Feedback from Bob Boelhouwer

This is feedback from Bob Boelhouwer Bob.Boelhouwer@ivdnt.org, a computational linguist at Instituut voor de Nederlandse Taal in Leiden, Netherlands. Bob didnât want to go through the motions of signing up for this mailing list etc. but he has given me permission to submit his feedback on his behalf and he has agreed to the OASIS Feedback License.

I found the description of DMLex very interesting. It would be good if it finds some uptake.
Below you will find some remarks, of which Iâm not sure if they are useful.

On the organisation of entries.

A distinctive feature of DMLex is that all entries are at the same level. There are no subentries. The main argument for this is that it is computational more efficient. But, Iâm not sure of for modern computational equipment this is a significant advantage. An argument for using embedding structures is that editors working on compiling a dictionary would prefer to have an overview of all the submeanings and collocations that can make up an entry. Moreover, they will have a better indication of how what the entry will look like when presented to the public.

The argument that it more valid in the case of MWEâs. Indeed it seems more logical to treat these as separate entries with links to the constituent words.

On hierarchical vs. complex graphs

Indeed has the data in a dictionary more complex relations than can be expressed by a strict hierarchical format like XML and JSON. If the data contains other relations that means extra programming.

However, using XML has some very great advantages.
The advantage of using a standard formatting language like XML is that there are a large number of tools available for important manipulation functions like editing and querying but also for data exchange. It is important that human editors have tools at their disposal that offer constraint guidance as to enforce structural constraints on the content, and methods to query the data. Of course, there are many existing graph formats and processing tools, but they are rather limited to niche applications. It would be great if there exists a graph-based data format that is well supported by tool-developers.

Another advantage of XML, which I think added to its success, is that the data is readable. You can have direct access to the data without depending on tools. That is different for graph languages like RDF which is notoriously difficult to read.

lexidma-comment message