A computer assistant for learning UNIX commands-synthesiging candidate commands from specifications using agent models as the intermediary

Bibliographic Information

Other Title
  • UNIXコマンドの学習を支援するシステム : エージェントモデルからのコマンド列生成とその検証

Search this article

Description

The paper is an interim report on our efforts in constructing a computer assistant that helps novice but ambitious learners of UNIX commands to acquire the ability to specify the task and compose UNIX commands accomplishing the task.For that purpose we propose to use our agent models as the intermediary.To be specific, the system lets the user input the initial and final states of the task to be carried out as far as he recongizes them.The system collects such agents as will attain the final states.Through arranging them as well as inserting between them other agents necessitated to realize the preconditions of the agents employed. The idea comes from the triangle table used in action planning. Nextly essentially the same program as is used in parsing command sequences into agent sequence can be run to generate candidate command sequence implying the given agent sequences.The failure of generation serves to filter out the infeasible agent sequences At any phase of the process the learner is requested to interact with the system via selection and simulation in order to amend,complement and refine his specification.

Journal

Keywords

Details 詳細情報について

  • CRID
    1573387452167007872
  • NII Article ID
    110003192401
  • NII Book ID
    AN10013163
  • ISSN
    09135685
  • Text Lang
    ja
  • Data Source
    • CiNii Articles

Report a problem

Back to top