Proposal Information

University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Project Title: Developing a Data Trust for Open Access Ebook Usage

Amount Requested: $1,200,000

Grant Start Date: January 1, 2020

Duration (in months): 24

Program: Scholarly Communications

Description of Proposed Work: This project will put into action the recommendations of the white paper “Exploring Open Access Ebook Usage,” published by the Book Industry Study Group in May 2019 with support from the Foundation, by building a pilot data trust for usage data on open access (OA) monographs. As an international cooperative managed by the community of stakeholders in scholarly communications and operating a secure data repository and member dashboards, this data trust will be designed to align with the priorities of authors and institutions while respecting emerging ethical norms in the use of metrics.

1 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Proposal Narrative

Cover Letter:

When invited to submit the final proposal, please upload a cover letter on the organization’s letterhead, signed by the principal investigator(s). A draft may be provided here for review by Foundation staff.

DRAFT BELOW

Dear Patricia:

On behalf of the project team, I am pleased to submit this proposal for a two-year project entitled “Developing a Pilot Data Trust for Open Access Ebook Usage.” This project is a collaboration led by Kevin Hawkins at the University of North Texas (UNT), with co-PIs Katherine Skinner (Educopia Institute), Lucy Montgomery and Cameron Neylon (Curtin University), Rebecca Welzenbach (University of Michigan), and Brian O’Leary (Book Industry Study Group).

With $1.2 million in funding from the Foundation, the project team will take into account the findings of our previously funded environmental scan, build on its recommendations, and leverage the strong relations established during our community consultation, collaborating with publishers, platforms, and standards bodies in the US and around the world in scholarly communications to create a secure pilot data repository and member dashboards.

The University of North Texas will adhere to the Foundation’s Grantmaking Policies published at https://mellon.org/grants/grantmaking-policies-and-guidelines/grantmaking-policies/ as of today’s date.

Sincerely,

Kevin S. Hawkins Assistant Dean for Scholarly Communication Interim Head of Library Research Support Services

2 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Endorsement Letter:

When invited to submit the final proposal, please upload a signed endorsement letter from the chief executive officer, vice chancellor, rector, or president of the organization if the head of the institution is not the principal investigator for the proposed grant. Please see the proposal guidelines for further details about the letter. A draft may be provided here for review by Foundation staff.

DRAFT BELOW

Dear Dr. Hswe:

On behalf of the University of North Texas, I am pleased to endorse this application for a two-year, $1.2 million grant project entitled “Developing a Pilot Data Trust for Open Access Ebook Usage,” to be led by Kevin Hawkins of the University of North Texas Libraries. As in his letter from August ___, 2019, he will lead a team of co-PIs from a range of institutions that will collaborate with publishers, platforms, and standards bodies in the US and around the world to create a secure data repository and member dashboards and lay the foundation for an international cooperative managed by the community of stakeholders in open-access monograph publishing.

Sincerely,

Neal J. Smatresk President

3 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Proposed Activities and Rationale:

In as many pages as you need, please explain in detail the project’s rationale and proposed activities. This document should address the questions and topics set forth in the Foundation’s proposal guidelines for the relevant program area. These guidelines can be found here.

Background The promise of open access (OA) to scholarship is that published work will be more often downloaded, used, and cited. Authors, publishers, and funders of works of scholarship have increasingly experimented with OA in the hopes that the work will be more often accessed, consulted, and cited. However, these stakeholders in scholarly communications need reliable and comprehensive information about usage in order to make such claims.

To take an example, Branislav Jakovljević, a faculty member in theater and performance studies at Stanford, published a monograph with the University of Michigan Press about the interplay of artistic, political, and economic performance in the former Yugoslavia from 1945 to 1991, arguing that they were in fact inseparable. Since it is a profoundly interdisciplinary work, Jakovljević was interested in ensuring that it would be discovered by readers outside of theater studies, and since it addresses the role of soft power in dictatorships, he wanted the work to reach audiences in various countries where English-language university-press monographs are rarely available. Therefore, he was happy to allow the University of Michigan Press to make the work OA through Knowledge Unlatched, though he also wanted to be able to understand how readers used the book and to make a case about the impact of the OA version of the work for his promotion dossier—especially in comparison to similar work—beyond quoting sales figures, book reviews, and book awards. At the same time, the University of Michigan Press wanted to reassure the author that he had chosen the right publisher and made a good decision in choosing OA.

How could both the author and publisher get the information they needed in relation to usage of this book? The University of Michigan Press has struggled to find a way to help its authors understand the impact of their publications in OA. Usage data—the numbers of downloads, views, searches, and online annotations, all broken down by time period and country of use—is available in many platforms that support OA monographs. However, ebooks are often made available in more than one online location, for example, via the publisher website, through digital content platforms such as JSTOR, and through dedicated OA monograph platforms such as the OAPEN Library. As a result, significant work is involved in aggregating data that relates to usage of an OA ebook from each of the websites hosting the content. Furthermore, individual platforms make OA books available in different ways—for example, as individual chapter downloads, full book PDF download, or in HTML on-screen read-only formats. As a result, usage data provided by different platforms must also be standardized in order to allow meaningful information to be distilled from raw usage numbers. This requires data-science expertise that most book publishers (which are generally smaller in size than journal publishers) and many other stakeholders in ebook publishing do not possess.

4 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 The problem goes beyond individual publishers working with individual authors. Even with standardization, the different stakeholders in OA monograph publishing, having only a small perspective on usage of ebooks based on their position in the ecosystem, need help understanding data relevant to them and how it relates to data from other stakeholders. Crucially, access to data must be carefully controlled to ensure that users have access only to what they are entitled to see, taking into account user confidentiality, international privacy legislation, and certain business sensitivities. A system for sharing data should be designed to ensure responsible use of ebook metrics, with knowledge of the potential, as well as the limitations, of usage data as a source of insight into the ways in which specialist scholarly books are being accessed, the communities using them, and the impact of the knowledge that they contain.

In sum, support for OA book publishing will not grow without investment in the collection and analysis of usage data. Little of the promise for increasing the reach of long-form digital scholarship through OA can be demonstrated without the availability of aggregated and standardized ebook usage data, agreement on what is important to analyze, and tools for reporting to authors and funders.

What is needed The project team believes that a positive step toward tackling this challenge is to create a data trust. Data trusts are an increasingly common feature of the information landscape, especially in cases where data from a range of different providers (who may have conflicting interests) needs to be shared. As documented in a recent report1 from the Open Data Institute, such structures facilitate data sharing while retaining trust among organizations, and researchers at the Digital Media Research Centre at the Queensland University of Technology have documented examples from industries, such as automobile manufacturing and pharmaceuticals, that have characteristics in common with the OA ebook ecosystem.2 Namely, they found industry stakeholders with an overarching question that there is mutual interest in answering, entities of various sizes with data to share, the need to cooperate across organizations with a range of value systems, and information requiring commercial confidentiality. In most cases a governance framework for sharing data is combined with a data repository where information can be ingested from data holders, normalized, and disseminated to data users. Indeed, the Open Data Institute defines a data trust as “a legal structure that provides independent stewardship of data.”3

The time has come for a data trust for OA ebook usage.

