DMOOC 论文冲刺 - 确实激发创新

登录注册

白羽/ DMOOC 论文冲刺

项目简介

截止日期：12月23日

白羽 TO DMOOC 论文冲刺 | 项目讨论区

初稿

发帖时间：2016-12-25 23:57

更新时间：2016-12-25 23:59

第一次提交

( 1.093 MB) 白羽, 2016-12-25 23:56

回复︿（1）▪ 赞

白羽 8年前

摘要

Online learning platform, which have taken higher education by storm, are an opportunity to observe student behavior and knowledge.

The vast majority of educational data mining research is based on the online learning platform in Europe and America and within a small-scale data. In this work, we collected 17,854 code submissions in CS1 from Trustie, which is a famous domestic online education platform. Then, we perform a preliminary exploratory inspect for code quality by SonarQube from the original data. Numerical experiments suggest that, logical training is more important than grammar training and "code smell" will seriously affect the quality of the program. Our understanding of educational data mining may also shed light on improving teaching quality and deserves additional study.

赞

登录后可添加回复

白羽 TO DMOOC 论文冲刺 | 项目讨论区

论文题目、摘要与大纲

发帖时间：2016-12-14 03:00

更新时间：2016-12-14 08:17

Predictive compilation results from student code smells with ELM

ABSTRACT

Code smells are not usually bugs, but it refers to any symptom in the source code that indicate weaknesses in design.In other word, It is directly related to the stability and robustness of the program. Therefore, new programmer should pay attention to these potential risks in particular，because a good programming style (avoiding code smell as much as possible) will lay a solid foundation for advanced programming. Over the past decade, we have observed significant growth of work in students' program data mining. However, the detailed course data is so rare that the majority of the studies focus on simplistic statistical analysis and are conducted within a small scale data set (less than 10000). In this work, we collected 17,854 code submissions in CS1, and use an open source platform SonarQube automatically inspect the code quality from the raw data. We then proposed an Extreme Learning Machine (ELM) based method to predict compilation results from processed data and educational platform data. In order to verify the accuracy of the forecast, we use the BP network as a comparison. The results show that our method is more accurate.

I. INTRODUCTION

II. METHODS

A. BP NETWORK

B. ELM NETWORK

III. PREDICTION BASED ON ELM

A. Dataset （平台介绍，数据来源，数据规模）

B. Data Preprocessing (经过sonar处理的流程，主要code smell举例)

C. Evaluation Metrics (预测精确度定义)

D. Parameter Optimization (参数优化)

E. Results (结果与对比)

IV. CONCLUSIONS

REFERENCES

回复︿（5）▪ 赞（1）

尹刚 8年前

尹刚 8年前
另外，标题也有语法错误。

赞回复︿

白羽 8年前
好嘞，我再改改

赞（1）回复︿

另外，题目可以不用elm，使用deep learning更好吧？

赞

尹刚 8年前

@16BaiY(baiyu)‍ @毛新军(xjmao)‍ @王涛(wangtao)‍ @yuyue_whu@foxmail.com(yuyue_whu@foxmail.com)‍

赞

尹刚 8年前

很好！修改一下：

I. INTRODUCTION

II. RESEARCH PROBLEM

说明你要解决的教学问题，一般2-3个即可

这个问题的解决程度，回答，要在III、IV、V中反复呼应

III. METHOD SELECTION

A. BP NETWORK

    B. ELM NETWORK

这一节要结合具体案例说明，为何要选择ELM，BP和ELM的关系，优缺点等等。

IV. PREDICTION BASED ON ELM

    A. Dataset （平台介绍，数据来源，数据规模）

    B. Data Preprocessing (经过sonar处理的流程，主要code smell举例)

    C. Evaluation Metrics (预测精确度定义)

    D. Parameter Optimization (参数优化)

V. RESULTS (结果与对比)

VI. RELATED WORK

VII. CONCLUSIONS

REFERENCES

  请补充参考文献

赞