Helping Digital Humanities scholars to find the tools and methods that are the most relevant to their work is a core mission of OpenMethods. By shifting perspectives from research outputs to the underlying workflow, procedures and tools, we aim to strengthen the culture of reuse of already existing resources in DH. But we do not only propagate the culture of reuse. Putting our money where our mouth is, we are also seeking ways to put DH tools in service of more effective content discovery and enrichment.
As a reusability exercise, recently we have created plugins to achieve interoperability with the entity recognition NERD service and the research discovery platform ISIDORE to increase the visibility and discoverability of our content
The NERD plugin
NERD is a service that recognizes and disambiguates named entities.
This plugin allows integration of the NERD service with WordPress. As a form of content enrichment, the plugin automatically creates tags from the named entities offered by NERD when provided the full text of the original article that has been republished on OpenMethods. The tags, in return, are used to propose extra information coming from Wikipedia and Wikidata. These tags contribute to the better discoverability and searchability of content on OpenMethods and add extra context layers to our content.
The plugin is free to download and is available for reuse form here: https://wordpress.org/plugins/nerd-wp/
It is reusable and only needs a NERD server to work with, one instance of this server is freely available at Huma-Num, please contact them for more information.
The Rich meta in RDFa plugin
The second plugin is for increasing the findability of our content on other discovery platforms via metadata enrichment. We wanted Isidore, an indexing and search service for the humanities and social sciences, to harvest our content and make it searchable and findable on their platform that serves as a single point entry for a wide range of open SSH resources, such as data, publications and other materials. To enable indexing by Isidore, OpenMethods has to be able to present information (i.e. a set of dublin core metadata elements such as title, description, date, etc.) in a format understood by Isidore. For this purpose we created this WordPress plugin that allows users to add Dublin Core metadata enrichments in RDFa within the HTML header of each post.
Currently it is only used for harvesting by ISIDORE, but it can easily be changed to accommodate other applications. In the future, it will be possible to change the namespace and element names of the RDFa data for a given input such as title or excerpt.
As a result, content from OpenMethods is now connected with and embedded in a large corpus of electronic publications, corpora, databases and scientific news, and thus Isidore users can find OpenMethods posts when looking for information about Digital Humanities. Another benefit of the plugin is that it indirectly, through OpenMethods, enables the integration of content that could not be indexed otherwise on research discovery platforms because of their language (other than English) or content type properties (e.g. blogs, videos or podcasts scattered across the web with insufficient metadata).
The plugin is free to download and is available for reuse from here: https://wordpress.org/plugins/rich-meta-in-rdfa/
The figure below allows for a sneak peek behind the scenes and gives an at-a-glance summary on how the plugins build on DH resources.
- As OpenMethods only offers users a very short introduction to and a quotation from the republished source content, first we have to go and look for the full text of the original content (the source of the entire article) from the article’s source site through an HTTP(S) request and keep the relevant textual elements (and exclude other, irrelevant elements of the page e.g. advertisements, menu etc.).
- Once the data is retrieved, it is followed by a call to NERD (runned on Huma-Num servers) so that we return the Named Entities that could be found in the text.
- These Named Entities are then ingested by OpenMethods to create tags and to be able to send the user back to entities on Wikipedia / Wikidata for more information.
To add an extra discovery layer to the OpenMethods platform, we also integrated a third plugin developed by the Isidore team: it enables listing of similar content recommendations on OpenMethods posts.
You are more than welcome to explore how the two plugins jointly contribute to the enrichment and greater visibility of the OpenMethods posts. Besides, as you would expect, both of them are free to download and are available for reuse. Take a look into their documentation and let us know if it inspired you to think about ways to put them to good (re)use in your own projects.
We owe special thanks to Pierre Mounier and Nicolas Larrousse for their idea to develop the two plugins, to Yoann Moranville who created them and to the Huma-Num team, especially to Laurent Capelli who helped us with the design and the Isidore indexing.
This post has been published under a Creative Commons CC-BY 4.0. license.