A very Happy New Year to all our readers. We wish you all the best for 2019. It’s going to be an exciting year for DOAJ and for open access. Much of the work which you can read about below will come to fruition in 2019 , as well as new opportunities presenting themselves.
With a revised set of Principles of Transparency and Best Practice and a new mission, DOAJ started 2018 by publishing its strategy to show the community where DOAJ is focussing its efforts: a) funding and sustainability; b) functionality, stability and scalability; c) education and outreach.
Financially, DOAJ has seen the benefits of the SCOSS initiative, with more than 60% of all monies being donated from the public sector:
| Income 2017 and 2018 per category in £ (GBPs) | | |
| | 2017 | | 2018 |
| Libraries | 89,000 | 26% | 282,867 | 46% |
| Library Consortia | 86,000 | 26% | 146,590 | 24% |
| Research Funders | 31,000 | 9% | 45,900 | 8% |
| Smaller Publishers | 8,000 | 2% | 4,471 | 1% |
| Sponsors | 123,000 | 36% | 129,555 | 21% |
| | | | | |
| Total | 337,000 | | 609,383 | |
Below are details of the major projects that we have undertaken in 2018. The terms in brackets correspond to the three arms of the strategy. If you have any questions about them, or would like to know more, please get in touch: dom@doaj.org.The full details of DOAJ finances are always posted on the IS4OA website.
1. Faster evaluation of journals (funding and sustainability)
We have made excellent progress in this area and this will be of real interest to many of our stakeholders, particularly the publishers. For the first time since before 2013, we do not have a backlog of applications waiting to be triaged*. Here are the changes we made which led to this excellent result:
From mid-December 2017 to mid-December 2018, Triage rejected without review over 2000 poorly completed applications, removing them from the system so that they wouldn’t end up with the editorial teams.
- We carried out a full editorial workflow review to identify bottlenecks and areas where we could become more efficient. The goal was to make the customer experience smoother and more uniform, and cut down turnaround times.
- We have created a monthly report which identifies applications which are stuck (we haven’t received a response from the journal contact), or which are reaching an age of 6 months. This has helped to focus activities on those applications which need it the most.
- *We expanded our “Triage process” to include many more preliminary checks and have re-organised how Triage assigns applications to editors for review. (The purpose of Triage is to filter out all poor quality, incorrect or incomplete applications so that they do not clog up the editorial queues.) This has resulted in faster turnaround times, since fewer low quality applications are making their way through Triage.
- We introduced a ‘Quick Reject’ email feature which allows journals rejected at Triage to get an instant notification that their application has failed Triage. The automated email contains the reason for rejection. Prior to this feature some applicants were having to wait before getting an answer that their application had fallen at the first hurdle.
- We made it easier for quality applications to be allocated directly to a reviewer, rather than sitting in a holding pattern waiting to be processed. We did this by creating a method of grouping applications and enabling Triage to allocate all of them in the same way.
- We have reduced the amount of people that an account holder comes into contact with. This improves transparency and also cuts down turnaround times.
- We have strengthened API capabilities so that any account holder can send us applications in bulk automatically via the API. The same goes for uploading or correcting article metadata and updating journal metadata.
The primary goal of these changes is to free up resource, resource which is better focussed on reviewing and processing the “good” applications. This project has been incredibly effective and we are very happy with the results here.
2. Improved monitoring of journal compliance and systemic weeding campaign (funding and sustainability)
The introduction of an update function allowed us to make systematic journal entry reviews more focussed and more effective. These are undertaken as each update is submitted. Further reviews are taken across our larger multi-journal accounts where, as far as possible, we have tried to establish common metadata entries across all journals belonging to the same publishing entity.
As well as manual reviewing, we have undertaken 2 major automated reviews of our data: one for broken URLs and one for erroneous ISSNs. Both revealed problematic entries in the database which we were then able to target and either correct or remove. This was a new approach for us and one which is easily repeatable at any point in the future.
