【作者】 塔娜；

【导师】 松云；

【作者基本信息】 内蒙古师范大学 ， 课程与教学论（计算机）， 2025， 硕士

【摘要】 随着信息技术的持续发展,以及国家积极推行教育信息化、加速教育现代化和数字化转型的战略逐步落实,教育行业正迎来深刻的变革与进步。信息技术的迅猛发展和信息化进程的持续深化,导致各类电子文本呈现出爆炸式增长的趋势。这些海量且多样的文本资源中蕴含着丰富的信息宝藏,如何借助计算机技术高效挖掘并提炼出有价值的信息,将其应用于教育教学实践,已成为提升教育信息化水平的关键需求。在边疆地区教育信息化进程中,教育资源的内容与本地区的地区化元素未能充分融合,且在某些方面还与当地经济社会发展需求有些脱节的现象。边疆地区的教育信息化不仅需要应对优化资源配置和提供个性化学习体验的普遍挑战,还需确保教育内容的相关性和实用性。因此,构建一个既能传承地区文化又能紧密对接教育需求的教育资源库,显得非常必要且具有深远意义。利用蒙古文信息处理技术可以挖掘和利用本地语言资源的教育价值、增强教育资源库的功能和服务效能,助力保护和发展少数民族的民俗文化,为学生提供更符合地区特色的学习材料和技术支持,从而增强教育的多样性和有效性。此外,在蒙古文信息处理技术的发展及其教育应用方面,也具有重要的推动价值。在此背景下,本文提出了基于蒙古文领域术语识别技术的蒙古文领域术语教育资源库系统,旨在从蒙古文领域文本中确定并挖掘有教育价值的术语。该系统通过智能化管理与处理,提供了包括术语知识管理、术语知识发现、术语知识查询等核心功能,以支持教育应用和发展,实现知识的结构化组织,促进民族地区教育资源的共享与扩展。本文具体工作如下:(1)进行蒙古文领域术语识别实验,提出蒙古文领域术语识别方法。收集来自蒙古文新闻领域的文本数据,并通过深入分析确定了三类具有教育价值的术语:政策名词术语、头衔名称术语及组织机构术语。基于这些分类,构建了包含35,697个实体标注的蒙古文领域术语识别语料库。考虑到蒙古文独特的构词特点,采用了基于双向长短时记忆网络与条件随机场结合的模型进行术语识别。实验结果显示,该模型在蒙古文术语识别任务中表现优异,验证了其有效性和可行性。为了实现术语数据的有效管理和利用,选择SQLite数据库作为术语存储的载体,经过人工验证过程,初步确认并收录了具有1,735条术语的蒙古文领域术语资源库。这项工作推动了蒙古文自然语言处理技术的发展也为后续研究奠定了坚实的基础。(2)运用前期提出的蒙古文领域术语识别方法,设计并实现了蒙古文领域术语教育资源库系统。首先,从技术支持和教育需求两个角度进行需求分析,确保系统的可行性与功能完备性。特别关注于从具有教育价值的蒙古文领域文本中准确识别术语,以支持教育教学实践。其次,设计合理的系统架构,细化各个功能模块。在系统的设计与搭建过程中,着重实现术语的自动识别、分类管理、智能标注以及查询检索等核心功能。该系统旨在为民族地区基础教育阶段的学生和教育工作者提供符合其文化背景的特色教育资源,通过多语种展示方式支持课堂教学、自主学习及资源共享,同时对中华优秀文化传承发扬具有积极的促进意义,也有利于优质教育资源的普及。为验证系统的有效性,将其投入小范围用户群体中试用,并从定性分析和定量分析两个维度对系统进行评价。结果表明,该系统对民族地区的教育资源信息化和教育教学工作具有积极的促进作用。更多还原

【Abstract】 With the continuous development of information technology and the gradual implementation of China’s strategy to actively promote educational informatization,accelerate educational modernization,and drive digital transformation,the education sector is undergoing profound reforms and advancements.The rapid evolution of information technology and the deepening informatization process have led to an explosive growth of various electronic texts.These massive and diverse textual resources contain rich informational treasures.How to effectively utilize computer technology to efficiently mine and extract valuable information from them,and apply this knowledge to educational practices,has become a critical requirement for enhancing the level of educational informatization.In the advancement of educational informatization in China’s border regions,there remains a disconnect between the content of educational resources and regional cultural elements,with certain aspects failing to align with local socioeconomic development needs.The informatization of education in these areas must not only address universal challenges such as optimizing resource allocation and delivering personalized learning experiences,but also ensure the relevance and practicality of educational content.Consequently,building an educational resource repository that preserves regional cultural heritage while closely addressing pedagogical demands is both imperative and of profound significance.By leveraging Mongolian language information processing technology,the educational value of local linguistic resources can be effectively harnessed,enhancing the functionality and service efficiency of the resource repository.This approach contributes to preserving and developing the folk cultures of ethnic minorities,while providing students with regionally tailored learning materials and technological support,thereby strengthening educational diversity and effectiveness.Furthermore,it holds substantial value for advancing Mongolian language processing technologies and their practical applications in education.Against this backdrop,this thesis proposes a Mongolian Domain Terminology Educational Resource Repository System based on Mongolian domain terminology recognition technology,aiming to identify and extract educationally valuable terms from Mongolian domain texts.Through intelligent management and processing,the system provides core functionalities including terminology knowledge management,terminology knowledge discovery,and terminology knowledge query,designed to support educational applications and development,achieve structured organization of knowledge,and promote the sharing and expansion of educational resources in ethnic minority regions.The specific work of this thesis is as follows:Firstly,conduct Mongolian domain terminology recognition experiments,this study proposes a Mongolian domain terminology identification method.Text data from the Mongolian news domain was collected,and through in-depth analysis,three categories of educationally valuable terms were identified:policy-related terms,title-related terms,and organization-related terms.Based on these categories,a Mongolian domain terminology recognition corpus containing 35,697 entity annotations was constructed.Considering the unique morphological characteristics of the Mongolian language,a Bidirectional Long Short-Term Memory model combined with Conditional Random Fields was adopted for terminology recognition.Experimental results demonstrate that the model achieves excellent performance in Mongolian terminology recognition tasks,validating its effectiveness and feasibility.To enable efficient management and utilization of terminology data,SQLite database was selected as the storage medium.Following a manual verification process,a Mongolian domain terminology resource repository containing1,735 validated terms was preliminarily established.This work advances the development of Mongolian natural language processing technologies and lays a solid foundation for subsequent research.Secondly,building upon the Mongolian domain terminology identification method proposed earlier,this study designs and implements a Mongolian Domain Terminology Educational Resource Repository System.First,requirement analysis was conducted from both technical support and educational demand perspectives to ensure the system’s feasibility and functional completeness.Special emphasis was placed on accurately identifying educationally valuable terms from Mongolian domain texts to support pedagogical practices.Second,a rational system architecture was designed,with detailed specifications for each functional module.During system development,core functionalities such as automatic terminology recognition,categorized management,intelligent annotation,and query retrieval were prioritized.The system aims to provide culturally contextualized educational resources for students and educators at the foundational education stage in ethnic minority regions.By adopting multilingual display modes,it supports classroom instruction,self-directed learning,and resource sharing.Furthermore,it actively contributes to preserving and promoting China’s outstanding cultural heritage while facilitating the dissemination of high-quality educational resources.To validate the system’s effectiveness,it was trialed within a small-scale user group and evaluated through qualitative and quantitative analyses.Results demonstrate that the system significantly enhances educational resource informatization and teaching practices in ethnic minority regions.更多还原