The University of Michigan was awarded a grant of $93,000 (G-1802-05418) from the Foundation in 2018 to engage with the challenges of sharing data among stakeholders in the publishing of OA ebooks and to socialize the idea of a data trust. This year-long scoping exercise was conducted in collaboration with the Book Industry Study Group (BISG), KU Research (now known as Collaborative Open Access Research and Development [COARD]), the Educopia Institute, and the University of North Texas (UNT). During the year, the research term wrote and

1 “Huge Appetite for Data Trusts, According to New ODI Research,” Open Data Institute, 2019, https://theodi.org/article/huge-appetite-for-data-trusts-according-to-new-odi-research/. 2 Nic Suzor and Joanne Gray, “A Data Trust for Industry Sharing”, http://dx.doi.org/10.17613/gxa6-mg85. 3 “Defining a ‘Data Trust’”, Open Data Institute, 2018, https://theodi.org/article/defining-a-data-trust/.

5 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 disseminated an environmental scan and draft proposal for developing a data trust mechanism, conducted an online survey and a series of interviews with eight individual experts to gather feedback, organized an international summit of relevant projects and individuals connected to OA ebook usage in New York in December 2018, and presented on the proposal at a range of meetings and conferences. The research team found widespread support for the idea of a data trust, though there was some disagreement on the best architecture for such a system. These findings are documented in a white paper,4 the final deliverable from the project, which was published by BISG on May 6, 2019. An updated environmental scan of existing efforts related to understanding usage of OA monographs is included in appendix 1. The white paper also recommends key “next steps” for realizing a data trust for OA ebooks:

● Define the governance and architecture for the data trust and articulate the resulting priorities for development of the technical architecture. ● Create a pilot service that implements the defined governance and architecture. ● Implement and extend relevant technologies for gathering and standardizing usage data across a base of stakeholders in the US. ● Develop personas and use cases that demonstrate who benefits from OA monograph usage data and how a data trust addresses the needs of different user groups. ● Build engagement across multiple markets for OA monograph publishing, which are not yet pooling their efforts. ● Better document the supply chain for OA monographs to identify pain points and understand where usage data currently gets lost

The proposed project The project proposed here will move the recommendations of the white paper into action by building a pilot data trust for OA ebook usage data. The pilot data trust will consist of a repository of standardized usage data to be compiled from multiple sources but which will be accessible only through user interfaces (“dashboards”) and an API, for which access will be strictly limited according to roles determined through a governance structure to be developed in consultation with the community of stakeholders. This governance structure, and a related business model to be developed during the course of the project, will form the basis for a future self-sustaining data trust.

The project team members will continue largely intact from the previous project, though all partners have agreed that the new project would be better led by Kevin Hawkins at the University of North Texas (UNT), who has the time and expertise to lead this continuation of the work. This project will be an international collaboration involving multiple stakeholders, including a group with demonstrated expertise in creating governance structures and tactics for community building (Educopia), a major industry organization supporting the book industry supply chain (BISG), and a technical team embedded in the OA book landscape and data trust research (the COARD research team at Curtin University). The University of Michigan will be

4 “BISG Releases White Paper on Open Access Ebook Usage,” Book Industry Study Group, 2019, https://bisg.org/news/449950/BISG-Releases-White-Paper-on-Open-Access-Ebook-Usage.htm.

6 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 represented on this project by Rebecca Welzenbach, an expert in the use of research metrics. (See the section below for a full description of the project team.)

A significant part of this grant project will be the development of a governance model for the data trust that ensures that the trust will be controlled by its member organizations, who will be stakeholders in the publication of OA ebooks, and will maintain operations that are transparent to current and prospective members.5 Building on the community consultation in the previous phase of this research, and with the help of consultants hired as part of this project, the membership and business models will be designed to be attractive to stakeholders of various types and sizes, both in the US and abroad, and to ensure compliance with regulatory regimes in the US, EU, and elsewhere for data privacy and ownership. Furthermore, as explained in the discussion of sustainability below, membership in the data trust will always be open to new members to ensure that even those hesitant to participate initially will be able to join later.

Governance models do not exist in a vacuum, but are in fact interdependent on technical decisions, including those that impact how shared technical infrastructure and shared content components are created, managed, stored, and accessed. The project team intends to establish deep, lasting connections among the data trust’s governance structure, access requirements, and technical systems. During this pilot phase, with funding from the Foundation, a governance model and user personas will be fully developed, and browser-based user “dashboards” will be developed to allow publishers and authors of OA monographs to explore and understand ebook usage data in appropriate ways. The project team will engage in user testing of these dashboards with a representative sample of stakeholder organizations (see “Collaboration” below).

Technical architecture The data trust as an organization will operate a central data repository that will implement the data model developed by the HIRMEOS project (see appendix 1), which encompasses various types of data but stores them separate to ensure that no false equivalencies are created. The repository data will be presented via dashboards that accesses the repository through the API according to granular permissions, by which only users receive access only to those data points that they are entitled to access. This architecture will be extensible, allowing access by third-party analysis tools in the future as desired by members of the data trust.

The repository data will be populated through a range of mechanisms. Principally, software “drivers”, which import data from a particular source and transform the data to conform to the HIRMEOS data model, will be used. This data will then be directly deposited into the repository, with access control implemented to record provenance and (where necessary and appropriate) ownership. Drivers will initially be hosted by the data trust but in the future could be managed separately by partners or third-party data providers.

5 These and other principles for the governance of infrastructure are drawn from Geoffrey Bilder, Jennifer Lin, and Cameron Neylon, “Principles for Open Scholarly Infrastructures,” 2015, https://doi.org/10.6084/m9.figshare.1314859.

7 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Figure 1. Proposed technical architecture design (left). The current project will build a centralized system including a repository and access control systems for usage data. The use of HIRMEOS-derived standards for data models and data exchange means that it will be possible in the future for third parties to provide data to the repository or use their own analysis tools. In the longer term, a commitment to open standards and open source means that this infrastructure can contribute to a wider ecosystem of standards compliant repositories for diverse communities (right).

