Article Preview
TopIntroduction
The English teaching activities include five important components: listening, speaking, reading, writing and translation. For the second language students (ESL), reading is one of the most important parts. Previous research shows that 58.4% of students believe that the most influential factor in reading interest is the reading article itself. Among the factors such as teaching articles, teaching methods, teachers' quality, and teaching environment, 82.3% of students think that the tediousness of reading articles is the most unbearable (Lanhui, 2001).
Due to the great distinction of reading requirements and language foundations of each student, it is unreasonable to require every student to read the same text. For example, some students prefer philosophical articles, while others prefer economics. Therefore, it turns out that the dynamic changes in reading interests are difficult to handle. We analyzed the English reading preferences of sophomore students in an English course at one of our art colleges. Through a questionnaire, 81% of the students participated in our survey. Figure 1 shows the distribution of a student’s interests in reading topics during eight weeks. It can be observed that this student prefers criticism in the first week, politics and education in the 4th week, and literature in the 8th week. Also, some students have a relatively solid foundation in English when they chose the course.
Figure 1. Distribution of reading interests of a college student in eight weeks
Therefore, tailoring the most suitable reading articles for each student is an important factor in promoting reading ability. Designing suitable article choices for students is mainly based on two dimensions: students' interests and reading ability. If the reading articles can stimulate students' reading motivation and arouse interest in reading, the students' reading levels will be improved. Additionally, the difficulty level of the assigned reading article is also critical. If the reading article is too difficult, it is easy to bring excessive cognitive load to students, which often leads to negative emotions such as depression. If the reading article is too easy for the student, it’s unlikely to improve the students' reading ability, which is not only a waste of students' time, but also pointless. For each individual student, the reading assignment should be the appropriate difficulty according to their English level, that is, neither too difficult nor too easy. Through a step-by-step approach, students can gradually improve their reading ability.
To achieve this goal, a solution is for teachers to customize an exclusive reading study plan according to each students' personality, supplemented by corresponding reading articles of different ability levels. However, for English teaching, teachers usually have hundreds or even thousands of students, so it is difficult to conceptualize of such a solution given what an individual human is capable of. Therefore, it is necessary to design an automatic and personalized English reading recommendation system, which can recommend the most suitable reading articles for students based on their historical data. Previous researchers used recommendation systems to help students find the most satisfying reading articles, and such systems generally fall into two categories: content-based filtering and collaborative filtering (He et al., 2017). In a content-based filtering recommendation system, the content of the reading article and the user's portrait are represented as feature vectors by natural language processing techniques. Rana and Deeba (2019) use collaborative filtering technology to recommend books that are most relevant to user profiles, which are created based on users' reading behaviors and feedback. Wang and Liu (2023) designed and implemented a feasible news recommendation system, and a news recommendation model based on user collaborative filtering algorithm is proposed.