Skip to main content
SearchLoginLogin or Signup

Summer 2022: Plans for June and Beyond

Summer projects and collaborations

Published onJun 20, 2022
Summer 2022: Plans for June and Beyond

The following is a seasonal update on the progress we have made at Philanthrobotics. Our hope is that this document will serve as a more speculative piece of our work. We want to use this space to think about some of the reflections we have received from external groups and ideate where we go from here. Below are bright spots from our Spring work, future directions, and potential partnerships from the reflections on our current work. 

Where We Go from Here

As we build Philanthrobotics, we are compiling issues to address and discoveries to build upon. 

An early issue to address is data governance and transparency. Organizational leaders need a way to come together and discuss what they want from a grants and proposals commons with analytic capabilities. A recurring challenge is determining what a grantor can share with colleagues or fellow-travellers with similar goals, preparing proposal data for sharing, and understanding how shared data will be used. Data-sharing agreements and data repositories with clear governance helps us all meet “in the middle” instead of working in silos. 

On the flip side, one of the strengths of our work has been meeting individually with MIT Solve and Lever for Change data scientists. This is where a lot of ideation and analyses come alive, because we can talk to each organization about what analyses they are focused on at the moment across their entire universe of proposals and reviews. For instance, Lever for Change is building internal tools to help their program managers find past contests and see what gaps exist in the history of their projects. Solve is looking for metrics to inform their long-term impact, including theories of change and mission directives for new challenges. Combining communication here with training individuals at each organization to leverage the Philanthrobotics knowledge graph and visualization capabilities, will help advance all of our work. More of this is addressed in “A Path Towards Training a Data Network.”


Directions for Summer 2022

As we start to build out more of the grants and proposals knowledge graph, we also want to think of some of the theoretical work that can be done on the graph itself. Specifically, we are interested in exploring causality among grant solicitation outcomes and de-biasing the judging process of some of these contests. As we work with colleagues in Belgium to lay the foundations of this work this summer, we have other usability goals in mind for Philanthrobotics, outlined below, and listed in order of importance. 

  1. By August, have a solid 1-2 projects proposed to work on with Carmen Mazijn, a Ph.D. student in AI Fairness at Vrije Universiteit Brussel, during the Fall* 2022 semester, using Philanthrobotics Knowledge Graph.

  2. Improve the UI of the knowledge graph instance on Neo4j by sustainable a perpetually-running instance that can be rendered in D3 or some other front-end library with specific customizations for Philanthrobotics users.

  3. Expand the Grants Data Index, using data hosted on the web to build out collections of grants data in specific domains

  4. Drawing on lessons from Our World in Data, Build out 10-20 views of combined data from MIT Solve + Lever for Change + UN Sustainable Development Goals datasets, to inform grantmaking processes.

  5. Continue outreach to open-grants partners from the University of Florida and the Council of Library Information and Resources (colleagues on the Open Grants Advisory Group). 

  6. Implement data from OGrants environmental scans [link] into the Philanthrobotics knowledge graph

  7. Plan out internal tooling for a grants search engine with OTS colleagues at Lever for Change – likely an open-source Python library

  8. Begin to train colleagues on how to use the knowledge graph with specific Cypher language workshops (through I³)


Update from Open Grants Advisory meeting in May

On May 19-20, I had the privilege of being part of the first meeting of the Open Grants Advisory Meeting at the University of Florida. It brought together a diverse group of advisors who are thinking about making grants and proposals data public. From this meeting, we were able to brainstorm problems, solutions, interfacing communities, and technologies we can use to make a grants and proposals commons a reality. 

In addition, there was an initial environmental scan of existing spaces that are making grants and proposals data (in the US) open to the public. We hope to take this scan and combine data from its sources into our knowledge graph for use later this year.


Potential Partnerships Ahead

After the Open Grants Advisory Meeting, many groups came forward with potential partnership ideas, including:

  1. Working with the Laurie Taylor (Director of Library Technology and Digital Strategy at UF) to make a knowledge graph of research around Caribbean Studies grants and proposals.

  2. Working with Christa Willford (Council on Library Information and Resources) to share grant data which that organization hopes to make public from its archives.

  3. Leveraging both federal and state databases, in addition to working with colleagues from the National Endowment of the Humanities to get geographic-specific spending profiles of the federally-funded grants onto our knowledge graph. 


Training a Data Network

As we begin to take on more partners and have talked of forming more potential partnerships. I’d like to take a little bit of time to outline some of the important aspects of training our network of partners in the near future. More specifically, summarizing resources we can create at Philanthrobotics to make it easier to use, and inspire partners to use it in their grantmaking. 

  • Having workshops through Knowledge Futures and/or I³ around Neo4j graph databases and instances, + how to query our data in meaningful ways

    • Creating a shared repository of queries for community partners to use in the earliest phases until more polished frontend technologies are made. 

  • Requesting to work with students or data scientists at the organizations we form partnerships with to gather questions on what visualizations might be useful for impact analysis, post hoc challenge/solicitation analyses, and following up with these professionals on how to use Datawrapper and/or D3 to get desired visuals

  • Build on the following list of projects for the data network to work on in tandem. This includes non-machine learning work that can be done at the Philanthrobotics level as well as at the individual organization level (i.e., to benefit the commons wholly, or community partners, individually):

Project

Expertise

Knowledge graph visualizations and auto-querying for more usable frontend

Javascript and front end knowledge; some Neo4j and Cypher, AWS/Redis queue knowledge

API building for grants and proposals search capabilities 

NLP similarity scoring knowledge, REST API and backend capabilities

Knowledge graph building
(various views and structures)

Some theoretical knowledge of DAGs, databases and backend technologies


Comments
0
comment
No comments here
Why not start the discussion?