Collaboration Building on strong relations established during the previous Mellon-funded scoping exercise, the project team will actively collaborate with a range of types of stakeholders around the world in OA ebook publishing as described below: ● University presses and commercial publishers: The project team will seek feedback from publishers both large (such as Springer Nature) and small (such as the University of North Texas Press) on the governance and business models and on the functionality of the dashboards. Many university presses interested in understanding usage of OA books are participating in Towards an Open Monograph Ecosystem (TOME), but university presses outside North America, including UCL Press (UK), Göttingen University Press (Germany) and Wits University Press (South Africa), have also pledged to provide feedback to the project team. ● Library publishers: The Library Publishing Directory 2019,6 published by the Library Publishing Coalition, counts 138 publishing programs based in academic libraries around the world. A promising point of engagement for this community is Next Generation Library Publishing, a grant-funded effort to develop a robust, sustainable open infrastructure for use by library publishers. The participants in this effort, like university presses and commercial publishers, will need to be served by the data trust’s governance and business models and functionality. ● Scholar-led publishers: The ScholarLed coalition of publishers and stakeholder groups has argued forcefully for removing commercial interests from scholarly communication. Its recent Community-led Open Publication Infrastructures for Monographs (COPIM) project has reinforced this group’s skepticism of centralization, such as envisioned in this project’s model for a data trust. The project team will engage this group and others

6 The annual Library Publishing Directory is available in various formats at https://librarypublishing.org/lp-directory/.

8 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 skeptical of conventional publishing to ensure that the project’s governance and business models meet their needs. ● Platforms: Book publishers of all types often distribute their content through third-party platforms, which in turn collect usage data and share it with client publishers. Project MUSE, ProQuest, JSTOR, and Fulcrum are especially interested in contributing usage data to the pilot. ● Others working with usage data: Open Book Publishers, Ubiquity Press, OpenEdition, and COARD are developing the OPERAS Metrics Suite based on one of the work packages of the HIRMEOS project. Having developed a data model for usage metrics and software (“drivers”) for importing usage data from various platforms, these organizations plan to co-develop these further with us through a form of shared governance to be determined. ● Standards bodies: The work of the data trust relies on publishing standards maintained by groups such as COUNTER, EDItEUR, and Crossref, and the project team expects to identify bottlenecks and inefficiencies in the aggregation and analysis of usage data caused by deficiencies in such standards. The project team will work with these organizations to propose improvements to the standards that will facilitate OA publishing and the collection and analysis of OA usage data.

The project team will actively monitor work related to usage data of OA ebooks from other community stakeholders as well and will keep all stakeholders informed of the work of the project through quarterly updates written and distributed by the community liaison.

In addition, representatives of key stakeholders in usage data will be appointed to an advisory board for the project. These project advisors will be selected by the project team to provide a range of expertise, and to represent different stakeholder perspectives, geographic locations, and academic disciplines. This group will meet virtually on a regular basis during the period of this grant to provide feedback on our work.

While the EU has funded the HIRMEOS project to develop a mechanism for sharing usage data for OA ebooks, the design principles applied within this project are suited to the politics of collaboration within Europe. Specifically the OPERAS, HIRMEOS, and COPIM projects strongly focus on a federated architecture in which individual actors manage their own data and shared systems are focused on only that data which they are willing to share openly. The project team believes that significant progress can also be made by creating an environment in which partial data sharing is possible. Doing so requires a focus on the intersection between data sharing, access control, and governance arrangements. This focus will address two issues. First, the project team will be able to engage more effectively the participation of for-profit platforms and aggregators (such as EBSCO and ProQuest), which hold most of the usage data of interest, and of commercial publishers that produce the majority of open access ebooks in the humanities (such as Springer Nature and Taylor & Francis). Second, the project will facilitate the adoption of the technical infrastructure for usage data by publishers such as university presses with limited technical capacities while still addressing the sensitivities of data sharing. The Foundation’s support for this project is critical as a way to develop a solution that is pragmatic for US

9 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 stakeholders, both commercial and non-commercial, and that can provide a foundation for an infrastructure that goes beyond the US.

Further detail is provided in the timeline below.

Project team The project will be led by Kevin Hawkins, assistant dean for scholarly communication and interim head of library research support services at the University of North Texas Libraries. With over fifteen years of work experience at the intersection of academic libraries and publishing, and experience as a PI on an NEH-funded Humanities Open Book Program grant, he will coordinate the work of the subcontracting institutions, convene meetings of the project team, and oversee the work of the project’s advisory board.

The project team includes co-PIs from four institutions: ● Katherine Skinner, PhD, executive director of the Educopia Institute, brings to the project team fifteen years of experience in building and sustaining communities in the cultural-heritage sector. She will facilitate development of a governance structure and legal architecture for the data trust, advising on the available pathways for creating and managing an international cooperative (an independent legal entity or activity of an established organization) that can engage in agreements for data ingest and extraction. Since Skinner has deep experience hiring and supervising facilitators for various communities in the cultural heritage sector, she will also supervise the work of a community liaison for this project (to be hired). ● Lucy Montgomery, PhD, associate professor and the program lead for Innovation in Knowledge Communication in the Centre for Culture and Technology (CCAT) at Curtin University, and Cameron Neylon, PhD, professor of research communications at CCAT, are members of Collaborative Open Access Research and Development (COARD). As part of the grant, Curtin University will hire a product manager and data scientist to leverage and extend open-source technologies developed by the HIRMEOS project to develop tools and a workflow for aggregating, cleaning, and normalizing usage data and to develop a pilot repository that implements the governance structure, including any necessary software development work (“data scientist” and “product manager” are internal job titles that cover a range of skills including a range of relevant development skills). The team at Curtin will also coordinate with the community liaison on the development of personas and conduct user-experience research with authors and with staff of publishers and aggregators in order to design dashboards used to access data in the repository; based on this work, the team at Curtin will build dashboards for these audiences that present data aggregated from various sources. Furthermore, to complement the software developed by the HIRMEOS project, Curtin will arrange for creation of additional “drivers”, through staff at the Curtin Institute for Computation, for pulling data from platforms. ● Rebecca Welzenbach serves as research impact librarian, an emerging and rapidly growing position in academic libraries, at the University of Michigan. Her role at U-M is to help researchers understand and communicate the influence and impact of their work. She will provide expertise on the use of research metrics by researchers and institutions,

10 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 ensuring that the development of the data trust addresses researchers’ desire for metrics to capture usage of their OA publications while respecting emerging ethical norms in use of metrics.7 She will also provide connections to a number of relevant ebook initiatives at the University of Michigan, including the hosted ACLS Humanities Ebook (ACLS HEB) initiative, which involves over 70 publisher rightsholders, and Lever Press, which brings together more than 50 liberal arts colleges to fund the production of OA ebooks. ● Brian O’Leary, executive director of the Book Industry Study Group, brings more than 30 years of experience in the publishing industry to this role in the US’s major trade group for book publishers. He will work with BISG’s sister organizations (such as EDItEUR, Book Industry Communication, and BookNet Canada) to generate interest in the data trust among a diverse group of stakeholders (different sizes, commercial and non-commercial, domestic US and international). He will hire a consultant to map the supply chain for OA ebooks and will host stakeholders at a community meeting to refine this work. He will also hire a consultant on business models, who will work closely with the Educopia’s legal counsel.

