty, students, post docs and staff at Princeton and Columbia meet twice a month during the semester to present papers and to review their research. Participants here and at Columbia can view the speakers and their Powerpoint slides, and communicate very much like a real gathering. Regina Leidy, Communications Coordinator in Population Research reports that the videoconferences support a healthy discussion and interaction.

Of course, as these examples suggests, videoconferencing is about more than saving money. It may also permit you and other speakers from other campuses to make very productive use of your collective time. You can use videoconferencing for interactive teaching, for research collaboration, for meetings with colleagues at a distance, for interviews on network television, and for sharing your expertise with outside groups.

Imagine bringing a guest lecturer from a distant locale into your class without infringing on their time or the huge expense of travel, or sharing your own expertise in a lecture at Harvard. You could also use campus’s videoconferencing technology to consult with outside staff and faculty who share your interests, to even to conduct interviews with potential job candidates or students. Videoconferencing simply involves two-way audio and video communication.

Princeton makes available to its faculty, staff, and students several different types of videoconferencing facilities that support a full range of audio-visual interactions.

But be sure to plan ahead. Socolow emphasizes: “With each use of videoconferencing, we are reminded that the technology is in its infancy. Each connection takes advanced planning, and most connections require at least a dress rehearsal. In most cases, Princeton is more advanced than our partner, who struggles to make the connection work. One of the reasons for choosing videoconferencing is to encourage the development of capability in other institutions.”

Videoconference Rooms at Princeton

There are five Videoconferencing rooms on campus with specialized equipment that will permit individuals or groups to interact with a remote group or even multiple groups.

One of the rooms is at Friend Center, two are in Robertson Hall, one is in Wallace Hall, and one is at the Peter B. Lewis Science Library. All five rooms support Internet (IP) and specialized telephone (ISDN) connections. There is an additional charge for telephone connections, but Internet-based sessions have no connection fee. Internet connections are encouraged whenever possible because Telephone videoconferences can become quite costly.

Desktop Videoconferencing

The University supports desktop videoconferencing, essentially one-to-one interactions or one-to-many interactions directly from your own computer. Software, notably WebEx and Skype, permits you to communicate directly with others in academia so long as you have a camera and microphone attached to (or built into) your computer. With WebEx, for example, you can host a meeting or conference with an unlimited number of attendees and share any application or content on your desktop or on any of the attendees’ computers. To get started, see the Knowledge Base articles on WebEx FAQs and setting up a WebEx session or contact OIT’s Help Desk.

Portable Videoconferencing Systems

When you want to communicate with a group of people, but need to do so from your own office or classroom, it is also possible to bring in portable videoconferencing equipment.

The University maintains two portable Video Teleconferencing units that are best used in an office, a small classroom or conference room with an active network connection. The entire setup is usually brought into the room on a small cart outfitted with a computer monitor for the other site. Rental and labor charges apply, and can depend on the host room’s capabilities.

For additional details on costs and getting started with Princeton’s videoconferencing facilities, look at the Videoconferencing at Princeton web site. The site previews all five videoconferencing rooms, contains a list of frequently asked questions (with answers), and offers useful tips for making sure that all of your videoconferences proceed smoothly.

The last question in the FAQ is: “Why don’t more people do this?” The answer: “Good question. It saves on travel, hotels, and time. It can also bring a world of content to your classroom or office.”

Backing Up's Not Hard to Do

2009-05-28T20:38:33Z

You don’t have to be a fan of Sex in the City to know that it’s important to backup your data. If you have not seen this wonderful vignette, take a moment to see what can happen if your intellectual property is not well protected.

Princeton uses a software application called Tivoli Storage Manager or TSM to back up campus computers. During the current academic year, TSM has backed up 600,000,000,000,000 bytes of stored data on 10,849 campus client accounts using 16 STK/Sun T10000 encrypted tape drives in two silos as well as eight TSM servers in two computer rooms.

]]> The primary goal of this important TSM backup service is to protect the personal work of members of the University community from loss owing to hardware failure or inadvertent deletion.

There is a version of the TSM software for most of the computer operating systems on campus. You should take note of the fact that a new version of the TSM client software that has just been released (version 6.1.0.0) and the TSM web pages are being changed this month to reflect the new software and its installation.

You can obtain the software from the TSM web page. There, you will be able to obtain client software for: Windows XP, 2003, Vista; Mac OS X (for Macs); Mac OS X (for Intel Macs); Linux x86; and Solaris.

You are eligible to use the service if you have a Princeton netid and meet the other requirements listed at the TSM web site. There is a limited, or personal, account that is free of charge. A complete backup account is included with DeSC computers or with an account number.

