Announcing the 2016-2017 Magic Grants!

We are pleased to announce the winners of the 2016-2017 Magic Grants. In these projects you will find a number of fantastic interdisciplinary collaborations that are formed within as well as between our two universities — from drone technology and advanced computer vision aiding video editing; to virtual reality creating visceral experiences of racism; to new “imager as sensor” research opening environmental reporting on our world’s coral reefs. This was our most competitive year to date!

Here is a brief description of the seven winning projects — fuller writeups are available below.

1000 Cut Journey from Harlem to Soho – A collaboration bettween the Cogburn Research Group at Columbia University and the Virtual Human Interaction Lab at Stanford University to leverage advances in virtual reality technology to create visceral experiences of racism

Re(ef)source – Combining new imaging modalities, crowdsourced data collection and analysis that provides a model for participatory story telling and citizen journalism in environmental reporting

The Comprehensive Database of Investor Ownership and Governance – A team from the Business and Law schools at Columbia will form the first-ever public, free data on the institutional owners of America’s largest companies (asking whether and how they serve as setters of average Americans’ savings)  and create an institutional governance index summarizing information on the advisor’s management, customer characteristics, and regulatory, civil, and criminal fraud history

Formalinfree: Tackling food adulteration in Bangladesh – In the markets in Bangladesh, high demand and low transparency means that manufacturers can hasten ripening processes with carbides and sellers can preserve fish in formaldehyde, and have little trouble finding buyers for the adulterated food — this project will help quantify, monitor and relay information about food contamination in these markets

We Can: A Geography of New York Canners – Oral history, illustration, “analog data visualization,” and sensor driven tracking, creates an immersive web platform to narrate the world of canners, people making a living collecting cans and bottles on the street

RoughCut: Developing Novel Video Capture and Editing Tools for Journalists – A video editor must have access to good footage to build an initial video sequence, called a rough cut — our project, a hybrid of drone technology and machine learning, addresses the key challenges of making a rough cut, planning, capturing, and annotating footage, finding relevant video clips, and pairing audio and visual content.

GenderMeme: A “Diversity-Aware” News feed – A product that takes in a stream of articles, and picks gender-balanced content, from which it makes a magazine — imagine Flipboard, but with a newsfeed that has been designed for gender diversity

Congratulations to our winners! Please check back to see how these projects develop over their funding year. Longer descriptions of the projects are available below.

1000 Cut Journey from Harlem to Soho: Examining the Psychosocial and Physiological Impact of an Immersive Racism Experience in Virtual Reality

The Cogburn Research Group (CRG) at Columbia University and the Virtual Human Interaction Lab (VHIL) at Stanford University have joined in collaboration to leverage advances in virtual reality technology to create visceral experiences of racism. The project integrates the groundbreaking research of Jeremy Bailenson, professor in the Department of Communication at Stanford and leading expert in human interaction in virtual spaces. with the emerging, innovative scholarship of Courtney D. Cogburn, assistant professor in the Columbia School of Social Work whose work challenges conventional notions of racism and broadly examines the role of racism in producing racial disparities in health.

The project team also includes a New York based media team, drawing on the community development and participatory media expertise of Linda Raftree and Peter Lee who has an extensive background in media-based social interventions as well as an interdisciplinary team of graduate students studying social work, psychology, communications and transmedia art.

With this project the team hopes to shift narratives of racism that produce feelings of “patronizing, well-meaning pity” at best and dangerous, life-changing decisions at worst to deeper, emotional connections that support critical social perspective taking and potentially move the needle of racial justice.

Re(ef)source

Re(ef)source will distribute cameras designed to capture fluorescence and multispectral images to scuba divers in Florid in order to mine data from images of coral that can help us gauge the health of reefs in the area. This innovative idea combines new imaging modalities, crowd sourced data collection and analysis in a new way that will provide a model for participatory storytelling and citizen journalism in environmental reporting. Environmental reporting typically relies on scientific and government reports. The team will add a new dimension to climate change stories by empowering journalists to deploy their own sensors and tools to collect data directly from the environment, enabling a new type of environmental interview. Instead of relying on experts, they are turning the environment itself into a source.

Re(ef)source will focus on stories about coral reefs because they are the “canary in the coal mine” for determining the health of ocean based ecosystems, as well as coastal areas. Not only does a quarter of the ocean’s animals and plants depend on healthy reefs for food and shelter, but entire economies lean on the tourism and fishing industries reliant on thriving coral populations. In the Florida Keys alone, $2 to 3 billion come in every diving season because of the coral reefs just off the coast. But these reefs are in perpetual danger of bleaching events and damage from human intervention, as well as rising ocean temperatures. At the same time, current coral reef monitoring strategies tend to be slow, expensive and reactionary, targeting reefs with already noticeable damage. People’s livelihoods, hobbies, and even their health hang in the balance along with the life of the coral.

