November 30, 2016 at 2:19 pm #68871
I am working on testing Nodegoat on behalf of Brill for a reference work on Greek inscriptions. Types include the inscriptions themselves, entries on these inscriptions and publications on on either the inscriptions or the entries.
Each entry on an inscription includes a number of index terms which can be classified as one of several categories (Proper Names, Kings, Emperors, Geography, Religion, Military terms, Greek and Latin Words, and Selected Topics). Ideally, I would like each entry to refer to the relevant index term, and from there be able to see the other entries in which this index term also appears.
However, I don’t know how to express this in my design exactly. I have Index terms as a classification now, but this doesn’t seem quite right. Would making a separate Type be easier/more helpful? And if so, how would I ensure the different categories are still distinguished?
I hope I’m being clear (and not overlooking something very obvious)! Thanks in advance.December 1, 2016 at 2:35 pm #68872
Yes, this is certainly possible. What you first need to do is start working with individual index terms. You will need to make a classification in which each object is one index term, this will allow you to create a list of all the terms that are used and use this to explore connecting terms (or make visualisations of networks of terms).
To link these terms to your inscriptions or entries, you create an object description in each type that links to this index term classification. You can allow for multiple index terms to be linked in this object description by checking the first tick-box of the additional options of these object descriptions.
To classify these index terms, you just create one object description in the classification index term that links to another classification (e.g. ‘Kinds of index terms’) which contains the nine categories you mentioned (Proper Names, Kings, Emperors, Geography, Religion, Military terms, Greek and Latin Words, and Selected Topics).
This should allow you to construct a network of interlinked and classified index terms.
Does this help you?December 6, 2016 at 1:25 pm #68874
Yes, it works now. I had not thought of creating two separate classifications but now it makes much more sense.
Thanks a lot!January 10, 2017 at 10:02 am #68886
I have one more question with regards to the importing of data. Matching the CSV-files to the design works really well (and is much easier than I feared!). However, the data in one type that I have linked to another type (for example, I’ve connected the object description ‘Editor’ in the type ‘Publication’ to the type ‘Editor’) does not show up after importing; the boxes where the information should be remain blank. It only seems to work when I change every object in my design to ‘String’–but of course, this means I lose all the relations between Types that enable a useful visualisation. Could you give me a hint as to how to import all the data with my design remaining intact? (Or does this mean by design must be flawed in some way?)
Thanks again!January 17, 2017 at 8:58 pm #68891
Yes, you should be able to import this data without needing to change your data model. Because you are importing relational data, it might be best for you to do the import in multiple steps. First you would import all the editors (based on a file with all the relevant/unique editors, preferable with an ID). Next you can link the publications to these editors while importing by using the IDs that refer to the editors and import this into the relational object description ‘Editor’ in the type ‘Publication’.
You must be logged in to reply to this topic.