ResearchGate and Academia.edu
ResearchGate and Academia.edu are known as academic social networking sites (ASNS). Their stated aims are to connect researchers with common interests based on their institutional affiliation and the type of research content they make available online. However, as commercial entities, they offer a service at no initial cost to the consumer as a way of establishing a foundation for future transactions including offering premium features to fee paying members (such as extra analytics or advanced search) or advertising space for companies and other organisations.
In 2016 for example, Academia.edu trialled a service which would enable authors to pay for their work to receive a ‘recommended’ status in addition to disclosing readers’ ranks (Bond, 2017, paras. 5 & 6, cited in Makula, 2017).
UWL Repository
The UWL Repository is a non-commercial digital archive of research and scholarship maintained by the University of West London library service, enabling authors to showcase a version of their output (usually in the form of an author manuscript) for the benefit of anybody with an Internet connection. The primary aims of the UWL Repository are to make these outputs as widely available as possible (with limited restrictions over reuse) and to ensure their long-term preservation. These conditions also allow staff members at UWL to comply easily with national, institutional and funder mandates..
Key Differences
1. Open access
While Academia.edu and ResearchGate encourage users to upload content to their respective platforms in order to make them publicly accessible, they do so by declaring ownership of the content they receive (upon accepting the terms of use on registration*).
In order to protect these interests, they are obliged to distance themselves from openly supporting re-usage rights and actively prohibit the use of robots, crawlers, and data mining tools.
Furthermore, lack of clear guidance on self-archiving on both websites (in addition to general misconceptions around open access) has resulted in users routinely infringing copyright or breaching publisher agreements as well as journal policies on embargos. In 2013 for example, Elsevier sent almost 3000 DMCA takedown notices to Academia.edu on account of what it claimed to be unauthorised sharing of the publishing giant’s content (Hall, 2015).
The UWL Repository on the other hand is an extension of the research and scholarly communication support provided by the library and is committed to openness as defined by the Budapest Open Access Initiative (BOAI). In addition to access for example, the UWL Repository recognises the importance of the availability of research and scholarship under conditions which permit users to read, download, copy, distribute, print, search, or link to the full texts of articles, crawl them for indexing, pass them as data to software, or use them for any other lawful purpose. Moreover, in situations where access to the full text of an output on the repository is restricted (this is sometimes the case with monographs and book chapters or where content is generally subject to an embargo,) anyone may access the metadata free of charge.
2. Discoverability
Many have observed that the majority of works posted to academic social networking sites lack robust metadata and that upload interfaces are geared towards enriching proprietary content rather than towards promoting the sharing of correct and complete scientific information (Dingemanse, 2016). As a result academic social networking sites often fall short of complying with minimum guidelines for metadata exposure of indexing services such as Google Scholar which affect search rankings and overall research findability. Up until recently, Academia.edu for example did not provide a field for digital object identifiers (DOI) and still does not allow PMID codes or links which direct users to sources outside of their platform.
One important way in which the UWL Repository optimizes discoverability of content is by using standardised records, based on established metadata schema. These make it easier for web indexing services (like CORE and Google Scholar) that use “parsers” to identify bibliographic data, to recognise papers and improve their overall visibility. Unlike ASNSs, dedicated library staff are available to review and improve the metadata on the UWL Repository, to increase the discoverability of these author outputs. Moreover, integration with CORE services (which stands for COnnected REpositories) allows repository users to benefit from a recommender feature which displays results for similar titles across other repositories, meaning that those users are not simply confined to one research or scholarly archive.
3. Interoperability
Hostility towards interoperability (standards which facilitate technical reciprocity between other products or systems) by Academia.edu and ResearchGate means that many preventable errors relating to paucity of bibliographic data and record duplication are never resolved. On Academia.edu and ResearchGate there is no way to import a record via a DOI or PMID code via the upload interface which would help to harmonise records and address version control. In instances where ResearchGate permits users to import publications from some applications, it provides no method for getting that same data out of the ResearchGate ecosystem (Fortney and Gonder, 2015).