Comprehensive Database of Investor Ownership and Governance

Federal securities law aims to give investors in American public companies transparency into the operation and ownership of the firm through disclosure. Historically, regulators and investors have focused these transparency efforts upon a classic problem of corporate law: the separation of ownership of the firm from control of its operations, due to the potential conflict of interest related to the use of corporate resources by managers. Thus, over the past two decades, a great deal of work has emphasized information regarding governance of public corporations.

By contrast, little is known about the owners of public corporations. In most work on this subject, owners are assumed to be the retail investors described in Depression-era scholarship. However, over the last several decades, ownership of American corporations has grown increasingly concentrated in large institutions such as mutual funds, pension funds, and hedge funds. Almost nothing is known about how these entities are governed. Yet the governance of those institutions represents the next frontier of American corporate law, since they will cast the critical votes that determine the outcomes of corporate elections, shareholder proposals and takeover plans.

Only few scholars have examined this area because, in part, little data exist that would allow empirical examination of institutional investor behavior. Although significant investment institutions are required to disclose their ownership in public companies on a quarterly basis on the Securities and Exchange Commission (SEC) Form 13F, each institution’s Form includes hundreds of entries, and thousands of such forms are filed each quarter, making it impossible for an investor to ascertain institutional ownership in a particular firm. Moreover, Form 13F includes no corresponding information about the institution’s governance—that is, the identity and background of its managers, directors, customers and employees. That information is disclosed separately in Form ADV, but the SEC provides limited access to those Forms.

Under this Magic Grant, the team will create the first comprehensive dataset of institutional ownership of American public companies, and the governance characteristics of those institutions. Understanding these two components is critical for both scholars and investors in American companies. For the former, understanding the investor base of public companies is necessary to examine the law that governs the balance of power at U.S. public companies. For the latter, knowing whether one is investing alongside a mutual fund, pension fund or hedge fund could be important information even for a casual investor. No previous research, however, has attempted to collect or analyze that information.

Formalinfree: Tackling food adulteration in Bangladesh

We propose to build a tool to provide real‐time updates about the food quality in the markets in Bangladesh. Food adulteration in Bangladesh is a deliberate ‐‐ if not outright malicious ‐‐ move to contaminate food with toxic substances. It happens because there is always a shortage of food ‐‐ Bangladesh is roughly as large as the state of Illinois but has a population of 157 million people ‐‐ greater than that of Russia. Thanks to high demand and low transparency, manufacturers can hasten ripening processes with carbides and sellers can preserve fish in formaldehyde, and have little trouble finding buyers.

These contaminants can lead to long term health consequences ‐‐ for example, formaldehyde has been classified as a probable carcinogen by the EPA. Due to a relatively low literacy rate, the consequences of consuming adulterated food is poorly understood by the public. Sometimes it simply stems from a lack of awareness about what substances are even toxic not an uncommon scenario in Bangladesh where the education system leaves much to be desired. The situation is made even more complex by the fact that big chain grocery stores are not common there ‐ people do their shopping largely from sprawling kitchen markets, which are difficult to monitor.

Bangladesh has several laws to protect consumers from adulterated food, and some have drastic sentences like the death penalty. But these laws, though severe, are rarely effective in dissuading the practice because of lack of awareness. Bangladesh passed the Mobile Court Act in 2009, empowering magistrates to deliver verdicts on food adulteration outside of a traditional courtroom setting. Currently, mobile courts raid food manufacturers and distributors randomly, test the food sold for adulterants, and deliver immediate verdicts.

However, the news on raids is delivered very late. Television broadcast happens hours after, and the newspapers publish it the next day. This provides little help to the buyers ‐‐ kitchen markets in Bangladesh are informal gathering places where sellers setup shops at random and move from one location to the next on a daily basis. In order to avoid buying or eating adulterated food, consumers need to steer clear of the location precisely when a raid is in progress. This is where the project comes in: it is designed to quantify, monitor and relay information about this difficult situation.