The personal work documents stored on faculty, staff and graduate student workstations is backed up at no charge and with no limit on size. The definition of personal documents is based on the location of the documents (for example, the “My Documents” and “\usr” folders on a Windows NT or Windows 2000 systems).

Undergraduates are provided with 250 MB of shared network disk storage that protects their data with two different backup/restore capabilities. Self-service recovery of individual files that have been deleted or damaged will be provided by the shared storage server using a “snapshot” technology, which maintains copies of changed or deleted files in a special storage area on disk. TSM backup will be used to guard against catastrophic failure of the entire shared storage system.

For those campus systems that need backup service beyond the personal documents directories, a complete backup option is available for a fee of $5.00 per month. This complete backup service encompasses all directories on all disk volumes, with the exception of a few such as Temporary Internet Files and other similarly temporary directories. A $5.00 fee applies for any system whose total TSM occupancy is less than 150GB. Since data is compressed before it is stored on TSM’s tapes, the 150 GB limit represents a much larger occupancy on a user’s disks. Usage above the 150 GB limit will be charged at $.50/GB month in addition to the $5.00 charge for complete backup. Owing to their special standardized software, campus computers participating in DeSC ( Desktop Systems Council) will continue to receive their regular backups at no charge.

Some departments store their users’ personal work on a central server, and therefore do not backup individual workstations. To accommodate such departments, OIT has defined a special mechanism. If you are interested, contact OIT for further information.

What does TSM do and why does it matter?

In its most elemental function, TSM simply sends a copy of the files on your computer to a server for safekeeping. TSM compresses the data with a proprietary algorithm on your computer before forwarding it to the TSM servers. A separate TSM password for your data, different from anything else you use, also helps to ensure that your data is safe during the transmission process. Further protection comes from storing the data on servers that are kept in computer rooms with secure limited access and environmental controls.

Most backups occur automatically, using TSM’s scheduler. This is a service or daemon, depending on your operating system, that wakes up every twelve hours and communicates with the TSM server. Most desktop computers have a backup window that runs from 6:30 pm to 6:30 am, with the bulk of the backup initiations occurring in the first half of this window. This is a startup window only. Some backups may run into the morning hours after 6:30 am. Laptops are not usually left on overnight, so their backup window is during the day, typically over lunch.

In order to use TSM effectively, it’s necessary to understand how the backup process operates. TSM is a true incremental process; it doesn’t need to run a ‘full’ or base backup every month or week. If a file has been backed up once, TSM will not back it up again - it’s already on the server.

If you make frequent changes to a file, it is important to know that TSM will store the final version as well as the last three versions. If you need to restore an older version, TSM will not have it unless you gave the file a different name.

Know that all of the inactive versions will be removed from the TSM server’s storage (30 days by policy). The active versions will NEVER, EVER be deleted. In other words, if a computer is not backed up for 30 days or more, the backup in storage will contain all of the files that the client computer had on it when it was last backed up. The files will remain there indefinitely unless you email tsm@princeton.edu and have them removed.

It may be important for you to know that the data on machines that are now inactive (surplused or even destroyed) but were backed up by TSM can still be restored. Indeed, a backup program wouldn’t be of much good if it weren’t able to restore data from a computer that no longer existed. You can have that data removed by sending e-mail to tsm@princeton.edu.

How are files restored from the TSM backup service?

The key, of course, is backing up, because otherwise restoring files will be impossible. If you are experiencing a crisis, please call the OIT Help Desk (8-HELP) or e-mail tsm@princeton.edu. The actual procedure for restoring files is well documented in the OIT KnowledgeBase.

Computer Modeling of the Mind and Brain

2009-05-06T15:19:07Z

For those who still assume that the Psychology faculty analyze subconscious thoughts and place rats in mazes, Matthew Botvinick represents an eye opening cup of java. Building on the foundations of cognitive psychology, Botvinick’s laboratory works at the intersection of neuroscience, psychology and computer science, seeking to clarify the computational and neural foundations of human behavior. They employ a diverse set of research tools, including functional neuroimaging (fMRI), behavioral techniques (reaction time, error, and decision analyses), and computational modeling (neural networks, reinforcement learning models, and belief nets), typically applying multiple techniques to a single problem.

They are leveraging these tools to investigate a range of specific research questions, spanning the topics of cognitive control, working memory, decision making, sequential action, and language processing. Current projects include the monitoring and control of cognitive processing, the control of sequential behavior, and the representation of sequential order in working memory.

]]> At the May 6 Lunch ‘n Learn seminar, Botvinick placed his work within the context of developments within Psychology. At the onset of Psychology as a field, he began, one of the central developments was introspection, observations and reports on the contents of the unconscious mind. That work generated interesting findings and got the field rolling, but Psychology concluded that this early approach provided no meaningful mechanism to confirm or deny such subjective hypotheses. Simply put, there was no objective data to analyze.