According to the Confederation of Open Access Repositories (COAR):
A single scientific repository is of limited value, real benefits come from the ability to exchange data within a network … interoperability allows us to exploit today’s computational services, and generate new knowledge from repository content (COAR, n.d., cited in petrknoth, 2019).
In contrast to ACSNs, the UWL Repository, respects users’ personal data whilst being fully interoperable by design (this is thanks to the power of something called the Open Archives Initiative Protocol for Metadata Harvesting or OAI-PMH). This allows it to communicate with other systems and transfer information, metadata, and digital objects between those systems as a matter of course.
The screenshot below for instance, shows how easily researchers are able to pull existing metadata records into the repository using a variety of sources including reference managers, scripts and unique identifiers.
The UWL Repository also supports ORCiD (Open Researcher and Contributor ID) integration, allowing researchers to easily import and export records from the two systems, reduce duplication of effort and embed their unique persistent identifier within their submissions.
The statistics area also offers a range of options to customise repository data (by author, work, deposit, requests) and export the results as XML, JSON or CSV files, all without even logging in.
4. Searching
According to Makula (2017) ‘Academia.edu offers only the ability to browse by title, not subject, and a basic (not advanced) search mechanism that provides no guidance on how to conduct an effective inquiry’. Advanced searches are reserved most notably to premium account holders only.
Conversely, institutional archives such as the UWL Repository are adapted to support both browse and advanced search functionality depending on user preference. Users are invited to browse items by year, subject, academic school, author and item type as well as carry out simple and complex searches using faceted techniques and a range of category filters such as item type, publication status, journal titles, etc.
Burnett (2013) also points out that institutional and subject repositories have common standards for describing their contents which enable their metadata to be harvested by search engines and repository-specific search services. This capacity allows users to cross-search around 3000 repositories worldwide with just one search query.
Just a few examples of such search services include:
- BASE – from Bielefeld University Library, who provide more than 150 million items comprising a mixture of both “metadata-only” records (which contain no full–text) and actual full–text records.
- CORE – from the Open University, who index about 11 million records with extracted full–text, with content harvested mostly from repositories.
- Unpaywall – from non-profit Impactstory, who harvest content from most journals as well as institutional repositories, with an index which currently hosts up to 20 million full-text and prioritises accuracy and precision (Burnett, 2013; Priem, 2019).
5. Business model
Both Academia.edu and ResearchGate are currently supported by venture capital funds. Although Academia.edu has a ‘dot.edu’ URL, it is not operated by a higher education institution or body. According to a blog post from University of California’s website, the domain name was registered before the rules that would now prohibit this use went into effect. On its filings with the Securities and Exchange Commission for instance it uses the legal name Academia Inc. (Fortney and Gonder, 2015).
Many commentators have ventured that the financial rationales of both Academia.edu and ResearchGate rest ultimately ‘on the ability to exploit the data flows generated by the academics who use the platform as an intermediary for sharing and discovering research’. Trending research data from academic social networking sites can then be repackaged to R&D institutions in a market which currently spends $800 billion on R&D in the private sector (including pharmaceuticals) globally (Hall, 2015).
6. Values and mission
Academic social networking sites have been widely criticised for attempting to recentralise disparate scholarly-led approaches to open access, in order to derive commercial value, particularly in the case of user-driven data. They are perceived by some to perpetuate individualistic practices which are in tension with the ethos of earlier community-orientated open access movements and consequently reproduce many of the problems associated with the pressure to demonstrate impact within the confines of legacy publishing. (Dingemanse, 2016; Hall, 2015). The ‘RG score’, a proprietary algorithm developed by ResearchGate is for instance reported to reach high values under highly questionable circumstances (Jordan, 2015; Murray, 2014).
