In a year where we moved to full time home working for the entire Figshare dev team, that didn’t slow down the pace of development of the Figshare platform! We had another bumper year of changes, from exciting new features to ensuring regulatory and legal compliance. Read on to find out more.
The standout release of the year was ensuring all public pages are fully-compliant with a range of international accessibility standards (outlined below). This project was a culmination of around 2 years of effort and not only ensures regulatory and legal compliance for our customers, but forms the technology bedrock for the next half decade of Figshare. There were many ancillary benefits from this project, including but not limited to:
The new public pages are now very responsive on mobile and desktop, changing their appearance to suit multiple screen sizes down to mobile. It’s not perfect, and we’ll be cleaning up areas of this in the next few months, but you can play with this today by resizing your browser window.
Try clicking into an object from the Figshare search page and you will see a significant increase in loading speed. This comes from a partial refactor of the search engine service in order to provide facets.
Additional benefits come from a migration to graphQL middleware server, where multiple queries can be made into a single call, thus reducing the http overhead. The final speed improvement comes from a UX tweak of pre-rendering part of the page on the server, showing earlier progress for the user. In short, faster search for faster results.
The new infrastructure that facilitated the accessibility changes also will allow for true multi-language support within the interface for labels and menus
This important new functionality is being done to better facilitate indexing in external search engines such as Google Scholar, especially as increasingly more institutions are using Figshare as an all-in-one repository.
A typical item page currently look like this:
https://yourdomain.edu/articles/object_title/numeric_object_id
We will be changing these to the following to allow an "item type" variable to populate the URL: https://yourdomain.edu/articles/item_type/object_title/numeric_object_id
The old URL will still work and will redirect to the new one. This change will take place for all items, both historic and new.
We are looking to deprecate support for the desktop uploader at some point in the future and as part of this, we are introducing new functionality of an FTP uploader. This will enable robust uploading of many files to figshare via any FTP client.
You can find information on the FTP uploader on the Figshare Knowledge Portal (https://help.figshare.com/article/upload-large-datasets-and-bulk-upload-using-the-ftp-uploader-desktop-uploader-or-api). Dependent on the method of login, you may have to create a password to access this functionality or you will use your existing account password. For each folder you add, you will create a new Figshare object.
As there is quite a lot around the new search page, I have broken this out into its own section. To start, why have we invested such effort into a new search page?
Content types & volume
Figshare search was built primarily with visual search in mind. As the volume of text-based content has increased dramatically, the need to offer a list-based view of results, as well as tile-based for our more visual content types, has grown.
New uses of Figshare
With institutions using Figshare as a paper, thesis, and dissertation repository, as well as being the infrastructure for numerous high-volume preprint services like ChemRxiv (https://chemrxiv.org), SAGE Advance (https://advance.sagepub.com/), and TechRxiv (https://www.techrxiv.org/), the data/visual focus to search was not entirely suitable.
Exposure of advanced search
Figshare’s existing advanced search facility required the use of operators and a strict search syntax that could pose a barrier to new users. By creating a new faceted search page, a lot of this power has been translated visually into the new facets and filters.
Future-proofing
With many plans for the future of search, like full text search and custom facets (explored later), these features would have been very difficult to implement on our current search implementation.
Accessibility
We have an ongoing accessibility project running across the entire Figshare platform. This has meant changes big and small, and the other reasons outlined meant this was a good opportunity to tackle accessibility for this page at the same time.
Visually, you can see some immediate changes. It’s cleaner and we believe it’s easier to use. We offer list and grid view support to allow for an optimized search experience dependent on the type of content you are looking for.
From a UX point of view, there have been many performance optimizations applied, so you should have a smooth and enjoyable experience, regardless of the amount of results returned.
One of the biggest differences is the new sidebar. It is:
Accessibility
The importance of accessibility is embedded in the foundation of all new Figshare developments. This new page fully complies with:
This manifests (among the visual and programmatic elements) as having some great keyboard navigation elements such as:
Using Ctrl+Home/Ctrl+End to enter/exit the search results listing:
A long-requested pair of new features have also been launched alongside the new search page. The first is the ability to continuously browse through results, whereas before the user was limited to 80. New items will be loaded from the server as you scroll through the page.
Another request is that it can be difficult to navigate to the footer of the page, something that would only have been exacerbated by the release of infinite scrolling. To ameliorate this, you can now show (“lock”) the footer as always displayed or minimize to focus on your results.
Searchable Citations on Figshare
At Figshare, we’re very fortunate to be part of the Digital Science family. As such, it means we often get to work in close collaboration with other sister companies, such as Altmetric doughnuts on Figshare pages and our widely-adopted Symplectic integration.
Working in conjunction with our friends at Dimensions, we’re pleased to be able to offer daily citations updates on any object within any Figshare instance. Not just papers, but datasets and videos and, well... anything! To illustrate this and before we look at how it works, let’s take a moment to showcase some of the interesting NTROs that we we have found as part of this feature release:
Good examples of “classic” data being cited:
Interesting to see yet not surprising, code is more frequently cited as it is designed to be reproducible and reused:
Great to see our institutional partners seeing their content being reused:
And to finish, a great example of content that previously would not have been tracked having a big impact:
How it works - in brief
We receive daily updates from Dimensions of the latest citation data. Dimensions are looking in the full text of documents and capturing any mention of a Figshare-affiliated DOI string. Citation data is actually stored within the Figshare database and is not pulled live, meaning we’ll be able to expose this in all kinds of interesting ways in the future e.g. citations over time, most cited, cumulative citations by group/lab.
We are not presenting this as an exhaustive and infalible list of all citations for Figshare objects. With continual refinements in coverage, indexing and parsing in the future, we can expect to pick up not only more mentions going forwards but also to increase the quantity and quality of the existing pool.
Why it’s special
At Figshare, we are fully aware of the power of citations to an academic. It has been highlighted in our State of Open Data reports as the main motivator in sharing open data - more impact in the form of more citations.
However, data citations still struggle to find their way into reference lists, but both funder and publisher requirements around open data have resulted in an exponential growth in citations to data across data repositories. For the most part, these ‘citations’ are found in the full text of the article, the methods section - or the data availability section.
As such, we have partnered with our sister company Dimensions to identify these citations in the full text of articles. This ensures that researchers, and our clients, are aware of all of the citations to their research.
We’re incredibly proud of the work we’ve done for our community in 2020 and cannot wait to share what’s in store for 2021.
If you’d like to talk more about Figshare’s features, releases, and what’s in store for the platform, please get in touch at info@figshare.com.
Christopher Blumzon
product director, Figshare
chris@figshare.com
Jan 12, 2021 10:13
This is a reposted article from our blog, the original article can be found at:
https://figshare.com/blog/Figshare_Product_Development_2020_-_A_Year_in_Review_part_1_/605