We Can: A Geography of New York Canners

“I know that if I go this way, in the morning and walk that far, it’s 25 dollars. And I know that if I go down that street, I’ll make 50”. Pier A. Simmons has his own way to interpret Bushwick, the neighborhood where he lives and works. He is a canner, one of the hundreds of people in New York who make a living picking up cans and bottles on the streets. After walking for hours and redeeming his haul, he likes to sit for a while in a tiny park on McKibbin street. He says the park has a musicality. Indeed, Pier is a former jazz musician from the Bronx who began canning after falling into A depression and quitting his job in a music academy. “Canning saved my life. I’m a free man now. I can go where I want, I can stop when I want, and I make money,” is his chant.

Working in collaboration with Sure We Can, the Brooklyn-based redemption center and community of canners Simmons belongs to, we are going to develop an innovative way to experience urban geographies and memories. Stories of canners will unfold across metal maps of the city that will take shape through oral histories, drawn illustrations and data visualization. We aim to create engaging multimedia narratives within an interactive web-platform that explore a hidden urban layer.

Each personal itinerary will be mapped through a sensor put on a piece of trash collected by the canner. Building upon the experience of the MIT’s project Trash Track, we will unpack a segment of the journey of a discarded object that becomes meaningful for a human being, drawing its path through the urban landscape as well as its unknown future.

RoughCut: Developing Novel Video Capture and Editing Tools for Journalists

Producing video content has become one of the primary responsibilities of news organizations. Many traditional news organizations, such as The New York Times, are increasing the size of their video production teams to meet the world’s growing demand for audio visual media. But video production is a complex, slow task. The quality of the final edited video relies on having good footage, which is characterized by having adequate coverage, i.e., multiple camera angles, and relevant audio recordings that the editor can use to craft the narrative. One way of extending the number of shots we produce is through the use of drones. Drones enable the capture of additional shot types, such as aerial cutaways, that the editor can incorporate into the video sequence.

An editor must have access to good footage to build an initial video sequence, called a rough cut. This is quite labor intensive. Editing a single minute of a rough cut can take even a skilled editor around 90 minutes. If there are large amounts of footage, it can take an editor several days to organize the content. This leaves journalists little time to make the important artistic and stylistic decisions that contribute to expressive storytelling.

In our project we plan to develop a system to support the process of making a rough cut. Some of the primary challenges of generating a rough cut include planning, capturing, and annotating footage, finding relevant video clips, and pairing audio and visual content. We are taking initial steps toward building a system to address these challenges and make the process of producing a rough cut easier and more time efficient. In our project we will support video teams by encouraging the camera crew to capture new perspectives. Having multiple viewpoints enables editors to have more flexibility in how they construct the narrative. We plan to explore new ways to interact with drones as professional journalistic tools.

Once editors find the video segments of interest, they annotate the shots to make it easier to construct a video sequence. Labels editors add might include the speaker in the shot and the shot type, such as establishing, medium or close-up shots. Our system will use computer vision face detection techniques to identify the speaker and characterize the shot type automatically. We will also allow the editor to talk over the video to provide annotations. Having videos labeled will help reduce the amount of time it takes editors to find the raw footage they need to construct the narrative.

Through this project, we will make the process of capturing footage simpler and the process of editing videos easier and more efficient.

GenderMeme: A “Diversity-Aware” News feed

How much media do we read that is written from a male perspective? A lot. Do we wish that we could hear more diverse voices in the media we consume? Yes.  And can we do something about that? We believe so.

We propose to build GenderMeme,a product that takes in a stream of articles, and picks gender-balanced articles, from which it makes a magazine. (Imagine Flipboard, but with a newsfeed that has been designed for gender diversity). To do this, we will first perform a large-scale analysis of news stories, in order to understand it better, and we will also create a “thermometer” that summarizes the state of the media in real time.

In much of the world, men’s voices are predominantly heard in many aspects of public life and society. This extends to the media as well: it is gendered. For example, most articles that focus on politics and business are written about and by men, whereas other topics like health and beauty are predominantly covered by women. This is reflective of the structural inequalities in our system.

As a result of this partitioning of topics, consumers of media might be missing certain valid points of view and perspectives. The media we consume might also be subtly reinforcing our existing biases about gender. For example, when most subject experts on, say, science, quoted in newspapers are men, a newsreader might end up strongly associating science and technology with men alone.

Further, in the data age, the proliferation of information about user preferences has enabled the construction of personalized news feeds, which show a user articles that he/she is likely to find interesting. However, the personalization of news feeds can often lead to a “filter bubble”, where users are shown articles that align with their existing viewpoints, ultimately leading to a harmful lack of diversity in the news feeds exposed to a typical user.