Is a practical dissertation on the topics Information retrieval, text mining, information extraction and others like it defendable?

Question

IT education

Essence: There is knowledge at the level of nlp.stanford.edu/IR-book in these areas. There are practical (pretty good) developments using well-known open solutions (full-text search, fact extraction, semantic network building, and something else).
It is possible to spend up to 8-10 hours a day on work. There is a protracted postgraduate study of the author on a topic that is extremely far from these topics, and which, most likely, will have to be completed unprotected - it takes a lot of time, but it is still very far from any results. And there is a persistent desire to still get a Ph.D.
Therefore, I ask knowledgeable people to suggest whether it is possible to defend "practical" work in these areas? I don’t want to get into the theory of processing languages, construction languages, and so on and so forth. If it is possible and you are ready to suggest topics \ institutions \ specialists, I will be extremely grateful to you.