基于马尔科夫模型的聋生阅读输入分析  

Analysis of Deaf Students' Reading Input Based on Markov Model

作　　者：姚茂建 李晗静 吕会华[2] Yao Maojian;Li Hanjing;Lyu Huihua(Beijing Key Laboratory of information Service Engineering,Beijing Union University,Beijing 100101,China;College of Special Education,Beijing Union University,Beijing 100075,China)

机构地区：[1]北京联合大学北京市信息服务工程重点实验室,北京100101 [2]北京联合大学特殊教育学院,北京100075

出　　处：《北京联合大学学报》2018年第3期86-92,共7页Journal of Beijing Union University

基　　金：国家语委重点项目(ZDI135-31);北京教育科学规划重点课题(ADA14121)

摘　　要：以不同类型文本阅读输入角度出发,分析聋生的阅读输入规律,以反映聋生阅读时的表现。通过聋生阅读输入的文本与中文分词标准匹配率在不同等级的转移情况建立马尔科夫过程,预测聋生的阅读输入结果。通过阅读实验以验证马尔科夫模型预测的结果,实验结果显示,聋生更习惯于阅读较小的组合词块,单音节词、双音节词、三字及四字多音节词对聋生阅读有较大影响,这与马尔科夫模型预测结果比较符合。This article starts with different types of text reading input and analyzes the reading input rules of deaf students to reflect the performance of deaf reading. The Markov process is established through the transfer of the matching rate of deaf students＇ reading input and the Chinese word segmentation standard at different levels,and finally predicts the reading input of the deaf student. Through reading experiments to verify the results of Markov model predictions,the experimental results show that deaf students are more accustomed to reading smaller combined word blocks,and monosyllabic words,two-syllable words,three-syllable and four-syllable words have a greater impact on deaf reading. This is in line with the Markov model predictions.

关 键 词：聋生 自然输入标注 阅读输入 马尔科夫模型 

分 类 号：G762[文化科学—特殊教育学]