چکیده
مقدمه
چالش ها و اقدامات متقابل در برابر داده کاوی رفتاری دانش آموزان
قوانین انجمن کاوی داده های رفتاری دانش آموزان
الگوریتم قانون کاوی انجمن داده های رفتاری دانش آموزان و بهبود
مورد درخواست
نتیجه گیری
منابع
Abstract
Introduction
Challenges to and Countermeasures Against Student Behavioral Data Association Mining
Mining Association Rules of Student Behavioral Data
Association Rule Mining Algorithm of Student Behavioral Data and Improvement
Application Case
Conclusions
References
چکیده
با پیشرفت فناوری جمعآوری دادههای دانشگاهی هوشمند، دادههای رفتاری دانشآموزان در اندازه، تنوع و توان عملیاتی در زمان واقعی رشد میکنند و ظرفیت ذخیرهسازی و قدرت محاسباتی روشهای تحلیل دادههای رفتاری سنتی را با چالشهایی مواجه میکند. این مطالعه بر کاربرد قانون کاوی تداعی در تجزیه و تحلیل داده های رفتاری دانش آموزان متمرکز است. جمعآوری دادهها، ذخیرهسازی، محاسبات و تجزیه و تحلیل، همگی بخشهای جداییناپذیر یک معماری چهارلایه دادهکاوی را تشکیل میدهند، و فرآیند کاوی سه مرحلهای از «پیشپردازش دادهها» تا «یافتن قوانین ارتباط» تا «کسب دانش مرتبط» شرح داده شده است. الگوریتم استخراج موجود برای رسیدگی به مسائل اسکن بیش از حد مجموعه داده اصلی و تکرارهای اضافی به روز شده است. یافتههای حاصل از مطالعه موردی نشان میدهد که تعداد تکرارها در الگوریتم استخراج اصلاحشده به میزان زیادی کاهش مییابد، و به طور موثری کارایی استخراج مجموعه دادههای رفتاری عظیم دانشآموزان را بهبود میبخشد.
توجه! این متن ترجمه ماشینی بوده و توسط مترجمین ای ترجمه، ترجمه نشده است.
Abstract
With the advancement of intelligent campus data acquisition technology, student behavioral data are growing in size, variety, and real-time throughput, posing challenges to the storage capacity and computing power of traditional behavioral data analysis methods. The study focuses on the application of association rule mining in student behavioral data analysis. Data collection, storage, computation, and analysis all comprise integral parts of a four-layer data association mining architecture, and the three-step mining process from “data preprocessing” to “finding association rules” to “acquiring relevant knowledge” is described. The existing mining algorithm is updated to address the issues of overscanning of the original dataset and excess iterations. The findings from the case study reveal that the number of iterations in the modified mining algorithm is greatly lessened, effectively improving the mining efficiency of the massive student behavioral dataset.
Introduction
The mass data generated with the continuous development in Internet technology exhibit a discrete and isolated state. It is also difficult to deeply integrate and intelligently process them by computer due to the lack of semantics. Association knowledge represents the relationship between events. Analyzing and refining association knowledge can reveal some potential laws between real-world things and provide guidance for work practice. Association rule mining can implement semantic association between different data sources through data integration and achieve the purpose of comprehensive data sharing, making it convenient for users to further analyze and mine data to access valuable information, and providing effective data support for users' scientific decision making.
Conclusions
Association rule mining of student behavioral data is an important area of smart campus data analysis. It can expand the methods for smart campus data analysis and deepen the research on student management in the process of smart campus construction. In view of the shortcomings of traditional data analysis methods in the face of mass data analysis, the distributed parallel processing technology has been introduced into student behavioral data association rule mining to construct the framework of distributed student behavioral data association rule mining, the relevant process has been clarified, and the existing mining algorithms have been improved accordingly. The case analysis results show that, the improved mining algorithm has effectively boosted the data mining efficiency. Association rule mining can intuitively reflect the relationship between students’ behavioral factors and further analyze the student management knowledge contained in the data, thereby providing an effective basis for campus managers to make sound decisions. In addition, the improved association rule mining algorithm has been applied to different datasets to verify its effectiveness.