The fleet of analytics conceived by academic social networking sites increasingly normalise surveillance. Academia.edu have perfected this with their increasingly intrusive notifications (“Someone in Sheffield, United Kingdom, mentioned you”) and ResearchGate have recently acquired a US patent for “Enhanced Online User-Interaction Tracking and Document Rendition” (Price, 2020). Both also promote social ranking over potentially meaningful engagement in a way that ‘plays to our vanity [with] elements of its design [being] built to satisfy and amplify our craving for external validation’ (Dingemanse, 2016).
This is sustained by a campaign of aggressive emailing designed to incite users to engage with the platform further, often on the basis of fairly mundane activity.
A 2014 Nature report also revealed that a proportion of the profiles on the site are not maintained by real people, but are automatically generated – though usually incompletely and with little regard to the individual. This understandably annoys researchers who do not want to interact with the site but who then feel pressured to engage because the page misrepresents them.
Institutional and subject repositories, by contrast, focus on balancing support for open access to research and scholarship for the public good with the needs to meet conditions for future research funding. The University of California observes that ‘open access repositories are usually managed by universities, government agencies, or non–profit associations. Affiliation with a larger institution (that has a public service mission) frees repositories from shareholder pressures meaning they are less likely to make decisions with future market potential in mind. The UWL Repository was also conceived to help respond to national and funder mandates of which the most widely applied in the UK is the Research Excellence Framework (REF).
7. Long-term preservation
Institutional repositories often employ librarians who specialize in ensuring long term archiving. As such, the preservation policies developed for institutional repositories, including the UWL Repository, typically enable items to be retained over an indefinite period and take reasonable efforts to ensure continued readability and accessibility. This is vital where ‘some publications in the repositories may not be published elsewhere. This is often the case with student theses and doctoral dissertations, as with report series, digitised material and open educational resources’ (Ball, 2010). The UWL Repository also regularly backs up its files according to current best practices and regulates withdrawal of records according to a stringent set of criteria (such as proven copyright violation, plagiarism or research that otherwise breaches the University’s code of conduct).
Academia.edu and ResearchGate on the other hand, are independent for-profit companies that could cease operating at any time. While Academia.edu disavows any duty to warn users if they shut down, ResearchGate places unreasonable demands on the user to be kept informed of changes to their terms of service:
- Academia.edu “reserves the right, at its sole discretion, to discontinue or terminate the Site and Services and to terminate these Terms, at any time and without prior notice.”
- ResearchGate “We reserve the right to modify the Service or to offer services different from those offered at the time of the User’s registration at any time. […] You are obliged to check your account regularly for a notice about changes to these Terms.”
8. Recognition of open access policies
The Research Excellence Framework (REF) 2021 Guidance on submissions (2019) sets out the criteria for open access compliance in order to help assess the types of research being produced by higher education institutions in the UK. One of the requirements states that the ‘output must have been deposited in an institutional repository, a repository service shared between multiple institutions, or a subject repository’. Since Academia.edu and ResearchGate are not institutional repositories, do not have a clear subject-disciplinary focus, and do not satisfy other funder requirements, embodied by pan-European funder initiatives such as Plan S (this stipulates the need for OpenAIRE compliance, quality metadata in interoperable format, continuous availability, etc.), they are widely considered to be unsuitable venues for open access impact evaluation (Hubbard, 2016).
Furthermore, REF guidance states that an ‘output must be deposited within the repository within a specific timeframe, determined by the date of acceptance’. Academic social networking sites such as Academia.edu and ResearchGate however provide no option to upload an accepted manuscript (also known as the post-print) as the article file type. Academia.edu for example, only supports ‘published works’ and non-descript ‘drafts’ for articles while ResearchGate similarly provides a binary choice between ‘published research’ or ‘preprints’ which have not yet been subject to peer review. In the case of both platforms, authors are limited to assigning just one date per record. These limitations and poor metadata standards make it exceptionally difficult to determine the point at which an article was accepted for publication, particularly as an article’s publication status changes over the course of its transition from first submission to final publication. This has clear implications for research and scholarly auditing processes which ultimately affect allocation of future funding to institutions.