In addition, the project team will include the following positions to be filled: ● A community liaison, to be employed by the Educopia Institute, who will create and distribute quarterly updates on the project to stakeholders in OA ebook publishing and will attend appropriate conferences and other meetings throughout the two years of the project to spread the word about our pilot project, invite participation in the international cooperative to be formed, and work with potential members on signing membership agreements. ● A data scientist, to be employed by Curtin University, who will gather, integrate, and analyze data from a wide range of sources and will develop software code, rules, and other models for matching ebook identifiers across data sources. ● A product manager, to be employed by Curtin University, who will work with the data scientist on data aggregation and analysis and with partners to ensure that the data trust meets their needs.

See appendix 2 for CVs of all PIs and appendix 3 for draft job descriptions for positions to be filled during the grant period; see appendix 4 for letters of commitment from each collaborating institution.

Project timeline The anticipated timeline for the project is shown below. In addition to the activities listed here, the PI team will meet virtually on a regular basis to coordinate our work, and the community liaison will create and distribute quarterly updates on the project to stakeholders in OA ebook publishing. The community liaison will also invite participation in the international cooperative to be formed; and work with potential members on signing membership agreements.

7 The most prominent statements on appropriate use of research metrics is the San Francisco Declaration on Research Assessment (DORA, https://sfdora.org/) and the Leiden Manifesto for Research Metrics (http://www.leidenmanifesto.org/), though the Mellon-funded HuMetricsHSS initiative (see appendix 1) has focused on creating a framework for appropriate use of metrics in the humanities and social sciences.

11 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Q1 2020 ● PI team finalizes a vacancy announcement for a community liaison; Skinner circulates the vacancy announcement and hires. Community liaison begins work at Educopia Institute. ● PI team formulates RFPs from Educopia for legal counsel and from BISG to hire consultants to do business-model development and to map the supply chain for OA ebooks. ● Hawkins invites representatives of key bodies of various types of stakeholders, plus domain experts, to join the advisory board. Hawkins convenes first virtual meeting of the advisory board to review the work plan for the grant project and provide input. ● PI team finalizes a vacancy announcement for a data scientist and product manager. Montgomery and Neylon hire for these positions. Incumbents begin work at Curtin University. ● Team at Curtin: ○ Carries out preliminary audit of relevant, publicly available data sets of usage data ○ Performs technical review of HIRMEOS APIs ○ Begins conversations with Project MUSE, JSTOR, ProQuest, and Fulcrum, about access to data from these multi-publisher hosting platforms

Q2 2020 ● Hawkins and community liaison identify partners (authors, publishers, and aggregators) to invite to test dashboards beginning in Q4 of 2020. ● Legal counsel and business-model consultant begin developing recommendations for the data trust, collaborating with each other. ● Supply-chain consultant begins work. ● Product Manager and Data Scientist at Curtin will: ○ Identify data sources, their data types, and the ways to acquire them ○ Apply the HIRMEOS data model and expand as necessary to handle the multiple identifiers (ISBNs, DOIs, and repository-specific identifiers) used for the same content and for part-whole relationships ○ Audit existing drivers and plan any new development requirements for drivers ○ Collect sample data from platforms and other key partners. This sample will need to contain data of all types found in the metadata schema and contain enough records to express the range of values expected for each element. The key characteristic is therefore diversity, not volume. ○ Match related identifiers in the datasets ○ Begin development of first version of dashboards with sample data

12 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

● Meeting in Denton, Texas, of all project staff, plus business-model and legal consultants. See appendix 5 for draft agenda. ● Hawkins convenes virtual meeting of the advisory board to provide an update on our progress. ● Community liaison attends 42nd Annual Meeting of the Society for Scholarly Publishing (US), Open Repositories 2020 (South Africa), and the 2020 Annual Meeting of the Association of University Presses (US) to spread the word about the pilot project to key stakeholders and influencers.

Q3 2020 ● Legal counsel and business-model consultant produce report with recommendations. ● Supply-chain consultant produces maps that illustrate how OA ebooks and their usage data are managed. These will be shared with the community for feedback. ● Hawkins convenes virtual meeting of the advisory board to discuss the business-model recommendations. ● Team at Curtin begins collecting data feeds from platforms and runs data trust infrastructure in test mode. ● Community liaison attends the 2020 meeting of institutions participating in Toward an Open Monograph Ecosystem (US) and the Open Access Scholarly Publishers Association 2020 Annual Conference (Europe) to spread the word about the pilot project to key stakeholders and influencers.

Q4 2020 ● Team at Curtin releases test dashboards for community feedback. ● Hawkins convenes virtual meeting of the advisory board to provide an update on our progress and to preview the dashboards. ● Community liaison attends OpenCon 2020 LatAm (Latin America) and the 2020 Charleston Conference on issues in book and serial acquisition (US) to spread the word about the pilot project to key stakeholders and influencers.

Q1 of 2021 ● Community meeting in New York of stakeholders to discuss the work of the project to date, including the study of gaps in the supply chain for OA ebooks. A meeting of all project staff will be held adjacent to the community meeting. See appendix 5 for draft agendas for both events. ● Data Scientist and Product Manager at Curtin will: ○ Refine data structures in accordance with feedback from the advisory board and other members of the community ○ Test security models for the dashboards and implementation approaches ○ Design the authorization and data-access model for users of the data trust

13 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

● Hawkins convenes virtual meeting of the advisory board to provide an update on our progress. ● Community liaison attends the 135th Annual Meeting of the American Historical Association (US), Academic Publishing in Europe Nr. 16 (Europe), and the fourth University Press Redux Conference (UK) to spread the word about the pilot project to key stakeholders and influencers.

Q2 of 2021 ● Team at Curtin: ○ Implements authorization and data-access model ○ Releases updated dashboards for broader community testing ○ Conducts integration testing (full end-to-end testing of compatibility) of data workflow and dashboards components ○ Conducts testing of access and security for both user-facing and repository systems ○ Re-implements and develops new drivers as required ● Hawkins convenes virtual meeting of the advisory board to provide an update on our progress. ● Community liaison attends the 2021 Library Publishing Forum (US), the 43nd Annual Meeting of the Society for Scholarly Publishing (US), the 3rd conference of the Association of European University Presses (Europe), and the 2021 Annual Meeting of the Association of University Presses (Canada) to spread the word about the pilot project to key stakeholders and influencers.

Q3 of 2021 ● Product Manager at Curtin: ○ Produces documentation of all components of the full stack of the data trust ○ Conducts user testing with Curtin team and project partners of the documentation ● Data Scientist at Curtin ○ Makes an initial release of the full stack of the data trust ○ Implements infrastructure for ongoing system testing ● Hawkins convenes virtual meeting of the advisory board to provide an update on our progress. ● Community liaison attends the 2020 meeting of institutions participating in Toward an Open Monograph Ecosystem (US) to spread the word about the pilot project to key stakeholders and influencers.

