The project is devoted to the modeling of the phenomenon of internal question processing (hereafter: IQP) by means of erotetic logic (i.e. the logic of questions). The main objective is the development of tools which are necessary in the formal modeling of IQP.
Roughly, IQP shows up when a question is not succeeded by an answer, but by a new question, or by a strategy of reducing it into auxiliary questions.
The initial tools will be provided by Inferential Erotetic Logic (IEL), elaborated on in earlier publication of the applicant. In order to accomplish the task IEL has to be developed and, possibly, revised to some extent.
The phenomenon we are interested in occurs when an agent is supposed to answer a question or solve a problem, but he/she cannot accomplish the task by means of informational resources which are directly accessible to him/her. In such a situation it happens that an initial question is internally processed by an agent, where the outcome is either a new question concerning the subject matter or a tentative strategy of reducing the initial question into auxiliary questions. In both cases inferences leading from questions to questions play a substantial role. When answer(s) to question(s) raised that way (at least some of them) are still inaccessible to an agent, the process goes further in an analogous way, possibly on the basis of data just collected.
There is no logical theory of IQP. On the other hand, the ability of performing IQP is one of the key features of human intelligence. IQP is involved in such fundamental cognitive activities as problem solving (proof-search included), question answering and communication between agents. A formal representation of IQP may result in a progress in AI research.
Although there is no logical theory of IQP, we have in our disposal a logical theory of inferences in which questions play the role of conclusions, namely, IEL. The results obtained by researchers working within the IEL paradigm indicate that at least some aspects of IQP can be modeled in terms of tools either taken from IEL, or definable in it.
The basic areas of applicability of (partial) results include: the search of information in distributed resources, question answering (in particular, cooperative answering), problem solving (in particular, problem solving by interrogation), proof theory and automated deduction (proof search, complexity issues).
The project Erotetic logic in the modeling of ultimate and distributed question processing. Theoretical foundations and applications is supported by funds of the National Science Council, Poland (DEC-2012/04/A/HS1/00715).
Andrzej Wiśniewski (Coordinator)