Behaviorism emerged, providing data from observable actions and reactions, an attempt to apply the methods of the natural sciences to the study of human and animal behavior. B.F. Skinner and his contemporaries worked with animals, mainly rodents, placing them in boxes with levers, and rewarded them with various treats. It was an article of faith, says Botvinick, that behavioralists would talk about observables, the stimulus conditions and the resulting behavior, but not what was happening inside the subject animal. The key theoretical construct in behavoralism was the construct of the stimulus-response association. It was a view of the function of the mind and the nervous system that had to do with direct observation of the relationship between stimulus events and observable responses.

Voices of dissent, even in those early days, notably Edward Tolman, pointed to empirical phenomena that could not easily be reconciled without thinking about what was happening within the subject animal. A rodent in a maze, for example, appeared to maintain a cognitive representation of the maze, not a simple stimulus or observable response.

And then came the computer, spurring a cognitive revolution for Psychology says Botvinick. Here was a machine that contained internal representations, patterns of electrical activity within the machine representing information and which determined the machine’s responses.

Early on, the new field of cognitive psychology used the computer as a metaphor for the mind. It could store short and long term memory, it had control elements, a central processing unit. And within its software, it had the ability to achieve tasks algorithmically. Some have even held that the machine well mimics human decision making processes. Do computers and neurons equivalently process information; Can we write programs that mimic human response to stimulate? These questions spawned research into neural network models and several other approaches in computational neuroscience.

To give a sense of what researchers can derive from computational models, Botvinick presented one case study of computer modeling from his own work, a study of the seemingly mundane task of making a cup of coffee. Using the computer more as a tool than a metaphor, modern machines implement models of such behavior.

Looked at in detail, says Botvinick, the study of coffee-making provides significant insight into how people function, and why and how they occasionally fail to accomplish even simple tasks.

A podcast and the presentation are available.

The Cairo Geniza: Ancient Papers in the Digital Age

2009-04-29T20:45:51Z

The Cairo Geniza is a collection of an estimated 750,000 manuscript pages found discarded for “burial” in the Geniza chamber of the Ben Ezra Synagogue in Cairo in the late 19th century. In addition to holding religious poems and fragments of Torah scrolls, the Cairo Geniza contains approximately 15,000 mundane papers that reflect the daily life of the Jewish community in Cairo during the medieval period (mainly in the 11th to 13th centuries) - letters, contracts, wills, and other legal documents preserved in the area’s arid climate. These “Geniza documents” range in size from a few words to long letters of 80-100 lines.

For more than two decades, Mark Cohen and his Princeton colleagues have been working to bring these ancient papers into the digital age. Their work, called the Princeton Geniza Project, has created the world’s only online, searchable-text database of the Cairo Geniza’s historical documents.

]]> At the April 29 Lunch ‘n Learn seminar, Mark Cohen, Professor of Near Eastern Studies, discussed the background and challenges of the project. In 1986, Cohen and his colleague in the Near Eastern Studies department, Avrom Udovitch, proposed the creation of a computerized database of Geniza documents. IBM (through its Princeton Pegasus Project) and Princeton University’s Near Eastern Studies department supported the effort, and in the past 20 years, with help from technology upgrades and recent grants from the Friedberg Genizah Project and the University, the database has grown to include more than 4,000 documents (as much as a quarter of the historical Geniza), available online and searchable in Hebrew and Arabic script or English keywords. The database used for this purpose, called TextGarden, contains transcriptions of Judeo-Arabic, Hebrew, and Arabic documents in XML format and allows for the storage of not only the transcriptions themselves, but also of the images, genres, news stories, essays, locations, and people involved with these documents.

The project has transcribed documents from film copies, photocopies, draft texts typed by S. D. Goitein, and printed editions, creating a full text retrieval text-base of transcribed documents. The project has developed new tools such as dictionaries, semantic categories, and morphological aids to aid the study of Geniza texts. The project disseminates its materials freely through the web to the international community of scholars who have an interest in the life of the medieval Middle East, as well as to all with an interest in Judaica. Ultimately, the project hopes to link digitized images of manuscripts in the corpus, as libraries pursue the imaging of their collections.

The fragments from the Cairo Genizah are dispersed in more than two dozen libraries worldwide. The three most important collections are in Cambridge, St. Petersburg, Russia, and in New York. Princeton’s Near Eastern Studies Department has copies and microfilms of most of the historical fragments. The virtual environment of the Geniza provides provides obvious advantages. Scholars gain much easier access to the manuscripts and have a honed index. But one advantage has been especially exciting. Cohen and other scholars have been able to reunite fragments of the manuscripts, restoring their integrity for future academic research.