Q4 of 2021 ● Team at Curtin: ○ Deploys the full stack of the pilot data trust into full production ○ Provides the documentation and source code that would enable an expert user to create their own instance of the

14 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

technical system to prevent lock-in (that is, ensure “forkability”)8 ● Hawkins convenes virtual meeting of the advisory board to provide an update on our progress. ● Community liaison attends FORCE2021 (likely in the US), the 2021 Charleston Conference on issues in book and serial acquisition (US), and the Fall 2021 Coalition of Networked Information Membership Meeting (US) to to spread the word about the pilot project to key stakeholders and influencers.

Subgrant management and payment UNT’s Office of Grants and Contracts Administration (OGCA) has written policies and procedures for funding and oversight of subgrantees that carry out a portion of grant-funded projects. Payment terms can be structured either as cost-reimbursable or as firm, fixed-price based on successful performance of work milestones. Once a grant is made, payment terms are negotiated in order to conform to terms and conditions of the granting agency, UNT, and the individual subgrantee institution. Subgranting procedures include a process for OGCA and the PI at UNT to monitor subaward activities, payments, and invoices to ensure that work is performed according to the terms of the grant.

Technology to be used and developed The project team will leverage existing capabilities developed through the HIRMEOS project and through the Curtin Open Knowledge Initiative (COKI) project to build a pilot system for usage data collection, management and analysis. The project team expects to devote considerable effort at the outset to questions of metadata quality and completeness and to distinguishing and connecting chapter and full book identifiers, developing systems to handle and manage identification and deduplication and the mapping of metadata sources. Details of architecture and design will align with the governance structure and legal architecture developed in the course of the project.

The HIRMEOS project has made available a set of open source “drivers”—software written in Python that retrieves usage data from a particular source, such as OAPEN or JSTOR—and a community-validated data model for usage data. The team at Curtin has already developed the capacity for managed data collection and processing at high velocity in the course of their work on the Curtin Open Knowledge Initiative (COKI) project. The COKI project, which currently manages more than six trillion data points, is implemented on the Google Cloud Platform (GCP), which provides high service availability and a low-cost implementation of open-source, or open source compatible (such as Apache Beam and HBase), technologies. This service stack provides stability and reliability without significant vendor lock-in, ensuring that any future migration away from GCP would be possible if necessary or desirable in the future. The custodians of the HIRMEOS outputs—Open Book Publishers, Ubiquity Press, OpenEdition, and COARD—have

8 This is one of the principles for the governance of infrastructure in Geoffrey Bilder, Jennifer Lin, and Cameron Neylon, “Principles for Open Scholarly Infrastructures,” 2015, https://doi.org/10.6084/m9.figshare.1314859.

15 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 also expressed an interest in developing shared community governance structures of the data model and software with the project team. Collaboration with them on shared governance is outside the scope of this proposal but will ensure that our further development of their work is contributed back to a community-owned resource.

Drivers developed by the HIRMEOS project, plus additional drivers to be developed by this project, will be implemented as cloud functions—and contributed in that form back to the HIRMEOS community—with data collection triggered by cloud events. Cloud events are a technical set of “hooks” which lead to actions. These can be triggered by user interface events (such as a “refresh” button in a dashboard) a scheduling system (data collection once a week), or other external triggers (deposit of fresh data by a partner). The project team expects to develop drivers for a small number of project partners that have their own platforms for providing access to content as well as for Crossref Event Data. Data will be collected by the drivers, normalized to the HIRMEOS data standard, and stored in Bigtable, a GCP hosted cloud database system. The cloud implementation can in principle be cloned to provide private data collection for specific community members as well as to enable the provision of multiple instances of a driver if this is desired (that is, forkability as described above). GCP also enables highly granular data-access capabilities, although this functionality is less portable to another platform.

The governance model developed as part of this project will affect which types of users will have access to which types of data, both through dashboards and through APIs.

Dashboards will be provided for project partners, and the project team will implement a semi-automated but customizable reporting system generating regular reports for partners. Depending on project partner requirements these might be based on the previously developed HIRMEOS dashboard (open-source code using the Shiny R package), a stand-alone dashboard platform developed in Python (also open-source and more scalable), or through Google Data Studio (which is highly scalable and flexible but not an open-source platform).

Expected outcomes and benefits This project will result in three major outcomes that will form the basis for a future self-sustaining data trust:

● An organized and committed community of stakeholder member organizations, both non-profit and commercial, who are willing to contribute to the data trust and make use of its products, with clear policies for governance, decision-making, and communication ● Technological infrastructure, released as open-source code with full documentation, to enable the process of securely aggregating, cleaning, normalizing, and communicating data about the usage of OA ebooks to a variety of audiences ● A business plan to sustain the maintenance and growth of the trust

In addition, the community liaison will publicize the work of the project through quarterly updates on the work of this project and outreach presentations at conferences and other meetings, and BISG will publish a report of the community meeting on the supply chain for OA ebooks.

16 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 Benefits of these three major outcomes for publishers, platforms and aggregators, and funders—three key types of stakeholders—are described below.

For publishers Publishers struggle to gather usage data from many different platforms and then consolidate it—often by hand—into time-consuming, bespoke reports that ultimately may not mean much to end users. In contrast, the proposed data trust will enable publishers to: ● Consolidate usage data from multiple platforms (or, better still, authorize those platforms to do this on their behalf) ● Trust that the data is managed—cleaned, normalized, and maintained securely—by data-science experts working for the data trust ● Produce readable, meaningful reports for other stakeholders (authors, funders, institutions) from one place ● Appropriately benchmark usage data against data of peer publishers also contributing to the trust, giving context and meaning to the numbers ● Share the technological and human costs of this work with other participating publishers

For platforms and aggregators Publishers increasingly request usage data from the platforms and aggregators that provide access to their books. But these platforms are often constrained in how quickly and robustly they can respond to this demand by internal feature request and development cycles. If they’re not able to make a clear business case for this feature, it may not be taken up. The proposed data trust will allow platforms to: ● Respond to publisher demand for usage data reporting using technology that is decoupled from any particular platform used to host content and that is designed especially for this purpose ● Gain access to aggregated data about usage of other platforms (market share) that they might otherwise not have access to ● Move the industry more generally toward a common understanding of what usage data is important to gather for OA ebooks and how best to structure its collection

For funders Funders want to know about the impact of their investment in OA monographs, but there are no consistent standards for tracking or reporting this data, nor are there known expectations for outcomes. The proposed data trust will enable funders to: ● Articulate clear expectations for their grant awardees to track and report usage data by aligning with standard practices to be established by the data trust ● Make use of the reports generated by the data trust to assess the outcomes of funded projects, and inform strategic decisions about the value of funding the publication of OA monographs ● Compare funded projects across different publishers in order to establish meaningful benchmarks for what kind of usage to expect

Sustainability plan

17 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 As previously explained, a major focus of this project is the development of a sustainable business model and governance structure to be used by a future self-sustaining data trust. The involvement of consultants and legal counsel is designed to ensure that the model chosen will allow for participation by stakeholders of various types of sizes, provide a cost-effective service for its members, and ensure control by its members. While participation in the pilot will be by invitation, and while some invited stakeholders may not feel comfortable participating in the pilot, we aim to establish a community-driven infrastructure that will be as transparent as possible to encourage future participation by others once it becomes established.

