Library Digital Scholarship Group and NULab receive $500,000 NEH grant

The Northeastern University Library’s Digital Scholarship Group and the NULab for Texts, Maps, and Networks received a $500,000 grant from the National Endowment for the Humanities as part of the NEH’s American Rescue Plan program.

The American Rescue Plan aims to provide funding to organizations conducting humanities projects that were adversely affected by the coronavirus pandemic. The grant awarded to the DSG and NULab is specifically focused on supporting humanities organizations.

This grant will help fund a series of digital projects currently underway through the DSG and NULab, but that were delayed or postponed due to the COVID-19 pandemic. It will support efforts to conduct collaborative research, digitize and process archival materials, create metadata, increase web accessibility, and more, while creating many graduate and undergraduate student research positions to conduct this work.

The projects that will benefit from this grant all involve collaborative engagement with communities outside of Northeastern, with many of them focused on resources related to underrepresented groups and social justice efforts. These include:

The grant also includes funding for additional projects organized through the NULab.

Julia Flanders, the director of the Digital Scholarship Group, is excited to get started: “We are honored and energized by this award. It creates wonderful research opportunities for students and will help the entire digital humanities ecology at Northeastern.”

A brief overview of machine learning practices for digital collections

Northeastern University Library’s procedure for digitizing physical materials utilizes a few different workflows for processing print documents, photographs, and analog audio and video recordings. Each step in the digitization workflow, from collection review to scanning to metadata description, is performed with thorough attention to detail, and it can take years to completely process a collection. For example, the approximately 1.6 million photographs in The Boston Globe Library collection held by the Northeastern University Archives and Special Collections may take several decades to complete!

What if some of these steps could be improved by using artificial intelligence technologies to complete portions of the work, freeing staff to focus more effort on the workflow elements that require human attention? Read on for a very brief overview of artificial intelligence and three potential options for processing The Boston Globe Library collection and other digital collections held by the Library.

A three-part cycle, with "Input" leading to "Model Learns and Predicts" leading to "Response" leading back to "Input"

What is artificial intelligence and machine learning?
Artificial intelligence (AI) is a broad term used for many different technologies that attempt to emulate human reasoning in some way. Machine learning (ML) is a subset of AI where a program is taught how to learn and reason on its own. The program learns by using an algorithm to process existing data and find patterns. Every pattern prediction is evaluated and scored according to how accurate the prediction may or may not be until the predictions reach an acceptable level of accuracy.

ML may be supervised or unsupervised, depending on the type of result needed. Supervised learning is when instructions are provided to assist the algorithm to learn how to identify patterns expected to the researcher. Unsupervised learning is when the algorithm is fed data and discovers its own patterns that may be unknown to the researcher.

Ethics
As we undertake this work, it is important to be aware that AI technologies are human-made and therefore human biases are embedded directly within the technology itself. Because AI technologies can be employed at such a large scale, the potential for negative impact caused by these biases is greater than with tools that require standard human effort. Although it is tempting to adopt and employ a useful technology as quickly as possible, this is an area of research where it is imperative that we make sure the work aligns with our institutional ethics and privacy practices before it is implemented.

What AI or ML techniques could be used to help process digital collections?
OCR: The most widely known and used form of AI in digital collections practices may be recognition of printed text using Optical Character Recognition, or OCR. OCR is the process of analyzing printed text and extracting the text objects, like letters, words, sentences. The results may be embedded directly in the file, like a PDF with OCR’d text, or stored separately, like in a METS-ALTO file, or both.

Screenshot of the front page of the Winchester News
Image source: Screenshot of an OCR page of The Winchester News with METS-ALTO encoding opened in AltoViewer.

OCR works rather well for modern text documents, especially those in English, but a particular challenge for OCR is historical documents. For more about this challenge, I recommend A Research Agenda for Historical and Multilingual OCR, a fairly recent report published by NULab.

A screenshot of a search result that reveals the result was returned because the search term matched OCR'd text within the document.

We can already see the benefit of using OCR in the library’s Digital Repository Service, as files with OCR text embedded in the file have the full text extracted and stored alongside the text file. That text is indexed and improves discoverability of text files by retrieving files that match search terms in the file’s metadata or the full text.