The current edition of the Geniza database, and the TextGarden web application that hosts it, was developed in 2005 by Rafael Alvarado, then Manager of Humanities Computing Research Applications at Princeton (now Director of Academic Technology Services at Dickinson College). It replaced and incorporated the original browser developed by Peter Batke in the late 1990s. Ben Johnston from Princeton’s Humanities Resource Center, who has maintained the TextGarden database since 2006, spoke about the TextGarden database. Use of Unicode on web pages permits the project to archive the documents and transcriptions on the same page, even when both Hebrew and Arabic appear on the same document. The TextGarden database permits scholars to search for words and phrases and to explore often complex interrelationships among the documents within the collection.

A podcast and the presentation are available.

Archives and Manuscripts: Library Finding Aids

2009-04-22T21:28:21Z

The Department of Rare Books and Special Collections at the Princeton University Library holds world class archival and manuscripts collections. The Mudd Manuscript Library, with more than 35,000 linear feet of storage, holds two major collections: The Princeton University Archives and “Public Policy Papers” which include very significant collections in the areas of foreign policy, economics and economic development, Civil Liberties, Law and Jurisprudence.

Finding aids, descriptive inventories created by archival repositories in order to provide access to collections, serve as the entry points for scholars and researchers to discover and explore these collections. In order to provide a standard structure for finding aids, the archival community developed an international XML metadata standard, Encoded Archival Description (EAD) in 1995. Comparable to AACR2 and MARC for bibliographic records, the content standard for Finding Aids has now been adopted by numerous institutions. EAD reflects the hierarchical nature of archival collections and provides a structure for describing the whole of a collection, as well as its components. And the standard supports flexible searching by collection, creator, biographies, title, call number, or topic.

]]> With more than 5,000 patrons a year, the manuscript library has sought to reduce its backlog in recognition that it is unreasonable and even unethical to “preserve” material without making it available to researchers. For archivists, providing access requires processing, physically organizing the materials usually contained in boxes and folders, and describing the contents in finding aids. At the April 22 Lunch ‘n Learn seminar, Daniel Santamaria introduced the Library’s EAD website.

A quick search brought up the Finding Aid for the papers of George Kennan ‘25, a noted diplomat and historian. The extensive finding aid contains an informative biography as well as a detailed description of the organization of the papers. Like the other finding aids, it places the Kennan papers within the context of the era in which the documents were created.

Santamaria emphasized that it can be very challenging to maintain and gain control over these collections and efficiently and to describe their contents succinctly. Collections can be as small as a single box or thousands of boxes (the largest collection at Mudd, the records of ACLU, contains several thousand boxes. More than 2,000 finding aids representing all University Archives, Public Policy Papers and nearly all Manuscripts Division collections are now available at Mudd to assist patrons. Latin American Ephemera and Engineering Library finding aids are also available. Moving forward, the library hopes to develop a more dynamic interface to the finding aids and to increase the use of finding aids to provide access to digital content.

Daniel Santamaria is Assistant University Archivist for Technical Services at the Seeley G. Mudd Manuscript Library at Princeton University. In his current position he oversees accessioning, processing, and descriptive practices. He has overseen the processing of several thousand linear feet of organizational records and personal papers since 2005. Dan has previously worked at the New York Public Library and both the Special Collections Library and the Bentley Historical Library at the University of Michigan. He holds an MSI from the University of Michigan’s School of Information and a BA in History from Wesleyan University. He is active in the Mid-Atlantic Regional Archives Conference and Society of American Archivists, and has presented and taught workshops at numerous meetings and conferences.

A podcast and the presentation are available.

Flash Forward: The Rise of Small Tech Gadgets

2009-04-15T16:56:00Z

A compelling technology, flash memory continues its march through the consumer electronics industry, yet again doubling quickly in capacity and dropping in price. In its wake, the wreckage of other, once proud, technologies and products — the floppy disk wiped from computers by the USB drive, the CD Audio disc humbled by portable flash players, and tape-based video cameras that now seem clunky compared to smaller flash cams. And next in the sights: computer hard drives giving way to faster and more rugged Solid-State devices.

Again this year, Doug Dixon of Manifest Technologies worked the January Consumer Electronics Show to scope out the new products. This year’s show saw even more examples of the impact of flash memory: rugged HD camcorders, replacement solid state storage devices, Wi-Fi integrated on SD memory cards, new formats promising 2 terabyte memory cards, and card slots everywhere, from mobile phones to HDTV displays. Dixon returned to Lunch ‘n Learn on April 15 to explore the developing trends in the rise of flash memory and to show off dozens of fun, new, high-tech gadgets. His web site, contains his Lunch ‘n Learn presentation, as well as more than 200 additional articles and a blog about a range of technology topics. The site also contains more detailed information about the products that Dixon demonstrated during his talk.

