Exploration Of Eye Movement Model And Interactive Application Of Chinese College Students In Reading English Web Pages

In order to help Chinese college students to read Chinese web pages fluently,this research studies the regularity of natural eye movement of readers and develops an browser extension that can interact with natural eye movement.Based on the regularity of natural eye movements,this browser extension can found out unfamiliar words of reader automatically,and provide essential interpretation naturally,enabling college students to immerse themselves deeply in the cognitive processing of articles,without distractions for translation operations.In order to summarize the college students' processing models of target words,and build a set of mathematical models to reflect the influence of relevant factors on the gaze time of target words,and calculate the suitable intervention opportunity of translation,and mine the potential needs of college students in using translation,this research combines the research results and eye movement research results in the field of English reading at home and abroad to design an eye movement pattern inquiry experiment which including 34 subjects.The experiment requires each subject to read one low difficulty and two difficult English articles,and allows the subject to use the word translation function on one of the difficult articles.The difficult words filtrate by the Youdao page translation 2.0 were selected as the target words.A Tobii Eye Tracker 4C and a browser extension are used to capture and record the reader's eye movement data and translation behavior data.The results of data analysing of 30 subjects show that 96.06% of the subjects' inside-sentence regression for the target word occur within 5 consecutive word reading.According to the number of regression of inside-sentence regression,the processing modes of the target word are divided into single,double and multiple.The gaze time of the word is considered as the dependent variable,and the quantifiable factors,including corpora and individual ability,are set as independent variables,a set of classification models for distinguishing unfamiliar words form familiar words of readers is built.The unfamiliar-wordpredictive level of the classification model reaches 96.69% in full article.Based on the classification models,the needs and translation behavioral data of college students,an Chrome browser extension with Non-Command interface is designed and developed.It was named Chinen Translation.The usability testing results show that the unfamiliar-word predictive level of the classification models reaches 97.63% in full article.The versatility of the classification model is verified with variance analysis.Compared with the same type of application,Chinen's translation enables participants to achieve better reading comprehension in a shorter time.In the PSSUQ standard questionnaire,the score of information quality and overall quality of Chinen's translation were significantly higher than the same type of application.The eye control,which is based on the regularity of natural eye movements,is divided into two levels in this research.The first is the primary eye control interaction based on the physiological level.The second is the advanced eye control based on the physiological and psychological level.At the end of the research,10 proposals were proposed for eye control system design.