Luận văn Thạc sĩ: Xây dựng hệ thống truy hồi học liệu cho sinh viên ngành Điện - Điện tử

The continuous advancement of information technology has significantly reshaped the landscape of data and document retrieval. In the current educational environment, particularly with the widespread adoption of online learning, the demand for precise and reliable academic resources for students has intensified. Traditional search methods often yield a vast but uncurated collection of information, making it challenging for students to discern accurate and relevant learning materials. This thesis addresses this critical need by proposing the development of an information retrieval system specifically designed for students in the electrical and electronics major. The primary goal is to enhance the efficiency and accuracy of accessing learning materials, ensuring students can quickly find trustworthy academic content relevant to their studies.

Students and educators in the electrical and electronics engineering field, particularly those requiring efficient access to specialized learning materials.

This academic thesis details the design and implementation of an advanced information retrieval system tailored specifically for students in the electrical and electronics major. Recognizing the challenges faced by students in navigating vast and often unstructured online resources, especially during periods of increased online learning, the research aims to provide a reliable and efficient solution for accessing academic learning materials. The methodology begins with an in-depth exploration of core information retrieval concepts, covering the stages of text preprocessing, indexing, and subsequent retrieval. A significant component involves the application of natural language processing techniques, such as tokenization and stop word removal, to accurately prepare learning documents for indexing. The system leverages the open-source search library Lucene for robust text indexing, which transforms raw documents into a searchable Document Index based on relevant terms and their weights. Furthermore, the study investigates text classification using machine learning techniques to categorize documents by type and topic, enhancing search precision. The retrieval phase incorporates Boolean models and ranked document presentation, allowing users to efficiently query and receive highly relevant results from a curated collection of electrical and electronics learning materials. Performance is evaluated using standard metrics like Precision and Recall. The practical implication of this system is its ability to significantly improve access to accurate and specialized academic content, thereby supporting effective study and research for students in the electrical and electronics field, and mitigating the inefficiencies of general web searches.