Research And Implementation Of Block-python Programming System

With the advent of the Internet era and the constant update of educational concepts,programming education for children has become a research hotspot in the world.Institutions and scholars in more and more countries have begun to explore programming education for teenagers,especially for the K-12 population.Children’s programming software based on educational purposes emerged,and graphic programming technology gradually became popular,which began to be involved in teaching curriculum by more and more schools.However,there is still a gap between graphical programming and code programming.Aiming at solving this problem,this paper designed and developed a Block-Python programming system,which is based on the open source framework of Scratch.This system not only provides a block-text programming environment and debugging environment,but also provides the code evaluation and analysis function,providing a more comprehensive and systematic learning environment of Block-Python for learners.On the basis of in-depth analysis of block programming concept and code evaluation technology,this paper first designs the overall architecture of Block Python programming system,including programming subsystem and evaluation subsystem,providing block-based programming and evaluation functions respectively.Then,aiming at the key problems in the evaluation of block-style Python programming works,this paper studies the computational thinking(CT)ability and the readability of the code from two dimensions:1)Proposed the evaluation metrics for Block-Python works based on computational thinking.Through reconstructing CT evaluation elements and the calculation rules,taking the capacity of the program complexity and structural complexity into consideration,which makes evaluation metrics more suitable for Python code;2)Proposed a Python code readability evaluation model combining multi-class features.A feature set covering three types of features is proposed,and an iterative training based on XGBoost is designed to automatically screen the best feature set and train the classification model,which improves the classification accuracy.Finally,the design and implementation of the system and the test results are introduced.The system that this topic completes is efficient and easy to use,and the function is perfect.On the one hand,the programming subsystem enables beginners to learn Python programming in the way of graphical programming,gradually develop programming skills through the automatic conversion from block programming to language programming.On the other hand,the evaluation subsystem implements the evaluation metrics that more suitable for Python teaching,which can not only evaluate the computational thinking abilities in the user’s works more accurately,but also detect the user’s programming habits and give accurate evaluation of code readability.After testing,the system runs well,meeting the design expectations and has certain stability and robustness,which is a complete and efficient Block-Python learning system.