What then may outwardly appear as sleek and minimalist upload interfaces that simplify the process of depositing for users, are services that are largely stripped of the checks and balances that would otherwise allow authors, creators, auditors and funders to navigate complex questions of ethics, licensing, ownership, provenance, time-stamping and compliance.
9. Dedicated support
COAR (2019) suggest that while large, centralized infrastructure and services such as Academia.edu and ResearchGate may be easier to market, ‘they cannot be as responsive to a diversity of needs and priorities across regions and domains. In addition, local services can engage with the local researchers to help ensure that their outputs are being described and deposited correctly’.
In this regard, the UWL Repository represents a support service for researchers and scholars based at the university as well as a vehicle for providing green open access.
The Repository website provides clear author guidance on how to deposit, through the availability of step-by-step guides and a video tutorial; details on relevant policies, copyright and flavours of open access; advice on how creators can responsibly promote their work for impact and a section on frequently asked questions.
In addition to this repository administrators manually review and enhance deposited records to monitor open access compliance, ensure integrity of research and scholarly outputs and check records against publisher terms for copyright and self-archiving.
This contrasts starkly with the disclaimers around liability and customer support which are featured heavily throughout Academia.edu and ResearchGate’s terms of service:
ResearchGate: Scope of the Service (2019)
‘We do not preview or review or filter such information, neither manually nor automatically. Therefore, we do not and cannot have current knowledge of possible infringements, inappropriate content, or violations of law caused by information that is uploaded by and/or stored upon the request of Members’.
Academia.edu: Disclaimers (2019)
‘THE SITE, SERVICES AND COLLECTIVE CONTENT ARE PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED. WITHOUT LIMITING THE FOREGOING, ACADEMIA.EDU EXPLICITLY DISCLAIMS ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT, AND ANY WARRANTIES ARISING OUT OF COURSE OF DEALING OR USAGE OF TRADE. ACADEMIA.EDU MAKES NO WARRANTY THAT THE SITE, SERVICES OR COLLECTIVE CONTENT WILL MEET YOUR REQUIREMENTS OR BE AVAILABLE ON AN UNINTERRUPTED, SECURE, OR ERROR-FREE BASIS. ACADEMIA.EDU MAKES NO WARRANTY REGARDING THE QUALITY OF ANY PRODUCTS, SERVICES OR COLLECTIVE CONTENT PURCHASED OR OBTAINED THROUGH THE SITE OR SERVICES OR THE ACCURACY, TIMELINESS, TRUTHFULNESS, COMPLETENESS OR RELIABILITY OF ANY CONTENT OBTAINED THROUGH THE SITE OR SERVICES.
NO ADVICE OR INFORMATION, WHETHER ORAL OR WRITTEN, OBTAINED FROM ACADEMIA.EDU OR THROUGH THE SITE, SERVICES OR COLLECTIVE CONTENT, WILL CREATE ANY WARRANTY NOT EXPRESSLY MADE HEREIN’.
10. Standards
- International registries:
The UWL Repository is registered with Directory of Open Access Repositories (OpenDOAR) which is the quality-assured global directory of academic open access repositories. It enables the identification, browsing and search for repositories, based on a range of features, such as location, software or type of material held.
- Adherence to technical standards:
The UWL Repository ensures all metadata – the information about what’s in the repository – are interoperable and open by using an internationally-agreed set of technical standards. This common protocol to which it adheres is called the open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) and is stewarded by OpenAIRE.
- Persistent identifiers (PIDs):
The UWL Repository uses persistent identifiers (PIDs) such as ORCiDs and digitial object identifiers (DOIs) in an attempt to solve the problems of resource identification and long-term access to online digital materials. PIDs allow resources within the repository to be uniquely identified in a way that will not change if the resource is renamed or relocated, and will persist regardless of the protocol used to access it. This means that a resource can be reliably referenced for future access by humans and software.
- Indexing and search engine optimisation (SEO) standards:
UWL Repository content is searchable via major search engines including Google Scholar and allows indexing with services such as CORE.