As described in the Project Details section, the software resulting from this project will be released as open source on GitHub. Long term archival will be provided through snapshots deposited to Zenodo. The presence on GitHub will ensure discoverability and accessibility of the code, and archiving on Zenodo, a repository operated by CERN, will demonstrate our adherence to our vision of community control of the data trust.

Additionally, the research nature of our initiative may qualify portions of the project for matching funds through the Texas Research Incentive Program (TRIP). UNT is one of seven Texas-based research universities that participate in TRIP. TRIP exists to increase funds dedicated to research and research-enhancing activities at public research universities throughout the state. If this proposal is funded, the grant total will be submitted for consideration of TRIP matching funds.

18 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Proposal Details

Collaborators

Please comment in this section on any organizations collaborating with the applicant organization that would be significantly involved in the conception, development, and/or execution of the proposed activities. Consultants and contractors are not considered collaborating organizations for the purposes of this section and should be listed in the section titled Consultants and Contractors.

Does the proposal involve collaborating institutions and individuals?

Yes.

Collaborators Details: If yes, please list all collaborating organizations; for each organization, briefly describe the nature of the collaboration. If a collaborating organization would be receiving grant funds, please list the amount in the Grant Budget and in the Budget Narrative section.

● The Educopia Institute will facilitate development of a governance structure and legal architecture for the data trust, advising on the available pathways for creating and managing an international cooperative (an independent legal entity or activity of an established organization) that can engage in agreements for data ingest and extraction. Educopia will also employ a community liaison (to be hired). ● Curtin University will hire a product manager and data scientist to leverage and extend open-source technologies developed by the HIRMEOS project to create a pilot repository that implements the governance structure. The team at Curtin will also coordinate with the community liaison on the development of personas and conduct user experience research in order to design dashboards used to access data in the repository; based on this, the team at Curtin will build at least three dashboards for different types of stakeholders that present data aggregated from various sources. Furthermore, to complement the software developed by the HIRMEOS project, Curtin will arrange for creation of additional “drivers”, through staff at the Curtin Institute for Computation, for pulling data from platforms. ● The University of Michigan will provide expertise on the use of research metrics by researchers and institutions, ensuring that the development of the data trust addresses researchers’ desire for metrics to capture usage of their OA publications while respecting emerging ethical norms (such as those articulated by DORA, the Leiden Manifesto, and HuMetricsHSS) in use of metrics. ● The Book Industry Study Group will work with BISG’s sister organizations (such as EDItEUR, Book Industry Communication, and BookNet Canada) to develop interest in the data trust among a diverse group of stakeholders (different sizes, commercial and non-commercial, domestic US and international). BISG will hire a consultant on business

19 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 models, who will work closely with the Educopia’s legal counsel. BISG will also hire a consultant to map the supply chain for OA ebooks and will publish a white paper on the findings of this work. Finally, BISG convene stakeholders at a community meeting mid-day through this grant-funded project.

Contractors and Consultants

The Foundation’s guidelines for grants involving consultants and/or contractors apply to proposals where a significant portion of the grant funds would be paid to a third party in exchange for services. Note that the Foundation does not ordinarily consider universities that partner with a grantee to be consultants or contractors for the purposes of these guidelines.

Does the proposal designate a significant portion of the grant funds to be paid to consultants, and/or subcontractors?

Yes.

Consultants, Subcontractors, and/or Vendors Details: If yes, please briefly describe the proposed grant activities for which a significant portion of the grant budget would be paid to a third party in exchange for services. Amounts allocated to consultants and contractors should be provided in the Grant Budget and Budget Narrative sections.

The Educopia Institute will hire legal counsel and BISG will hire a business-model consultant, who will collaborate with each other and the PI team in developing a governance structure for the data trust. In addition, BISG will hire a consultant to map the supply chain for OA ebooks. See the Budget Narrative below for details on the budget and selection of these consultants.

Diversity and Inclusion and Anti-Discrimination

The Foundation is committed to diversity and inclusion in its grantmaking programs and to ensuring that its funds are deployed in workplaces and educational environments that maintain and enforce policies committed to safety, dignity, ethical conduct, and freedom from discrimination. Please complete the questions below.

Diversity and Inclusion Please describe how your organization defines and approaches diversity and inclusion in relation to its mission and operations. We ask that you include one or more examples of challenges and successes the organization has experienced with respect to diversity and inclusion. We ask that you include one or more examples of challenges and successes the organization has experienced with respect to diversity and inclusion. (Note: if the proposed grant is intended specifically to address diversity and inclusion, this should be described in the Proposed Activities and Rationale document).

20 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 According to the 2018 Campus Inclusion Climate survey (https://edo.unt.edu/climatesurvey), UNT utilizes the following definitions: ● Diversity: The multiple identities around which people differ that make one individual or group different from another and impact one’s perceptions, experiences, and interactions. These include identities such as race, ethnicity, sexual orientation, socioeconomic status, gender and gender identity, religion, etc. ● Inclusion: Active, intention, and ongoing engagement with diversity; embracing and affirming differences and offering respect in words and actions (such as language) for all groups and people. The University’s approach to diversity and inclusion is driven by the UNT Division of Institutional Equity and itsOffice of Diversity and Inclusion (ODI). The ODI provides learning and development opportunities to students, faculty, and staff, and develops meaningful and effective internal and external partnerships. Its mission is to facilitate the development of a socially just and intentionally inclusive environment for the university community through cultural humility, inclusion, education, identity development, community engagement, and strategic partnerships. UNT is currently in the process of developing its new strategic plan. This new plan will take into account our emerging Minority-Serving and Hispanic-Serving Institution statuses. Information gathered from the aforementioned Campus Inclusion Climate survey will be referenced to guide the development of this new strategic plan. UNT will also create a strategic plan specifically catered to diversity and inclusion. Example of a challenge: Despite the ODI’s efforts, there are many systematic issues in higher education related to racial equity that present challenges at UNT and most other postsecondary institutions. Just one is the limited number of students of color pursuing and graduating from doctoral programs, which makes it difficult to diversify university faculty in a way that represents a diverse student body. This example has a variety of root causes, including a lack of representation among faculty, a lack of diversified recruiters and admissions officers operating with a multicultural lens, and narrow definitions of merit in the graduate admissions process—all of which were confirmed by responses to UNT’s recent Campus Inclusion Climate Survey. In the 2019–2020 academic year, the Division for Institutional Equity and Diversity will draw on the findings from the survey to develop a strategic plan that seeks to address these prominent issues. The results of the Campus Inclusion Climate survey and the Division’s strategic plan will be utilized to apply a multicultural lens to the University’s strategic planning process, and large components of the Division’s strategic plan will be integrated into the University’s strategic plan currently under development. Both strategic plans will include steps to address steps to address the lack of diversity in graduate-level admissions and enrollment; the strategic plans will also infuse best practices in inclusive, diverse hiring of faculty and staff. Moreover, two prominent faculty members are spearheading an initiative to develop a pipeline of graduate and doctoral students of color, with a long-term goal of funneling these students to faculty positions. They are

