VNLP: Visible natural language processing

In general, Natural Language Processing (NLP) algorithms exhibit black-box behavior. Users input text and output are provided with no explanation of how the results obtained. order to increase understanding trust, users value transparent processing which may explain derived enable underlying routines. Many approaches take an opaque approach by default when designing NLP tools do not incorporate a means steer manipulate intermediate steps. We present interactive, customizable, visual framework that enables observe participate in pipeline processes, explicitly parameters each step, explore result visually based on user preferences. The visible (VNLP) design is then applied similarity application demonstrate utility advantages supporting understand justify both process results. also report feedback our from modern languages expert.