You will remember, no doubt, what was said about the dance routines which Ginger Rogers did with Fred Astaire: “Ginger Rogers did everything that Fred Astaire did, but she did it backwards and in high heels”.
The quotation came to mind as I read Brett Burney’s article Successful Predictive Coding in an unfamiliar linguistic landscape. That was written in conjunction with UBIC, the developers of eDiscovery software developed originally for Asian languages.
Predictive coding is complex enough. Add the need to deal with Chinese, Japanese and Korean languages, called collectively known as “CJK” and you have the equivalent of a Ginger Rogers dance routine, with the complexities of the CJK matching the “backwards and in high heels” requirement.
Brett Burney, whom I met for the first time at ILTA, makes a good job of explaining how predictive coding, including UBIC’s Lit i View, works, and describes the additional hoops added by the presence of CJK documents in a collection.
As Brett makes clear, the complexity is not simply one of language, hard though that is on its own. Where most English language (and most European languages) use familiar email environments such as Microsoft Exchange or Lotus Notes, CJK countries have developed their own email platforms, with their own architecture and storage protocols. “Backwards and in high heels” barely suffices as a description of the potential complexity with which UBIC and its software have to deal.