21 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 in the planning phase of this initiative and plan to seek grant funding in the near future to enact the project. Example of a success: One of UNT’s most successful examples of its commitment to diversity and inclusion is its annual Equity and Diversity Conference, which is now in its 20th year. Hosted by the Division of Institutional Equity and Diversity, the conference marks a day of professional development, networking, and coalition building. It includes keynote addresses and workshops from thought leaders around topics related to social justice, identity, equity, and inclusion. This conference is funded through strategic sponsorships, with more and more companies seeking to sponsor the event each year. Past guest speakers have included national news correspondent Soledad O’Brien, singer/songwriter John Legend, journalist Don Lemon, actress Vivica A. Fox, and speaker/author Dr. Melissa Harris-Perry.

Anti-Discrimination: Please briefly describe your organization’s equal opportunity, anti-discrimination, and/or anti-harassment policies.

UNT prohibits discrimination and harassment because of race, color, national origin, religion, sex, sexual orientation, gender identity, gender expression, age, disability, genetic information, veteran status, or any other characteristic protected under applicable federal or state law in its application and admission processes; educational programs and activities; employment policies, procedures, and processes; and university facilities. The University takes active measures to prevent such conduct and investigates and takes remedial action when appropriate.

Financial Health

Has your organization experienced any financial difficulties and/or deficits in the last three fiscal years?

No.

If yes, please address any financial difficulties and/or deficits your organization has experienced in the last three fiscal years.

Leadership Changes

Has your organization recently experienced, or does it anticipate, any leadership changes and/or significant staff turnover?

Yes.

If yes, please describe any recent or anticipated leadership changes and/or significant staff turnover relevant to the proposal grant activities.

22 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

UNT’s dean of libraries joined in April 2018 but has retained a senior leadership team with long tenure in the UNT Libraries.

Intellectual Property

Do proposed grant activities include the digitization of works or the creation of digital technology and/or digital products, such as software, databases, audio or video recordings, podcasts, and websites?

Yes.

If yes, please provide a detailed account of the intellectual property to be created, any rights or permissions that your organization would need to secure, the means by which the technologies and/or content would be distributed, including the type of license that your institution would issue to users, and how your organization would ensure the long-term sustainability of any digital or software products.

In accordance with the Foundation’s Intellectual Property Policy, all software created with project funding will be released as open source (under the Apache License, Version 2.0) and made available on Github. Citable DOIs and long-term archival will be provided through snapshots deposited to Zenodo. This will ensure both preservation and availability with minimal costs, providing for the long-term availability of the software. Furthermore, the project will actively work to identify which data contributed by members of the data trust can be shared with other members.

Investment Income

Please describe how grant funds would be managed, including the overall investment strategy and asset allocation, and how income would be calculated and allocated to the grant. If the organization cannot by law invest grant funds in interest- or income-generating instruments, please explain why.

Grant funds, if awarded, would be managed according to the University of North Texas System’s investment policy “Regulation of the University of North Texas System, 08.2000 Investment of System Funds”. Further, the grant funds would be invested in the University of North Texas’ Short-Term Pool in order to provide sufficient liquidity and safety of principal. The Short-Term Pool invests in assets that provide daily liquidity and are considered U.S. cash or cash-equivalent investments or vehicles. All individual securities must have a credit quality of AAA at the time of investment, and the Short-Term Pool must maintain a weighted-average maturity of less than 90 days. The grant funds would receive investment income allocation on a quarterly basis proportional to the Short-Term Pool balances.

23 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Strategies for Successful Grant Execution

Are there any factors that could potentially impede the timely and successful execution of grant activities and goals? (Consider, for example, the achievability of project timeline, availability of qualified personnel, ability to secure agreements, and capacity to obtain any remaining project funding).

Yes.

Please outline these factors and briefly explain possible strategies for mitigating their effects on grant activities.

Risk Likelihood Severity Mitigation

Project staff cannot be hired Medium Medium Curtin is running a recruitment according to the project for a similar job description for a timeline. separate funded project that will allow us to advertise early and build a pool of applicants. Curtin already has an excellent candidate for the second position available.

A governance model cannot be Low High Educopia has substantial agreed upon according to the experience with facilitating project timeline. community-consensus exercises, and this project will draw on that experience to maximise the chances of success. If the project team and advisory board cannot reach consensus, then a smaller group will need to be formed to agree on a model in time for use by a future self-sustaining data trust.

A business model cannot be Low High Educopia has substantial agreed upon according to the experience with facilitating project timeline. community-consensus exercises, and this project will draw on that experience to maximise the chances of success. If the project team and advisory board cannot reach consensus, then a smaller

24 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

group will need to be formed to agree on a model in time for use by a future self-sustaining data trust.

Some platforms are not willing Medium High Plan to allow for individual to share data. publishers to upload the data they receive from platforms if necessary.

Turnover of technical staff Medium Medium Curtin University will host the impedes development timeline. technical part of the project through the Curtin Institute for Computation. The CIC operates a pool of technical staff which the project can draw on.

Changing norms and legislation Low High The project focuses only on data around Internet privacy make which is not personally usage data sharing difficult. identifiable. The community liaison and PI team will monitor legislative and regulatory change that could affect the data trust.

Grant Payment to Third Party

If the proposed grant is approved, will you request that the Foundation make payment to another organization (e.g., a university foundation) to administer grant funds on your organization’s behalf?

No.

If yes, please provide the name of the organization that will administer the grant funds and a brief description of the relationship with your organization

Matching Requirements

Would the grant be subject to a matching requirement?

No.

25 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

If yes, please provide a description of your organization’s strategies and timetable for meeting the matching requirement, including details of the prospective donor base and project uses for the matching funds.

Please note: only newly received gifts of cash or securities - in hand valued at the time of transfer, and designated for the Foundation-supported project - can qualify as matching contributions. Pledges do not qualify.

26 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Budget Narrative

Budget Notes by Category

Please provide category-specific notes in the fields below, as appropriate.

Salaries and Wages – Fellowships and Residencies: Total salaries, excluding benefits, for all fellowships, residencies, and scholarships. The budget narrative should list the base salary or stipend amount for each fellow or resident, the number of individuals in each position, and, if applicable, the FTE percentage allocation.

Total Amount Budgeted: $0

Notes:

Salaries and Wages – Program: Total salaries, excluding benefits, for all programmatic employees, which may include salaries, stipends, and/or course releases for the principal investigator, project director, artistic staff, and/or faculty. The budget narrative should list the base salary, stipend, and/or course release amount for each position, the number of individuals in each position, and, if applicable, the FTE percentage allocation.