The back of a photograph from the Boston Globe Library Collection, featuring difficult-to-read handwritten descriptions.
Digitized back of a photograph from The Boston Globe Library collection.

HTR: Handwritten Text Recognition, or HTR, is like OCR, but for handwritten, not typewritten, text. Handwriting is very unique to an individual and poses a difficult challenge for teaching machines to interpret it. HTR relies heavily on having lots of data to train a model (in this case, lots of digitized images of handwriting), so even once a model is accurately trained on one set of handwriting, it may not be useful for accurately interpreting another set. Transkribus is a project attempting to navigate this challenge by creating training sets for batches of handwriting data. Researchers submit at least 100 transcribed images for a particular handwriting set to Transkribus and Transkribus uses that set as training data to create an HTR model to process the remaining corpus of handwritten text. HTR is appealing for the Boston Globe collection, as the backs of the photographs contain handwritten text describing the image, including the photographer name, date the photograph was taken, classification information, and perhaps a description or an address.

Computer Vision: Computer vision refers to AI technologies that allow machines to work with images and video, essentially training a machine to “see”. This type of AI is particularly challenging because it requires the machine to learn how to observe and analyze a picture and understand the content. Algorithms for computer vision are trained to identify patterns of different objects or people and attempt to accurately sort and identify the patterns. In a picture of the Northeastern campus, for example, a computer vision algorithm may be able to identify building objects or people objects or tree objects.

A black and white photograph of a man being arrested by two police officers next to an analysis of the photo's contents: Footwear (98%); Shoe (96%); Gesture (85%); Style (84%); Military Person (84%); Black-and-white (84%); Military Uniform (80%); Cap (80%); Hat (78%); Street Fashion (75%); Overcoat (75%)
Result of Google Cloud’s Vision API analysis for a black and white photograph.

When used in digital collections workflows, the output produced by computer vision tools will need to be evaluated for its usefulness and accuracy. In the above example, the terms returned to describe the image are technically present in the photo (the subjects are wearing shoes and hats and overcoats), but the terms do not adequately capture the spirit of the image (a person being detained at a demonstration).

There are a lot of ethical concerns about using computer vision, especially for recognizing faces and assigning emotions. If we were to employ this particular technology, it may be able to generate keywords or other descriptive metadata for the Boston Globe collection that may not be present on the back of an image, but we would need to be careful to make sure that the process does not embed problematic assessments into the description, like describing an image of a protest as a riot.

Computer vision is already being employed in some digital collection workflows. Carnegie Mellon University Libraries has developed an internal tool called CAMPI to help archivists enhance metadata. An archivist uses the software to tag selected images, then the program returns other images it identifies as visually similar, regardless of its box and folder, allowing the archivist to easily apply the same tags to those visually similar images without having to manually seek them out.

Many other aspects of AI and ML technologies will need to be researched and evaluated before they can be integrated into our digital collections workflows. We will need to evaluate tools and identify the skills that are needed to train staff to perform the work. We will also continue to watch leaders in this space as they dive deep into the world of artificial intelligence for library work.

Recommended resources:
Machine Learning + Libraries: A Report on the State of the Field / Ryan Cordell : https://blogs.loc.gov/thesignal/2020/07/machine-learning-libraries-a-report-on-the-state-of-the-field/
Digital Libraries, Intelligent Data Analytics, and Augmented Description / University Of Nebraska–Lincoln: https://digitalcommons.unl.edu/libraryscience/396/

Overcoming the Paywall: Radical empathy and making the Gay Community News accessible to all

When Jackson Davidow was looking for information on Boston’s gay community in the 1970s, he knew where to go.

“I’ve long been interested in the relationship between queer politics and queer art, particularly in Boston in the 1970s, a point at which the city was a crucial hub of gay discourse, activism, nightlife, and sex,” said Davidow, a postdoctoral fellow in the “Translating Race” Lab at the Center for the Humanities at Tufts University. Gay Community News “was grounded in the political, cultural, and social environments of Boston. For that reason, it is an invaluable resource for researchers who study gay and lesbian life and liberation in Boston and beyond.”