]]> In his talk, Dixon emphasized that flash-based Solid State Drives (SSDs) are beginning to encroach on the domain of hard disk drives (HDD): they weigh less, they are more rugged (with no mechanical parts or spinning platters), they are more power efficient (with 10 to 15 percent longer battery life), they can withstand more heat, vibration, and shock, and they offer significantly faster performance, so such systems will be able start up and launch applications twice as fast.

SSDs are moving from an extra-cost option (as in the Apple MacBook Air) to a cost-effective option for upgrading older systems. For example, rather than replacing an old laptop, you can swap in a solid state drive to make the systems feel young again, with a clearly visible performance boosts for disk-intensive operations.

While you can look forward to your next notebook using SSD storage, the industry has plenty of uses for SSD right now.

Dixon showed off and passed around iPods, iPhones, digital cameras, camcorders, portable computers, and portable game stations, but it’s the mobile phones that have truly captured the market. “There are billions of mobile phones out there,” says Dixon, and they are much more than just telephones. They play music and movies, they support a growing range of applications, and of course, they reach the web.

Ten years ago, he noted, we stored our documents on floppy drives. Our cameras held 36 photos on a roll of film. Tape players recorded music. Today we are seeing a remarkable increase in storage capacity that is outpacing even the predictions of Moore’s Law. Very tiny USB drives can store thousands of documents and photos and songs. In 2006, says Dixon, $14 bought a 128 MB SD memory card. Today, that same price buys 4 GB — five re-doublings in just three years, with an enormous impact on consumer electronics.

Douglas Dixon is an independent technology consultant, author, and speaker specializing in digital media. A graduate of Brown University, and previously a product manager and software developer at Intel and Sarnoff in Princeton, he is the author of four books and has published over 250 feature articles. Doug is currently editor-in-chief of Mediaware magazine and writes for Digital Photographer and Condé Nast Traveler magazine and the U.S. 1 Newspaper in Princeton. He has presented over 95 seminars and talks on digital media over the past eight years. Doug makes his articles and technical references freely available on his Manifest Technology blog and website.

A podcast is available.

All Aboard! Teaching Engineering with Computerized Toys

2009-04-01T14:52:57Z

Imagine being paid, or getting Princeton credit, for playing with trains and Legos™.

For more than 25 years, Professor Michael G. Littman, of Mechanical and Aerospace Engineering at Princeton, has taught MAE 412 Microprocessors for Measurement and Control, a course about microcomputer control.

In the class, students design single-board microcomputer controllers, and apply them for the automation of a modular n-scale model railroad. For example, a computer might be used to automate railroad switches to prevent collisions, facilitate traffic flow through a ladder network of tracks on a project board, or even regulate the loading of pipes onto train cars.

]]> In a three week portion of another course, EGR 194, freshman engineers program a LEGO™ Mindstorms controller to automate a LEGO™ vehicle to explore the environment. The LEGO™ Mindstorms controller is based on a single-chip PIC microprocessor. The student laboratory vehicle is a scaled-down version of the Mars Rover that has been exploring the surface of Mars for many years.

Littman believes that the use of the toys helps to engage students in important engineering concepts. Common to both courses stand-alone computers that are “embedded” as components within the toys. These components are akin to those contained within many engineering devices such as automobiles, aircraft, electronic instruments, and appliances.

In MAE 412, Littman makes use of a system that comes from England, the Hornsby Railway Company. Hornby, the English equivalent of our Lionel, has a digital train system that facilitates command and control of train set-ups. The system uses computers both in the master control unit as well as within the locomotives, to control both the speed and direction of up to six different engines. Each gains an identity, making it possible to talk to them individually. There are also “wayside” computers that are used for controlling switches or signal lights.

In the first demonstration, he sent a locomotive in motion, altered its speed remotely, and then sent a signal to switch the train to a different track. The train layouts support up to 99 accessories, permitting some very complex endeavors by the students, though Littman encourages them rather to design and execute simple tasks very well. He asks them to plan their systems, design them, build them, and then operate them. All designs and construction must meet specific standards, and all systems must comply with a set of rules.

In DC train layouts, you normally send voltage to a track, and the polarity determines the direction of the train. Trains facing each other would still go in the same direction, one forward, and one backward, at a speed determined by the voltage. By contrast, in the Hornsby system, since each locomotive has its own computer, you can command any car to move in either direction.

The tracks themselves form a broadcast network that carries both power and data. To change the speed, or to alter a signal, you change the data that the computer receives rather than altering the power to the tracks. Littman showed a live oscilloscope trace of the signal on the tracks. The repeating bursts of data each contain 32 bits, containing the speed and direction for four locomotives (four consecutive bursts control up to 16 locomotives), the state of a switch or signal, and a packet checksum to verify the integrity of the data.

