Research And Implementation Of Online Programming Platform Dataset

With the development of the times,programming education for children has received more and more attention in the field of education.Scratch and Python have been widely used as programming languages for primary and secondary school students in the field of progranuning education.As a result,there have been many online programming platforms for children.However,these platforms generally do not provide public datasets.At the same time,with the widespread application of machine learning and artificial intelligence,more and more researchers have begun to use big data analysis technology to study the programming learning situation of students in the field of children's programming education to help students improve their programming skills.However,the lack of datasets limits its further research and analysis.Therefore,it is necessary to build a multi-dimensional dataset in the field of online programming,and at the same time open up fine-grained data API retrieval services.This will provide researchers with a wealth of basic data in the field of computer education research,which has a very wide Application prospects.In response to the above problems,this paper first designs an architecture of online programming platform dataset,it is divided into three parts:data collection,data processing and data providing services.Then in the data processing part,a multi-label Scratch classification algorithm based on label correlation is designed to solve the problem of inaccurate category data of Scratch projects.A multi-label Scratch classification model is constructed.Furthermore,a multi-dimensional data model is proposed,which contains data in four dimensions:user model,work model,behavior model,and domain model.Finally,the construction of a multi-dimensional dataset is completed.Finally,based on the proposed architecture and the actual needs of users,a prototype system for data set construction and data service provision of the online programming platform is designed and implemented.The experimental results show that the multi-label Scratch classification algorithm based on label correlation proposed in this paper is superior to multi-label classification algorithms such as RAKEL in terms of classification performance and time performance.The constructed multi-label Scratch classification model effectively guarantees the data accuracy of the multi-dimensional dataset.The multi-dimensional data model proposed in this paper promotes the data diversity of multi-dimensional dataset.The prototype system designed and implemented in this paper,the functional test verified the correctness of the system,has a high level of performance,and can provide flexible and efficient data API retrieval services.