- Compliance with copyright and licensing conditions:
SHERPA RoMEO is an online resource which aggregates and analyses publisher open access policies from around the world and provides summaries of self-archiving permissions and conditions of rights given to authors on a journal-by-journal basis. The UWL Repository and its users navigate this service to find out which version of their articles can be self-archived under what condition, ensuring all outputs are licensed appropriately and archived responsibly.
- Open source:
The open source software upon which the UWL Repository is based (EPrints) leverages the care and knowledge of the community of developers, librarians and users that feed into its progress.
Blog post by Camille Regnault
References:
Academia.edu. (2017). Terms of Use. Available: https://www.academia.edu/terms. Last accessed 27 Aug 2019.
Ball, A. (2010). Preservation and Curation in Institutional Repositories. Available: http://1seminariopreservacaopatrimoniodigital.dglab.gov.pt/wp-content/uploads/sites/19/2015/08/recurso_02.pdf. Last accessed 01/10/2019.
Burnett, P. (2013). Cross searching of institutional repositories. Available: http://blog.inasp.info/cross-searching-institutional-repositories/. Last accessed 30 Sep 2019.
COAR. (2019). Repository Toolkit: Role of repositories. Available: https://coartraining.gitbook.io/coar-repository-toolkit/introduction/role-of-repositories. Last accessed 28/11/2019
Dingemanse, M. (2016). How Academia.edu promotes poor metadata and plays to our vanity. Available: http://ideophone.org/academia-edu-poor-metadata-vanity/. Last accessed 20/08/2019.
Fortney, K. and Gonder, J. (2015). A social networking site is not an open access repository. Available: https://osc.universityofcalifornia.edu/2015/12/a-social-networking-site-is-not-an-open-access-repository/. Last accessed 20/08/2019.
Hall, G. (2015). Does Academia.edu Mean Open Access Is Becoming Irrelevant?. Available: http://www.garyhall.info/journal/2015/10/18/does-academiaedu-mean-open-access-is-becoming-irrelevant.html. Last accessed 01/10/2019.
Hubbard, B. (2016). In the context of Open Access policies in the UK, what is a “repository”?. Available: https://scholarlycommunications.jiscinvolve.org/wp/2016/10/14/in-the-context-of-open-access-policies-in-the-uk-what-is-a-repository/. Last accessed 06/12/2019.
Price, G. (2020). ResearchGate Receives U.S. Patent For “Enhanced Online User-Interaction Tracking and Document Rendition”. Available: https://www.infodocket.com/2020/02/11/researchgate-receives-u-s-patent-for-enhanced-online-user-interaction-tracking-and-document-rendition/. Last accessed 12 Feb 2020.
Jordan, K., 2015. Exploring the ResearchGate score as an academic metric: reflections and implications for practice.
Makula, A. (2017). “Is it like academia.edu?”: Faculty perceptions and usage of academic social networking sites and implications for librarians and institutional repositories. Journal of New Librarianship. 2 (1)
Murray, M. (2014). Analysis of a Scholarly Social Networking Site: The Case of the Dormant User. SAIS 2014 Proceedings. 24. Available: https://aisel.aisnet.org/sais2014/24. Last accessed 02/10/2019
OpenDOAR. (2013). UWL Repository. Available: http://v2.sherpa.ac.uk/id/repository/2700. Last accessed 27/08/2019.
petrknoth. (2019). Data discovery opportunities: Lessons learned from CORE. Available: https://www.slideshare.net/petrknoth/data-discovery-opportunities-lessons-learned-from-core. Last accessed 28/11/2019
Priem, J. Scholcomm-request@lists.ala.org On Behalf of Jason Priem. Re: [SCHOLCOMM] List of OA indices. 03/05/2019
ResearchGate. (2018). ResearchGate Terms of Service. Available: https://www.researchgate.net/terms-of-service#serviceChanges. Last accessed 27 August 2019.
van Noorden, R. (2014). Online collaboration: Scientists and the social network. Available: https://www.nature.com/news/online-collaboration-scientists-and-the-social-network-1.15711. Last accessed 02/10/2019.