Students learn to build a computer, they learning programming skills, and they get experience in planning, building, and operating a multidisciplinary engineering system. By the end of six weeks, the students will have designed, built, and tested a single board microcomputer that is able to receive instructions (data) from the track signals. The computer development is carefully orchestrated - they first build a minimal computer using only a clock, a display, an EPROM, and a processor. They then add additional new capabilities, one at a time the RAM, then Interrupts, then input/output registers, and then a programmable gate array.

During the final six weeks, they design, build, and implement circuits and mechanisms for sensing and control. Topics include how to use a transistor to control an electric motor; how to read a data sheet; how to select a transistor based on switching requirements and power dissipation; how to wire up project so that it is not plagued with electrical noise between circuits. And integrated projects typically employ optical, magnetic, mechanical, and thermal sensors.

He showed one student project that solved a variation of the Towers of Hanoi. The idea was to resort the cars of a train while avoided collisions with ongoing traffic along the line.

EGR 194 provides an introduction to the various disciplines of engineering and the relationship to the principles of physics and mathematics. For their projects, the class uses Lego Mindstorm systems that employ robotic remote sensing. Littman demonstrated one project, a robotic vehicle that navigated by itself over a twisting dark-colored path.

“Why toys?” Littman posed. “Because I really like toys, and because it breaks down barriers. Students are often afraid of complicated equipment. If it’s a toy, they’re not afraid to touch it”

Michael G. Littman is a Professor of MAE. He received his BA in Physics from Brandeis in 1972 and his PhD in Atomic Physics from MIT in 1977. In 2005, he received the Engineering Excellence Award from the Optical Society of America. His research interests include Laser Spectroscopy, Tunable Lasers, and Quantum control and Robotics.

A podcast and presentation from Dr. Littman’s Lunch ‘n Learn talk are available.

Women in Research Computing

2009-03-25T20:15:53Z

How did three distinguished women in research computing overcome political and societal obstacles? How have they dealt with the different work/life expectations that our society places on women? Do they see progress toward equaling the playing field?

On March 25, three prominent members of the faculty at the University joined moderator Betty Leydon, Vice President for Information Technology and CIO to discuss their use of Princeton’s high-performance computing facilities and a range of varied issues, from the challenges of performing research in a male-dominated field to the importance of mentorship.

]]> Emily Carter (Arthur W. Marks ‘19 Professor of Mechanical and Aerospace Engineering and Applied and Computational Mathematics) originally trained as a research chemist though she works now in Princeton’s School of Engineering. For the last decade or so, her applications focus has been to determine how materials fail due to chemical and mechanical effects and how to optimally protect these materials against failure. At present, she is turning her attention toward materials design for energy applications, including exploring novel materials for solar energy conversion to electricity and water splitting, metal alloy design for fusion reactor walls, and optimization of lightweight metal alloys to improve vehicle fuel efficiency.

In this research, she has a research group of 16 people, including 14 graduate students drawn from many different departments and backgrounds and two post-doctoral fellows looking for training in new areas to enhance and enrich their own research capabilities before they assume faculty positions of their own. She has also undertaken collaborative research with other groups.

Jennifer Rexford is a Professor in the Department of Computer Science with a specific interest in issues related to networking and the internet, how to manage and deploy networks and especially how to employ a network that was not designed to serve many modern needs. From 1996-2004, she was a member of the Network Management and Performance department at AT&T Labs—Research. At Princeton, she has become interested in how networks can be redesigned more efficiently, how to make them programmable, how to encourage users to share ideas and applications, and how to move from point to point effectively. Her research has focused upon routing protocols, how to communicate well while avoiding congestion and promoting good performance. In her work, she collaborates often with others in Computer Science and Electrical Engineering in relatively small groups.

Olga Troyanskaya is an Assistant Professor in the Department of Computer Science and the Lewis-Sigler Institute for Integrative Genomics at Princeton University, where she runs the Laboratory of Bioinformatics and Functional Genomics. The goal of her research is to bring the capabilities of computer science and statistics to the study of gene function and regulation in the biological networks through integrated analysis of biological data from diverse data sources. Most of the work in her lab involves computational experiments, using very large computational clusters to examine biological data. In the past, she noted, it might take seven years to perform an experiment and two years to analyze the data. Today, with high performance computing facilities, we can generate comparable data much more quickly, in months. There was a need, therefore, for computer scientists like Troyanskaya to develop statistical, search, and visualization algorithms that could analyze results and test predictions more efficiently. She notes that nearly all of the work in her area is collaborative, with biologists, systems programmers, experts on parallel computing and visualization, and with data modelers.

