The project has demonstrated the feasibility of a monograph intersect tool that can pull down books from the Directory of Open Access books (DOAB), take an arbitrary number of these books as input and intersect them to find references cited in common. It has built a working prototype in a short space of time and shown that there are ways of using the citation graph for monographs for research benefit. As expected, there are a few technical challenges and these are described in detail in the project report. In short, to enable true scalability and practical implementation of the tool it would be helpful to improve the machine readability of open access books.
We discussed the project results in a meeting on 14th October with a group of library representatives, publishers, funders, service providers and researchers. The meeting inlcuded an overview of the open metrics lab project and presentations from Martin Eve, Birkbeck, University of London on the project results, from Pierre Mounier, OpenEdition on the OPERAS Metrics Portal and from Graham Stone, Jisc on how the two initatives are connected.
I have included a short summary that captures the discussion after the presentations on the day of the meeting. If you think there’s anything important we haven’t touched on, have any further thoughts, feel free to add your thoughts via the comments section.
Metrics for OA monographs inspire a range of concerns among researchers. Should we be developing bibliometrics for books in the first place?
- Monographs are important for many HSS disciplines. Their quality can’t be ‘measured’ in the same way as journal articles, so they often seem to get deprioritised, don’t get the same value attached to them. Citations can’t be used as indicators of quality for monographs but without them books might lose some of their significance as research communication moves online.
- We need to understand the added value of open access books to support the transition to OA.
- We need to demonstrate the value of HSS to those that make funding decisions. Citations could be useful to support stories about the quality and impact of individual books.
- The experiment demonstrated what a citation graph might achieve within a positive research context and how it can be used to benefit research.
- We can’t control how a citation graph for monographs would be used/abused – but as OA monographs are now becoming openly available others could move into this space anytime.
- Lots of tools and services are developed around OA research articles, we should also develop tools for books.
- For an overview of other OA monograph metrics initatives see the final project report
The experiment has shown the feasibility of a monograph intersect tool. To enable scalability of the tool minimum viable technical standards to improve the machine readability of OA books would be helpful. How can this be achieved?
- One way would be to get a set of technical standards defined and implemented through funder mandates.
- Smaller publishers would generally need assistance with this – possibly via a consortium model.
- There is also a role for authors to act as advocates for the implementation of standards – including e.g. the attribution of persistent identifiers to books – within their departments and for senior editors who are also academics.
- A number of large publishers are already providing good metadata for books. For some publishers this seems to depend on when the books were published, i.e. metadata for books have improved compared to a few years ago.
- The many different reference styles exist in many cases with justifiable reasons (e.g. references to a box number of manuscripts in an archive), so trying to change this doesn’t make much sense.
- There needs to be progress at different levels including:
(1) experiments such as the development of the intersect tool
(2) work with platforms to structure their content, implement standards and
(3) encourage researchers to use more innovative processes and tools (e.g. Zotero).
- A consortium approach is helpful to improve the infrastructure for open access book publishing. There are already strong initiatives, e.g. COPIM and OPERAS – but they are coalitions of the willing and it is totally different to transform an industry, which is what we are trying to do. The sharp end of funder mandates has a role to play in this respect.
Challenges around persistent identifiers for books
- There is some confusion around the role of DOIs and how to attribute DOIs to books (e.g. to what editions, different formats, book chapters –what is a chapter? etc.). CrossRef is in the process of rationalising the different ways of how book metadata is collected and will be providing advice. They would need to make the case for chapter level DOIs and would be happy to take community input on this.
- Identifiers for books are not necessary the same as for journal articles. We note that identifiers are crucial for other services e.g. do we need to think about an OA switchboard for monographs?
Is there a role for libraries?
- Yes, as advocates – they are often the first point of contact.
The conversation about this continued on Twitter after the meeting and further points that were made inlcude: Altough practice varies, lots of new roles are emerging. Libraries create, manage and exchange lots of book metadata. They facilitate OA book funding, they host OA versions of monographs on repositories, they run university presses, train students and staff on the topic and support infrastructures like the OAPEN library and DOAB. They also have a significant interest in all the points listed below under ‘what might a metrics dashboard look like’ and act as leaders, hubs and partners alonsigde advocates. [added on 24.10.2019]
What might a metrics dashboard look like for different stakeholders?
- A citation graph for books would be useful for institutions to e.g. map collaborations with other institutions.
- Geolocation (usage location) is a good news story for OA and also shows global impact in a visual sense.
- Authors require what is said about a book and by whom rather than the number of citations.
- The number of times a book features on course syllabi can support recognition of authors. (https://opensyllabus.org/)
- There might be a role for book reviews, but are they read? Also, publishers often tag them as articles to inflate numbers.
- We need to look at interventions at different levels, e.g. publishers/platforms, infrastructure, social/researcher – need to articulate the benefits to all different groups.
- Metrics need to be kept open to avoid lock-in.
- It would be useful to build case studies – there might be a link to Kudos here (which is encouraging people to tell stories).