Announcing the Economic Graph Research Program
March 15, 2017
Roughly two years ago, we launched an ambitious project called the Economic Graph Challenge (EGC). The idea was simple: There are approximately 3 billion people in the global workforce, and LinkedIn's vision is to create economic opportunity for every one of them. By mapping economic relationships, developing the world's first Economic Graph, we can make that vision a reality. While it is easy to explain the vision, realizing it is extremely difficult. It is hard to imagine that those of us at LinkedIn could do it alone, so we decided to reach out to all of the brilliant minds in the world who shared our passion and wanted to work together to solve one of the most difficult problems of our times.
Here's the challenge we posed to the world: Given the wealth of data that exists within LinkedIn, what research would you propose that has the potential to create greater economic opportunity?
We were humbled by the overwhelming response to our CFP and the quality of proposals we received. We had more than 200 submissions, and each had something interesting to say. We ended up working with 10 teams for about a year. These engagements produced some strong results, generated exciting lines of research and generated intriguing new ideas that can help us better understand the economy.
Encouraged by the success of the EGC, we have decided to launch the program again this year. However, based on our learnings from the first round of research partnerships, we are changing the nature of the engagement. We are no longer hosting a challenge to select proposals. Instead, we want to make this a research engagement with more focus on some broad areas of interest to researchers and ongoing assistance from specific teams at LinkedIn. We are renaming this effort the Economic Graph Research (EGR) program, to reflect the long-term partnerships that are the basis for this program.
There were several factors that motivated our decision. In the course of the EGC program, we have already established a process to engage with researchers. Hence there is no reason to host a one-time challenge with a firm deadline; instead, we want to engage with the broad community anytime they have a strong idea. We have also learned that for such a program to succeed, continuous interaction on a regular cadence between researchers and LinkedIn teams is essential. It is hard for researchers to make brisk progress without obtaining additional context and guidance on Economic Graph data and prior research on an ongoing basis. Such an interaction can only be provided with significant involvement from researchers at LinkedIn.
New research topics
Additionally, we want to focus our attention on areas of the Economic Graph that are of broad interest to researchers here at LinkedIn. Note that these areas are indeed very broad given the diversity of talent and teams at LinkedIn, so we do not anticipate losing much in terms of generality through this focus. For instance, every single proposal we received for EGC 2015 can be classified into one of these areas.
We break the EGR topics into three major areas with the following broad topics: analytics, economics, and artificial intelligence.
Analytics is the discovery, interpretation, and communication of actionable insights from big data. At LinkedIn, our mission is to drive understanding and impactful decision-making through rigorous use of data. Our analytics is deeply tied to core modules of our ecosystem, including product, marketing, and sales. We are looking for research proposals that leverage big data analytics and data science to understand relationships in the economic graph, preferably in the following areas:
- Role of connections and other relationships in shaping career success;
- Occupation skill set trends and predictions as well as how to get them;
- Estimating the talent supply and demand gap globally/by geo/country/industry;
- Relationship between the economy and professional mobility/talent migration.
LinkedIn aims to be the go-to source for economic research that creates opportunities for every potential member of the global workforce. Within the Economic Graph Research umbrella, we are interested in rigorously investigating economic and labor market phenomena.
The relative importance of credentials vs. skills in determining labor market outcomes.
We are interested in a variety of machine learning and predictive modeling problems. They are the fundamental building blocks to drive search, discovery, and recommendation across the LinkedIn ecosystem. We are looking for research proposals in this broad context, with preferences given to the following areas:
- Large-scale machine learning, including large-scale methods for massive graphs/networks, fast online computations of models, deep learning, knowledge graphs, Q&A systems, etc.
- Personalized machine learning, such as content recommendation for each member, job recommendations for each job seeker, member intent modeling, etc.
- Reinforcement learning, such as developing conversational interfaces, real-time exploration/exploitation, etc.
- Crowdsourcing for social networks, for social applications that are highly personalized (feed ranking, link recommendation, etc.).
- Causal inference, root-cause analysis and real-time problem triaging
- Mining semi-structured and unstructured data, for member information standardization, content understanding, recommendation, etc.
- Mining time-series and spatial data, for trend analysis, member life-cycle modeling, etc.
- Security, privacy and ethics in machine learning, such as detecting and removing fake accounts, preventing member data breach, diversity in workforce, human rights, etc.
Call for proposals now open
We are excited to roll-out this program and are looking forward to partnering with the best minds in the world to solve an important problem that can make a big difference to many lives across the globe. Starting today and running until June 1, interested researchers from universities, think tanks, and other non-profit entities can download the proposal template and submit their proposals for Economic Graph Research projects to the team at LinkedIn. After June 1, we will be making our selections for ongoing research partnerships.
Full details about the submission process, selection criteria, and how we protect member privacy can all be found on the new Economic Graph Research website. Please join us in our endeavor to make the world a better place.