“Are you working with more women today than when you started?” asked Leydon. “And what kind of gender-balance changes are you experiencing?”

In pre-calculus in high school, replied Carter, she and one other woman were looked upon skeptically. It wasn’t much better at Berkeley, with two women out of 350 in an introductory Physics course. It’s now much better at both the undergraduate and graduate level. Now in her 21^st year as a faculty member, Carter notes that there were few female students when she began teaching. There were a few female graduate students in the first decade but today, she notes that her group is nearly half female. The number of female speakers at conferences remains depressing, she adds, and while it’s getting better, there are still times when she is the only one.

When she switched from hardware-related work to her current focus on networking issues, Rexford noted that the representation of women was markedly different. She was one of two women in her graduate program in the early 1990s because women had gravitated to more software-oriented disciplines. At conferences, she was one of two women. After her switch to networking, nearly 10% of conference attendees were now women. As an undergraduate in Electrical Engineering here at Princeton, there were only three women faculty in the entire School of Engineering. When she arrived on the faculty, she was one of five in her department alone, a profound transition in a relatively short period of time.

Troyanskaya graduated from high school in 1995, from college in 1999, and graduate school in 2003. In her 1995 Calculus class, her teacher commented that girls are as good as boys in math but not as good in computer science. The interdisciplinary nature of her field seems to attract more women than were the effort only in Computer Science. Still, the conferences involve no more than about 25% women, and fewer still if you count only speakers.

“What are the main obstacles?” asked Leydon.

All agreed that confidence is an enormous barrier for young girls in math and science. Said Troyanskaya, if you ask a question of a random sample of male and female graduate students, the men who are 40% confident of an answer, they will just say it, but women who are 90% certain will still express some uncertainty. It effects basic confidence, and even manifests itself in salary negotiations.

Work-life balance, added Carter, because without wives, it’s much more difficult to arrive at an appropriate balance between obligations at home and work. Home life entails a second shift that often requires that women leave the work place earlier than their male colleagues. And many male colleagues have wives willing to provide support not only at home, but as personal assistants. Generally, female CEOs are not married and don’t have children. Many women unfortunately decide not to pursue an academic career owing to the tremendous personal sacrifices that such decisions would entail. It’s often seemed necessary, adds Carter, to mask human qualities, to act with unemotional professionalism in order to counter stereotypic images of women in higher education.

Added Troyanskaya, men who are doing their work but having fun are seen simply as having a personality. Women who express themselves naturally are often not taken seriously.

Rexford added that women faculty members are also under extra pressure to join committees that add an additional burden. The University wants our voices to be heard, and we’re pleased to be asked, but that adds to the pressure.

Troyanskaya recalled that, on applying for an NSF grant, a male colleague once remarked that she was likely to get the grant because she’s a woman. The decisions are gender neutral, of course, but the perception complicates reality. Get the grant and it’s because you are a women, while being turned down would be perceived more harshly than normal.

“Has there been significant improvement at Princeton?” asked Leydon.

At Princeton, Carter states, we’ve become quite gender blind, from the perspective of hiring, promotion, and salaries.” Rexford noted that women tend not to negotiate as aggressively for salary and that can have a huge impact on relative wages over a career. Fortunately at Princeton, she adds, there’s an expectation that you don’t need to, and that makes for a much more comfortable environment.

Betty Leydon has been Vice President for Information Technology and CIO at Princeton since June 2001. After graduating from Bucknell in 1967, she joined IBM where she worked as a computer programmer and systems engineer for seven years. In 1986 after seven years in France, she joined the University of New Hampshire to oversee initiatives in computer aided instruction and two years later became the Executive Director for Computer and Information Services. From 1994-2001, she was Vice Provost for Information Technology and CIO at Duke University.

The podcast is available.

Princeton University's Broadcast Center: First Cuts

2009-03-12T20:31:24Z

The new Peter B. Lewis Library contains a new OIT-operated Broadcast Center with a high definition video studio that features a green room with a 65 inch LCD screen, a professional audio recording studio, as well as the hardware and software to edit video, color correct footage, and sweeten and edit audio. The Studio also has a Broadcast van with full, mobile production capabilities.

The Broadcast Studio staff is happy to assist members of the University community from the beginning through the end of their A/V projects, from the actual shoots through video editing and the final distribution. Some of the projects involve location shoots (from single camera shoots through full production), live event productions (such as Commencement and Opening Exercises), and in-studio shoots that aid in control of lighting and other key conditions. The Center also manages lecture recordings (including he integration of lecture A/V and speaker slides), podcasts, and rich media content.

