Quantum computations for disambiguation and question answering

Abstract Automatic text processing is now a mature discipline in computer science, and so attempts at advancements using quantum computation have emerged as the new frontier, often under term of natural language processing. The main challenges consist finding most adequate ways encoding words their interactions on computer, considering hardware constraints, well building algorithms that take advantage architectures, to show improvement performance tasks. In this paper, we introduce framework starts from grammar can be interpreted by means tensor contraction, build word representations states serve input algorithm. We start introducing an operator measurement contract words, resulting representation larger fragments text. then go develop pipelines for tasks sentence meaning disambiguation question answering features. For first task, our contraction scheme deals with syntactically ambiguous phrases storing various different meanings superposition, solution not available classical setting. second obtain contains all possible answers equal implement Grover’s search algorithm find correct answer, agnostic specific question, implementation potential delivering result quadratic speedup.