Scan of the January 12, 1974 issue of the Gay Community News. It includes the headlines: New Gay Bills; UNH Saga Continues; and Maine Gays Attacked
The January 12, 1974, issue of the Gay Community News, one of its first published.

Gay Community News (GCN) was started in 1973 by eight Bostonians seeking to create a community voice for gays and lesbians in the Boston area. Originally published as a 2-page mimeographed sheet, the newspaper grew to have a national and international audience by the late 1970s and became one of the longest-running and most progressive national newspapers in the gay community. It was a natural place to start to gather the information Davidow needed. Issues of the GCN and records from its parent organization, the Bromfield Street Educational Foundation were subsequently donated to the Northeastern University Archives and Special Collections (NUASC).  

While today’s researchers can contact many archives by email and receive scans of collections remotely, there was a time when physically visiting an Archives was only possible for those who lived in or could travel to the area. To provide more access to collections in the 1980s and 1990s, some Archives made arrangements to microfilm high use portions of their collections. In recent years those microfilms have been digitized and are offered via subscription to libraries — usually at a high cost — and then made available to the students and faculty affiliated with that university, a practice commonly described as “paywalling.”

The August 2-8, 1987, issue of the Gay Community News. Its front page is an image of protesters standing in front of the U.S. Capitol with the headline "DC-Active! Coming out center stage to march on Washington"
The August 2-8, 1987, of the Gay Community News.

Unfortunately, this means that the many of the volunteers who wrote and edited articles, turned the crank on the mimeograph machine, or paid to advertise a queer night at a local club no longer have access to the content they created. It’s a trend that Giordana Mecagni, Head of the NUASC, knows all too well. Troubled, she recently published “Tear Down This (Pay)wall!: Equality, Equity, and Liberation for Archivists” in the Journal of Critical Library and Information Studies. The piece describes the negative effect paywalled archives have on institutions, archives, and researchers, and focuses on the GCN.

“Having the Gay Community News behind a paywall results in uneven access, where affiliates of universities can access the resource but members of marginalized groups within the queer community may not,” Mecagni wrote.

“Paywalls restrict who has access to archival materials. Many scholars are independent and unattached to academic institutions, or attached to academic institutions that do not have the money to subscribe to special historical resources,” Davidow added.

The NUASC recently completed an effort to made the Gay Community News freely available to anyone by re-scanning the GCN with help from the Boston Public Library’s “Library for the Commonwealth” program. This program provides free scanning services to Massachusetts libraries who have unique materials they want to share widely  and freely. Now researchers, students, members of the LGBTQIA+ community, writers, and anyone else can browse through 26 years of the GCN to get a glimpse of the gay community in Boston and around the world.

Researchers like Davidow are thrilled.

“The digitization of GCN helps scholars and community members learn about and revisit these important histories,” he said. “During my research for my recent essay in The Baffler, ‘Against Our Vanishing,’ I talked with many people involved in GCN, and everyone was thrilled to learn that the full run is available online.”

The GCN is available to access digitally through the NUASC’s LGBTQIA+ History Collection.

Introducing the New Northeastern Commons

Commons Redesign
The library is happy to announce that the Northeastern Commons is relaunching with a new look. The Northeastern Commons is an online platform where Northeastern University students, faculty, staff, and the outside community can come together to share ideas, explore common interests, foster creativity, and expand interdisciplinary thinking.

Screenshot of the Northeastern Commons website

The redesign was led by Northeastern Commons Coordinator Meg McMahon, with help from Web Developer Jeanine Rodriguez and Digital Accessibility and User Experience Assistant Vanessa Lee. As a team, they re-built the platform with a user-first approach and a focus on digital accessibility of the platform.

Using a variety of methods, including stakeholder and user listening sessions, the team focused on how the current platform was functioning. They took the data gathered during those sessions and created an affinity diagram of user needs for the rebuild. From this network of user needs, they turned to considering system requirements for the platform. Rodriguez pitched the idea of using the BuddyBoss Platform as the codebase because of the overlap between user needs and the features of that specific WordPress plugin.

During the build, Lee conducted an accessibility audit of the BuddyBoss platform, including browser checking, screen reader testing, and mobile testing, which Rodriguez then used as a roadmap for changes to the initial codebase. McMahon worked on user testing and internal testing of the platform to ensure users would be able to use the platform easily. Any issues found during the testing were added to the list of changes to make to the codebase.

