we have integrated the Dictionary Tagger into our text mining workflow.
For us it is not clear how we have to configure the tag types and tag values exactly or what the individual differences would do.
Since we are working with German text, we have used the STTS as tag type. What would be the differences with English text?
As tag value we first used UNKNOWN, which made it look least wrong. What is it doing exactly and what´s are the effects of the other possibilities?
For the way how the node tags your text, it makes no difference which tag/value you select, as it’s only adding a name to the tag. It will tag your text according to the dictionary that you provide via the second input port and of course if you want to display the text afterward using a tagger viewer it makes sense to select something meaningful.
I agree, finding the right tag for what one is tagging can be tricky. What is the content of your dictionary? Is there a word you could summarize the terms that you want to tag in the text?
thank you for the explanation.
We are doing the text mining on the sections “tasks” and “candidate profile” of job descriptions. Therefore we are looking for given key words to make a list and analyze them f.e. count them in further steps. We created a table with the keywords. We used the Node Tagged Document Viewer and every row it shows “false”.
Our idea was so filter the nouns in the following Node Tag Filter which unfortunately wasn´t successful because it showed no result. It makes no difference if its before the Node Case Converter or not, where we changed to lower cases. Also using a different tag f.e.“ART” didn´t give any result.
Do you have a recommendation on how to proceed?