This post concerns a major milestone in the long-running pursuit of Japanese scholarly archives for use with modern research tools.

The story began with the Second CiNii Web API Contest (CiNii ウェブAPIコンテスト) in 2010. CiNii is an aggregator of Japanese journal metadata run by the Japan National Information Institute. The CiNii API specification is a thing of beauty, capable of delivering properly tagged multilingual variants of every element of item data. I believed that leveraging CiNii data in a modified version of Zotero would be a straight path to research efficiency in Asian languages. Fueled by this prospect, a small group of us coded the original version of Juris-M [1] in three months of intensive work.

The “path” proved to be less than straight. While CiNii is a very convenient service for locating Japanese academic publications, the metadata that it delivers is often of indifferent quality. As CiNii developer Ohmukai Ikki has explained, the project is constrained by the “Yoshinoya Principle,” which holds that as between “fast,” “good,” and “cheap,” you can only have two; and given its role and position, the CiNii service is forced to be fast and cheap. This is costly, as it were, to Juris-M: when fetched items require time-consuming corrections and editing, the tool feels broken and it seems less useful.

Two recent developments promise to change things for the better. One is the appearance in Juris-M of RSS feed support (which it inherits from Zotero 5.0). The other is the deployment of NDL Online, a new service of the National Diet Library. These make a very nice combination that I think will be welcomed by researchers working with Japanese materials.

zotero-new-z-48px + Feed-icon + 

In Juris-M and Zotero, you can think of feeds as a modern equivalent of a table-of-contents service. Publishers, journals, newspapers, weblogs, and many other information providers syndicate summaries of newly released items as a machine-readable “feed” available at a particular URL. When registered in Juris-M or Zotero, the feed displays as a folder in the left-side column. Clicking on a feed icon opens a list of its current contents, with unread items shown in boldface. Clicking on at item opens a summary of the item in the information pane, as shown below.


There is a button labeled “Add to My Library” in the information pane. Clicking on it will fetch the item; and if the full content is available, it too will be added to your library. Metadata retrieved from NDL records is generally good, and little post-fetch editing is required.

The NDL hosts feeds for most of the academic journals published in Japan, all in a single location. To get started, find a journal, copy its feed to the clipboard, open Juris-M, click on the folder icon at the top of the left-side column, and register the URL.

Feeds are not the only way to access NDL Online records, of course. The service offers a clean search interface, and CiNii also provides links to NDL pages from their own records, so if you prefer the CiNii search engine, you can have the best of both worlds.

It has been a long road, but I think we may finally be getting somewhere with this.

Juris-M was originally known as “Multilingual Zotero” or “MLZ” for short.