Currently, the Commons team is still working on accessibility updates to the platform and feature updates and will continue to do so as the work on the Commons continues.

Commons Features
The Northeastern Commons runs on profile and group-based networking. That means users will be able to post, share, and create from their own individual profiles and within groups, which are the primary method of collaboration on the Commons.

Users who set up a user profile can share their research interests, publications, projects, talks, and press. Adding this information to a Commons profile makes it easier for other users to find people with similar research interests, which can lead to greater collaboration between Commons users.

Group collaboration on the Commons is unique based on choice and the subsequent use of those features. Furthermore, there is privacy built into the group design. Visibility of the group depends on the privacy setting of the group: public, private, or hidden. Public groups can be joined or viewed by anyone, whether they are signed into the platform or not. Private groups can be seen on the platform, but members must request or be invited to join. Hidden groups are only visible to those invited to join. Every group regardless of privacy status has the same features, which are:

Feed
The feed for groups is the activity feed. Activity can be an update from the organizers of the group, a notification when someone joins the group, a document added to the document table, and any action a member does within the group. Members of the group can also comment on the activity, leading to greater collaboration within the feed.

The feed acts as a living record of the progress and conversation the group is having and is searchable by keyword, which leads to greater discoverability of previous conversations.

Members
The members tab is a list of all the members of the group. Users will be able to search for members here, message them, and request a Commons connection, which is like friending on the platform.

Documents
The documents tab is a place for the group to upload documents that are relevant to the whole group. The file structure uses folders to sort and separate out documents.

Discussions
The discussions tab is a place where group members can create discussion board topics and reply to others’ discussion board topics. These can be subscribed to for easy access through a user’s profile.

Send Messages
This tab can be used to send a message to all group members using private messaging. It can also be used to send a message to only a few group members the message creator selects.

Subgroups
This tab appears if the parent group has subgroups within it. Subgroups function the same way that a parent group does; it is just nested within the parent group and does not show up in the group search.

Zoom
This tab is used to keep a running list of Zoom meetings for the group. If the organizers of the group choose to have the meeting recorded in the cloud, the meeting itself is accessible within the group.

Calendar
The calendar is a tab where organizers can create a list of group events which can be viewable in many different calendar forms. This feature must be specifically requested for a group using the Northeastern Commons Consultation form.

Static Pages
This is where a group can request to have a static HTML page within their group tabs. Group organizers will be able to add whatever they want to that page and continually update it based on their needs. This feature must be specifically requested for a group using the Northeastern Commons Consultation form.

Next Step for the Commons
Going forward, the Northeastern Commons will continue to utilize user needs assessments to grow and build further functionalities, leaning on the collective knowledge and desires of current group organizers and users.

For more information on the Commons, visit northeasterncommons.org or contact Meg McMahon at m.mcmahon@northeastern.edu.

DaVinci Resolve: Learning the Interfaces

DaVinci Resolve is a very powerful open-source video editing program. Its strength lies in its segmented workflow, allowing the user to work on the project in stages from beginning to completion. The variety of different interfaces the program presents you with may be daunting and confusing at first, but it gives you different opportunities to change the interface to suit your needs. In this tutorial, I will show you the different interfaces for each “stage” of post-production; then I will demonstrate how to customize those given interfaces.

If you don’t already have DaVinci Resolve, you can download the program for free here.

The Basics
Media Menu

Screenshot of the Media Menu in DaVinci Resolve with File Explorer and Media Pool highlighted

The first stage of post-production is assembling your project files. In the file explorer, it is important to keep all of the files you plan to use in your project in the same directory. From that directory, you can import you files by right-clicking and selecting “Add Into Media Pool” or by dragging it into the media pool (the bottom panel). You can also drag your files from the Windows File Explorer to the Media Pool. In the middle-center of the interface is the preview panel, where you can preview a file before importing it. On the right side is an audio panel, which provides equalizer and waveform representations of audio levels, and the metadata panel, which includes embedded information about the file.