The result of this work is that we can categorically state that the metadata is purer and cleaner than it has ever been before, both at journal and article level. The metadata is now more correct and up-to-date than it has ever been. For example, an automated review of URLs in the database recently revealed that 121 journals in the database had broken URLs. That’s just 1% of all journals in DOAJ.
3. Enable greater collaboration with organisations to enrich metadata (education and outreach)
Much of the ability to enrich metadata must come after the refactoring work (see point 7 below) is complete. However, this hasn’t stopped us making a little progress in this area. We had already met with Altmetric in February 2017 to discuss the integration of alternative metrics into DOAJ. A sticking point then which limited the usefulness of such a feature is that a large amount (maybe almost 50%) of the article metadata had no DOI. We are working to improve this figure.
In 2018, DOAJ renewed its partnerships with Research4Life, ISSN ROAD, OCLC and Think.Check.Submit. DOAJ entered into new partnerships with Digital Science, Library Publishing Coalition, Creative Commons, Science Afrique and C4DISC. We are also working closely with the OJS team since so many of the journals in DOAJ use the OJs software.
4. Strengthen coverage and diversity in terms of language and geography, including an extended Ambassador programme and translated materials (education and outreach)
This is another area in which we have made excellent progress. We have undertaken the following initiatives:
5. Governance (funding and sustainability)
The development and implementation of a governance structure is almost complete. In December, DOAJ published its new governance model which was created with input from the existing advisory board. The new governance structure enables all of those organisations who fund DOAJ to nominate individuals from their organisation to sit on the DOAJ board. DOAJ will be announcing the details of nominations and elections later this month.
6. Account holder survey (funding and sustainability)
We sent out a survey to the community of 6000+ DOAJ account holders asking for their opinions on what we should be developing next, benefits of being indexed in DOAJ, technologies that they employ etc. We had 1065 responses with a response rate of 25% which is excellent for an online survey. The results from the survey and some deeper analysis will be published later this month.
7. Refactoring Project (functionality, stability and scalability)
Last but certainly not least, we have undertaken a huge amount of work to stabilise the foundation that the DOAJ database sits on. It’s tecchy to a level which certainly won’t appeal to everyone but it is unarguably the most important piece of work that we have done this year and, it turns out, already puts us in an excellent position for whatever 2019 and 2020 will throw at us.
DOAJ receives more than 75,500 hits per day to the site. The API received 130,804,030 hits in the first 6 months of 2018 alone. The database holds 4.3 million pieces of article metadata
In 2014, when DOAJ was relaunched on its current platform, the database was receiving about 22,000 hits per day. Today, we are seeing figures upwards of 75,500 per day (averaging 1.2 million hits per month), plus an increase in traffic from referrals (75% of all DOAJ traffic comes via referrals), plus huge amounts of activity via our API (2018: 266,255,000 hits; 2017 total: 187,212,674). In 2014, the database held about 2 million pieces of article metadata; today it is about 4.3 million. The need for a holistic, strategic approach was very real to ensure that there would be no surprises along what is turning out to be a 3-year project.
In December 2017, DOAJ and its technical partner, Cottage Labs, finalised the list of tasks. To date, we have completed 20 of the 40 original tasks, although the remaining 20 tasks are bigger and meatier chunks of work which will take us well into 2019 and beyond. If you’d like details, let me know.
Some other things that we have done to ensure that we have a stable and scalable back-end:
- Upgrades required to protect DOAJ from the Meltdown and Spectre vulnerabilities.
- Machine snapshots, security updates, and resize operations across 4 servers
- Increased our number of developer days at our technical partner to 20 days per month which was seen as a necessary increase if DOAJ was to achieve everything that it sets out to do in its strategy document.
- Made data at DOAJ secure, following the recommendations laid out by the GDPR which came into effect in May.
- Limited unintended access to the query endpoint.
- Greater integration into Google Analytics.
