Dear all,
I am dealing with bilingual data and I developed an annotation scheme whereby the transcription is on a Parent Tier, and language tagging and POS tagging for each token are in two different Dependent Tiers. At the end of the transcription and annotation process, I would like to find a way to see, for instance, how many and which adverbs are expressed in language X in my data and how many and which adverbs are expressed in language Y. How could I do that?
I thank you in advance for your help.
Dear Marco,
I think you can use the following function available in the Search menu: Search->Structured Search Multiple eaf.... The appearing window has 3 tabs, select the tab Multiple Layer Search. Here you can define a query across tiers. Based on your description I would say you only need one column of query fields and three layers (rows). For the top layer you could select the token tier, for the second layer the POS tier and for the third the language tier. You probably need to use some regular expressions and run the query once for language X and once for Y, depending on the desired output. The query results can be shown in and exported in Frequency view.
Best,
Han
Dear Han,
Thank you very much for your swift reply. It is vey useful and corresponds more or less to what I came about with. However, I can’t understand which regular expression should I use. I think that in the second layer I can select the corresponding language label and in the third layer the corresponding POS label. But in the top layer, how could I select each possible token, so as to get all tokens that match the other two conditions (i.e., language X and POS X)?
Thank you very much and sorry for my question.
I guess the regular expression .+ should work fine (assuming you don’t have empty tokens that do have a POS and language set). In the Mode dropdown box you have to make sure that regular expression is selected. Don’t hesitate to follow up if this does not render the results you’re after!