Cut Menu

Screenshot of the Cut Menu in DaVinci Resolve

The Cut Menu is an interesting addition to the editing process. While you are able to cut your clips in the next menu (the edit menu), the cut menu has a specific view that allows you to focus in on a specific spot you want to cut. This is set up like an old school film cutting machine where the cutting line is fixed in the center and the clip is moved from left to right. This view is best suited for trimming down your clips to the length you want them to be before moving onto more complicated edits.

Edit Menu

Screenshot of Edit Menu in DaVinci Resolve, with Media Pool and Timeline highlighted

The Edit Menu is the menu that is most similar to other video editing programs. The interface has multiple audio and video tracks, and more can be added by right-clicking. By default, the timeline is on the bottom, the timeline preview is on top, and the media library is to the left.

Fusion Menu

Screenshot of Fusion Menu in DaVinci Resolve with Nodes highlighted

Fusion is Resolve’s interpretation of compositing and effects in the post-production process. In many video editing suites, the composites and effects would be applied directly to the timeline. Here, it’s on its own menu with its own interface and workflow to apply effects. The effects are applied by creating a chart with lines that connect to the in and out video points of the clip. You can add bubbles, called Nodes, to the chain of effects that represent text, noise, and other image transformations. Since this process may be unfamiliar to many, the Recording Studios has a video tutorial that explains how to use Fusion in more detail.

Color Menu

Screenshot of the Color Menu in DaVinci Resolve with Color Settings and Nodes highlighted

Like the Fusion Menu, the Color Menu also has a chart-and-node-based interface for applying effects. In this menu, the nodes panel is in the right side of the window by default instead of the bottom of the window. The bottom panel on the Color Menu contains several effects related to color correction, including wheels that tweak the values of different light and dark parts of the image, and a center channel that can be changed between multiple different menus, including color curves, windows, and qualifiers. There is a video tutorial on how to use the Color Menu.

Fairlight Menu

Screenshot of the Fairlight Menu in DaVinci Resolve with Track Volumes, Timeline, and Mixers highlighted

The Fairlight Menu is an in-depth sound mixing interface. At the top of the interface is a row of equalizer bars that displays up to 39 audio tracks at once, as well as control room and loudness levels. The bottom half of the interface displays a timeline of all the audio tracks so they can be trimmed and edited, and the right side of the interface has mixers for the output audio.

Deliver Menu

Screenshot of Deliver Menu in DaVinci Resolve with Export Settings and Timeline highlighted

Deliver is the final stage of post-production, in which you select the settings that best optimize the project for export to a video file. It provides a timeline view to make any last minute changes to the project, as well as a view above that of all the clips that you have added to the timeline. You can click on any of these clips and it will take you to the part of the timeline where that clip is located. On the left side of the Deliver Menu is the export settings. You can choose from a number of different presets that fit commonly used website formats, or make a custom choice of the format, resolution, and directory. After selecting those, you add the project to the render queue on the right side of the window, then select “render all” to start exporting your projects.

Customizing Your Interfaces
While it may seem that the multiple interfaces offered by DaVinci Resolve offer little room to change what the interface looks like, there are many ways to customize the windows onscreen and make the view suit your needs. The easiest way to do this is to click on the “workspace” option on the top window bar, and hover over the option “show panel in workspace.” This shows a list of all the available windows in your view that you can turn on or off.

Screenshot of DaVinci Resolve
Screenshot of DaVinci Resolve

For example, in the media view, I may think I want more space for the timeline, and I am finished with dragging items from the media pool onto the timeline. I can uncheck Media Pool in the list of viewable workspace items, and that makes more room to use the timeline with.

You can also click and drag the margins of each window to scale its size relative to other windows. Unlike Adobe Premiere, Resolve does not offer the ability to pop out a portion of the view into its own window. For this reason, the user experience of DaVinci Resolve is greatly improved by using multiple monitors.

Conclusion
The biggest hurdle for new users of DaVinci Resolve to clear is the unfamiliar interfaces and understanding what the new interfaces are used for. Once you understand that each of the new interfaces reflect their own discrete stage of the editing process, you should get used to the workflow of DaVinci Resolve quickly. Hopefully, this tutorial is able to explain what each of the interface views allows you to do in building a cohesive video project.

Listen to the audio version of this tutorial here: