A Collaborative System for Digital Research Environment via IIIF

Introduction
It is important for researchers not only of humanities but also digital humanities (DH) to observe primary resources or evidences to ensure the reliability of their research. Many cultural institutions such as research libraries and museums have provided their cultural properties in the form of digital images on their Web sites, but researchers have not efficiently taken good advantage of them until recently because the images have been kept behind their luxury Web interfaces. However, the situation has rapidly been improved due to the spread of the IIIF (International Image Interoperability Framework).
[1] Due to the specifications of IIIF, researchers have been able to produce various solutions for their research with digital methods. In this situation, the authors will present an attempt of a collaborative environment leveraging IIIF-compliant digital images for the DH.

The situation of digital resources
Some institutions have a team of capable specialists which ensures that their digital collections will be available for researchers. However, as many institutions do not have such specialists, their digital collections often end up being released without sufficient metadata and descriptions. Moreover, in both cases, it isn’t easy for researchers to find useful resources among the individual silo-like Web sites. while integrated search systems like Europeana and DPLA are offered. Researchers must often seek for a series of digital images of a work in digital collections on many Web sites with various interfaces. Hence the authors of this presentation have developed and managed an integrated collaborative system to gather IIIF-compliant digital resources for a restricted field to solve the situation.

The systems: IIIF-BS and J-IIIF
The system consists of open source software including CentOS7, Apache httpd, PostgreSQL, Apache Solr, and so on. As examples of the targeted fields, we adopted resources of Japanese studies and Buddhist studies which are included partially in many digital collections around the world such as Gallica and Bayerische Staatsbibliothek. The system offers a collaborative environment to aggregate IIIF-compliant digital resources so that users can search and leverage them through it. When an authorized user inputs a IIIF Manifest URI, the URI and its metadata are inserted into the database (PostgreSQL) and the search software (Apache Solr). After that, any user can search, browse, annotate, and crop the IIIF contents on the Web site with the included IIIF viewers.
The system for Buddhist Studies called “IIIF-BS” (Fig 1) (
http://bauddha.dhii.jp/SAT/iiifmani/show.php) includes 6846 items from 24 institutions added by researchers of Buddhist studies. For Japanese Studies called “J-IIIF” includes 193 items added by participants of a series of IIIF workshops in Japan. As the idea of the J-IIIF has recently been updated to the other system (
http://iiif3.dl.itc.u-tokyo.ac.jp/s/iiif/page/home), it will be used only for the workshops. The IIIF-BS has continuously been developed further in order to improve the digital research environment for Buddhist studies.

(Fig 1. IIIF-BS)
IIIF-BS provides a function to add some specific metadata to the registered IIIF Manifest URIs so that the resources can be leveraged with a de-facto method in the field.(Fig 2) The method means that readers can find a certain text in the scriptures by one or a pair of unique line numbers which are embedded in most of the major scriptures of Buddhism
[2]. Then, digital images of fragmented manuscripts included in large digital collections can be identified and collated on this system. IIIF-BS also provides JSON-style data including the IIIF Manifest URI and collaboratively added metadata like start and end line numbers of the textual images by use of certain URLs.

(Fig 2. Adding annotation and metadata to a IIIF resource on IIIF-BS)

The current achievements of the system
Currently, IIIF-BS has achieved two things: (1) an integrated viewing of the digital textual images and (2) offering detailed metadata to a university library. The former is that of an integrated Buddhist text database, SAT2018. Here we show a table of the images of each scripture by use of the JSON data on-the-fly.(Fig 3, Fig 4) The latter is that of Kyoto University Library which releases several digital collections including Buddhist scriptures with little or no metadata adopted from the detailed information embedded by IIIF-BS.

(Fig 3. A table of digital images including fragments which are in put an appropriate location)

(Fig 4. Comparing digital images by click on the images on the table)

Conclusion
As this case shows, while IIIF is a general technology, specific function for a certain field is efficient and useful not only for researchers but also for cultural institutions as providers of research resources. The authors will seek for ways to make such functions more general through both aspects of the system and the IIIF specification itself.

Bibliography
[1] International Image Interoperability Framework, About IIIF https://iiif.io/about/ (accessed 2018-11-26)
[2] Kiyonori Nagasaki, Toru Tomabechi, Masahiro Shimoda, Towards a digital research environment for Buddhist studies, Digital Scholarship in the Humanities, Volume 28, Issue 2, 1 June 2013, Pages 296–300, https://doi.org/10.1093/llc/fqs076