Total Amount Budgeted: $38,882

Notes: The grant project will fund 20% of PI Kevin Hawkins’s effort in years 1 and 2. His base salary is $95,769.35, and the budget includes an estimated 3% increase for year 2.

Salaries and Wages – Administration: Total salaries, excluding benefits, for all administrative employees involved in the grant-funded project or activities, which may include salaries for executive, finance, communications, and/or clerical staff. The budget narrative should list the base salary for each position, the number of individuals in each position, and, if applicable, the FTE percentage allocation.

Total Amount Budgeted: $0

Salaries and Wages – Other: Total salary, excluding benefits, for any other staff not included in the above salary categories. The budget narrative should list the base salary for each position, the number of individuals in each position, and, if applicable, the FTE percentage allocation.

Total Amount Budgeted: $0

27 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Notes:

Benefits: Total benefit costs related to personnel involved with the proposed grant activities, including any pension contributions, health insurance, and other benefits. The budget narrative should describe how benefit costs are calculated.

Total Amount Budgeted: $8,895

Notes: Fringe benefits for the effort of Kevin Hawkins have been calculated according to the formulas used by UNT’s Office of Grants and Contracts Administration based on the benefits choices made by Hawkins.

Travel: Total travel costs directly related to the major activities of the proposed grant, including any expenses for transportation, meals, lodging, mileage reimbursements, and per diem payments. The budget narrative should provide a breakdown of the costs by activity and include the number of individuals participating and travel days. Please include the sources for the travel cost estimates.

Total Amount Budgeted: $1,087

Notes: Kevin Hawkins will travel to New York in year 2 to meet in person with the project team and attend the community meeting on the supply chain for OA ebooks:

Lodging (2 nights @ $300 per night) $600

Per diem for meals and incidental expenses (3 $57 on day 1, $76 on day 2, $57 on day 3 days of travel, based on US federal rates for 2020)

Flight from Dallas to New York (current cost $297 of a flight booked four weeks in advance)

Total $1,087

Travel expenses for other grant personnel are included in subcontracts to those organizations.

Conferences and Meetings – Non-travel: Total non-travel costs associated with conferences, meetings, and workshops, including any registration fees. The budget narrative should provide a breakdown of the costs by activity. Please note: some program areas do not support catering and

28 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 facilities charges. Please refer to specific program guidelines on the Foundation’s website for any further conference and meeting requirements.

Total Amount Budgeted: $0

Notes:

Equipment and Supplies: Total equipment and supplies costs, including any anticipated shipping, licensing, maintenance, and insurance costs. The budget narrative should include an itemized list of any equipment and supplies to be purchased or leased along with the cost assumptions.

Total Amount Budgeted: $0

Notes:

Consultants and Professional Services (including Artist Fees): Total consultant and professional services costs (including artist fees) not included in other budget categories. The budget narrative should describe any existing and/or anticipated agreements with third parties, including consultants, artists, and/or other vendors. The narrative should include a description of the work to be performed, whether the amount is based on a fixed price, hourly, or per diem rate, and whether it includes any anticipated travel or other additional expenses. The Foundation’s Guidelines for Grants Involving Consultants and/or Subcontractors can be found here.

Total Amount Budgeted: $0

Subgrants/Regrants: If applicable, the total funds allocated for use for subgrants/regrants. For subgrants, the budget narrative should include the names of subgrantees, a description of the work to be performed by and the amount of allocated funds for each subgrantee. For regrants, the budget narrative should include the total funds to be regranted and describe the amounts expected to be awarded to each organization and/or individual.

Total Amount Budgeted: $1,149,031

Notes: UNT will have four subcontracts:

1) Educopia Institute Activity Year 1 Year 2

15% effort for co-PI $20,000 $20,560 Katherine Skinner

29 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

100% effort for community $100,000 $102,800 liaison to be hired

Legal counsel (52 hours in $26,000 $26,000 each of years 1 and 2 @ $500 per hour, the going rate for experts in international law)

Travel for the co-PI and legal $18,846 $27,997 consultant to attend the project meeting in year 1 and for the community liaison to attend various conferences (see appendix 6 for details)

Total $164,846 $177,357

2) University of Michigan Activity Year 1 Year 2

5% effort for co-PI Rebecca $4,773 $4,963 Welzenbach

Travel for co-PI to attend $1,025 $1,166 project meetings in year 1 and 2 (see appendix 7 for details) Total $5,797 $6,129

3) Curtin University Activity Year 1 Year 2

10% effort for co-PI Lucy $15,779 $16,016 Montgomery

5% effort for co-PI Cameron $9,222 $9,360 Neylon

100% effort for data scientist $91,596 $92,969 to be hired

100% effort for product $90,957 $92,324 manager to be hired

55 days of effort from a $19,712 $20,024

30 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973 computational specialist at the Curtin Institute for Computation

140 days of effort from $49,321 $50,060 technical staff at Curtin Institute for Computation for creation of software drivers

Laptops for data scientist and $6,102 — product manager ($3,051 each)

Cloud computing costs ($0.02 $3,390 $3,390 per gigabyte for Google BigQuery)

Travel for all participants $37,000 $34,000 from Curtin to attend project meetings in years 1 and 2 and to meet with select technical-development partners (see appendix 8 for details)

Total $323,078 $318,144

4) BISG Activity Year 1 Year 2

7.5% effort (year 1) and 5% $12,500 $10,280 effort (year 2) for co-PI Brian O’Leary

Business-model consultant $30,000 $4,500 (20 days in year 1 and 3 days in year 2 @ $1,500 per day, the going rate for non-profit business consultants)

Supply-chain consultant (20 $30,000 $19,500 days in year 1 and 13 days in year 2 @ $1,500 per day, the going rate for non-profit business consultants)

31 University of North Texas | Principal Investigator: Kevin S. Hawkins | Grant Reference Number: 1907-06973

Publication of white paper — $5,000 (costs for editorial and production expertise, per previous experience at BISG)

Hosting of community — $42,500 meeting (see BISG’s letter of commitment for details)

Travel for co-PI and $1,400 business-model consultant to attend project meeting in year 1 (see BISG’s letter of commitment for details)

Total $73,900 $81,780

Promotion, Marketing, and Advertising: If applicable, the total funds allocated for all promotion, marketing, and advertising costs associated with the proposed grant activities.

Total Amount Budgeted: $0

Notes:

Occupancy Costs: If applicable, the total funds allocated for utilities and space rental associated with the project or development of the project. The budget narrative should also designate the Foundation’s percentage contribution to the occupancy costs. Please refer to specific program guidelines on the Foundation’s website.

Total Amount Budgeted: $0

Notes:

Additional Budget Categories: If applicable, the total funds allocated for use in any additional budget categories. Please describe the allocated amount and purpose for each category.