]]>

Lance Herrington introduced a range of the early work produced at the Broadcast Center. Examples of Location shoots include a discrete recording with a roaming camera of PLOrk, the Princeton Laptop Orchestra, a narrated film without local audio about OIT Expressions, a yearly art exhibition, and a production involving factoids and interviews for the Human Resources Service Recognition Luncheon.

Live production events included the Provost’s interview with Ruth Bader Ginsburg in the University’s Public Lecture Series and the Freshman Opening Exercises.

An example of a studio shoot is an interview with Maria Garlock of the Department of Civil and Environmental Engineering. And Lance demonstrated a video editing project about the Iron Tiger Showdown, a cooking competition among the many residential college chefs.

Lance also showed the result of time-lapse photography in recording the construction of the Peter B. Lewis Library.

David Hopkins has been with the University for 17 years and has been working in video production for more than 20 years. He was formerly the manager of the New Media Center. David graduated from North Central University.

A graduate of the Rhode Island School of Design, Lance Herrington joined Princeton in 2001. He serves as a Senior Editor and Producer in the OIT Broadcast Center.

A podcast and the presentation are available.

The Foundations and Future of Information Search

2009-03-09T15:07:47Z

Launched in 1998, Google’s stated its mission: “to organize the world’s information and make it universally accessible and useful.” And so it is. Today, everyone Googles - in the U.S, about 12 billion times a month (including search engines that aren’t Google). We are mostly pleased with the results we get. How can it be that we give an automated system a couple of words and it finds reasonably relevant documents among one hundred billion or so possibilities? Will our satisfaction with these tools increase or decrease as the Web and our expectations grow?

At the March 4 Lunch ‘n Learn seminar, Computer Science Professor Andrea LaPaugh gave a peek “under the hood” of major search engines. Core techniques range from word occurrence analysis for text documents, which originating in the 1960s, to Web linking analysis, pioneered by Google’s 1998 PageRank document ranking method.

]]> It’s possible to score the relevance of documents by taking into account their length, by tabulating the frequency of words and phrases, and by giving greater weight to key words and phrases. The appearance of search terms in a section header, in bolded text or italics, within abstracts, or perhaps even within the URL itself can enhance the score in specific queries.

Before you initiate a search, retrieval systems have already recorded all information about web documents in an index that records a list, for each word, of all documents in which the word appears, the positions of the word in each document, and the various attributes that define the use of each occurrence. Indexes do not usually record pairs of words. Such indices would become prohibitively large. But by storing the location of each word, it becomes easy to determine if a searched for phrase, even a lengthy phrase, exists within a document.

PageRank is the algorithm that gave Google its killer performance. The technique uses the idea of random walks, essentially following links to and from web pages. As with scholarly literature, every link reference confers some legitimacy to the linked page. Every web page gets a stable PageRank score that reflects not only word frequency analysis, but also the number of links from and to the page and the probability of being at any given web page.

The CASS Project, within the Princeton Computer Science department, is one of many efforts to investigate retrieval of images based upon features within the images. Led by Kai Li, Moses Charikar, Perry Cook, Olga Troyanskaya, and Jennifer Rexford, the project permits you to click on a picture, for example, and generate similar pictures from its database.

LaPaugh predicted the emergence of a richer Web that could proffer topics without word clues, respond to natural language queries, and return information in ways that are more useful to us individually.

On February 22, 2009, the New York Times ran an article Exploring a “Deep Web” that Google Can’t Grasp. The point, she emphasize, is that much of the information on the web related to financial information, shopping sites, and even much academic research lies behind databases that must be directly queried rather than universally indexed. New technologies are emerging to extend the reach of search engines into these database collections.

The emerging web may use additional clues such as personal or aggregate behavior, essentially how users have interacted with pages, clustering and sampling methods, and individual link analysis. Clusty offers results based on clusters. A search on the word “bat” will cluster results around different topics such as baseball, mammals, images, and conservation. Another site, Cuil, provides useful summaries of sites.

Other research areas being pursued include machine learning to assign more appropriate catalog tags, network analysis of the web’s structure, semantic analysis of Web pages, sampling methods for image and video analysis, and architectures for large data collections.

Andrea LaPaugh is a Professor of Computer Science. Her research is in the development and evaluation of methods for searching and analyzing information. Much of Professor LaPaugh’s work is based on the principles of combinatorial algorithm design. In addition to her work in application areas, she has developed and analyzed algorithms for theoretical combinatorial problems such as graph structure problems. Professor LaPaugh has worked extensively in the development of algorithms for problems in digital design. A major area of research has been VLSI circuit layout: investigating the interactions between placement and detailed routing. She has developed algorithms that use these interactions to find better placements for circuit components.She is currently teaching COS 435: Information Retrieval, Discovery, and Delivery.

A podcast and the presentation are available.