Strategies for Improving College English Reading Comprehension Based on Text Mining Technology

English reading is an important way for people to acquire foreign culture and information. Therefore, English reading teaching occupies a rather important position in college English teaching, and its basic strategy is to emphasize the instrumental nature of language and cultivate students’ ability to acquire information through reading [1–3]. The cultivation of this ability requires first of all the mastery of some basic reading skills, such as finding specific information quickly, skimming to get a general impression, prediction, overview and so on. In addition, since reading in English is an active process of comprehending and accepting information, to read effectively, it is more important to strengthen the cultivation of reading ability in terms of word meaning, chapter structure and cross-cultural knowledge [4–8]. In college English teaching, English reading teaching has always been a very important part. The University English Syllabus points out that the purpose of university English teaching is to train students to have strong reading ability, certain listening ability, preliminary writing and speaking ability, so that students can use English as a tool to obtain the information they need for their specialties [9–12]. Therefore, it is of great practical significance that the university English syllabus puts the cultivation of students with strong reading ability at the top of the teaching objectives [13–15]. Text mining technology is an algorithm for extracting meaningful information from unstructured text data, which is able to obtain potential valuable information in the text that cannot be obtained in the classical structured data format. The use of text mining technology in the context of big data can quickly, efficiently and intelligently realize the mining of the characteristics of the reading needs of college English, which can be a reference for the cultivation of college English reading comprehension ability [16–20].

The evolution of educational concepts and the leap of technology have caused the traditional college English reading education to undergo profound changes, and how to efficiently enhance students’ reading skills is a major issue facing college English education. This paper establishes a model for teaching college English reading with the support of smart classrooms, and implements it in three stages: before, during, and after class. In view of the current lack of university English reading teaching resources, this paper establishes a reading preference model based on students’ learning behaviors, and conducts reading preference resource mining through the TF-IDF algorithm, and establishes an English reading teaching resource recommendation algorithm by combining the similarity of the knowledge structure of students’ reading resource preferences. In order to analyze the feasibility of the university English reading teaching model to improve students’ reading comprehension ability, students from the School of Foreign Languages of a university were taken as the research sample, and a teaching comparison experiment was set up for data analysis. Based on the results of the analysis, the strategy for improving college English reading comprehension ability is proposed from three perspectives: technology optimization, teaching mode improvement, and creating a digital resource library.

2

Benefits and Resistance of Cultivating Reading Comprehension Skills in College English

There are many problems in the current university English reading teaching, such as the lack of relevance of traditional teaching methods, the lack of diversity of reading materials, the insufficient cultivation of students’ reading interests and habits, and the lack of reading initiative, which leads to poor results in English reading teaching and fails to really improve students’ English reading ability. Since there are big differences in students’ English reading interest and reading ability, it is a big challenge to choose English reading teaching strategies that can ensure students’ reading interest and improve their reading ability, so that every student can benefit from reading.

2.1

Benefits of English Reading Skills Development

1)

Enhancing Comprehensive Language Application Ability

Under the guidance of modern information technology, university English reading teaching has undergone a remarkable change, which is aimed at comprehensively enhancing students’ comprehensive language application ability. By reading authentic English materials, students can not only enhance their profound mastery of vocabulary and grammar, but more importantly, they can flexibly use this language knowledge in real contexts, realizing the perfect combination of theory and practice. This teaching methodology not only encourages students to comprehend and utilize the language thoroughly, but also enhances their English expression in real-life communication. After such training, students are able to communicate more confidently and accurately in English when facing real English communication scenarios, which undoubtedly lays a solid foundation for their future international communication.

2)

Leading students to carry out independent reading

In the traditional English reading teaching mode, classroom teaching is mainly dominated by the teacher, who instills students with a single knowledge of English, so the students’ subjectivity is not fully reflected and their motivation for learning is not high. In modern English reading teaching, teachers pay attention to the improvement of students’ thinking ability, which helps to highlight the main position of students, so that students actively participate in English reading learning, constantly raise questions, analyze problems through a variety of ways, and ultimately achieve the purpose of solving problems. In the process of mobilizing students’ thinking, teachers lead them to read independently, so that they can master more reading experience and methods, so that they can truly become masters of English reading.

3)

Enhance students’ cross-cultural communication ability

Optimizing the ways and methods of college English reading teaching and improving students’ English reading abilities are essential to enhance their cultural awareness and cross-cultural communication skills. English reading materials can enable students to gain an in-depth understanding of the diverse cultural heritage, historical lineage, and social landscape of countries and regions around the world. This will not only broaden students’ knowledge horizons, but also deepen their understanding of and respect for cultural diversity, thus fostering a tolerant and open mindset in the context of globalization. In addition, by reading in English, teachers can help students to gain a deeper understanding of Western values and social institutions, so that they can not only gain a deeper understanding of historical origins and their significance, but also promote the understanding and acceptance of diverse social structures and cultural environments, so reading in English is an important way for students to develop a multicultural perspective.

2.2

Resistance to Cultivate English Reading Skills

Reading is one of the most frequent activities in learning a foreign language, and a lot of foreign language reading can improve the level of this language skill, and the same is true for learning English, if you want to improve your English level, you must often have the input of English-related knowledge in order to speak pure English. In order to improve English, you need to have a knowledge base related to English in order to speak pure English. Including improving English listening and writing, it is indispensable to accumulate reading in a gradual and orderly manner. Therefore, reading comprehension is especially important in college English teaching. However, there are still some inevitable problems in the current college English reading teaching, which are mainly manifested in the following aspects: 1)

Limited teaching resources. Traditional English teaching materials tend to be conservative in content selection due to the need to consider the average level of students nationwide and the requirements of education policy, resulting in relatively fixed content and form of English reading materials in the textbooks, which cannot meet the learning needs of students at different levels, and the richness of teaching resources is insufficient. Teachers usually need to spend a lot of time to screen and integrate English learning resources on the Internet when preparing lessons, while some teachers rely on traditional teaching resources in actual teaching due to time constraints, thus failing to provide students with rich and diversified reading materials.

2)

Single teaching mode. In the traditional college English reading teaching mode, teachers are the dominant players in the classroom, and students passively accept the knowledge. The teaching mode lacks effective interaction, which can’t fully attract students’ attention. At the same time, teachers in the teaching process pay more attention to vocabulary and grammar and other basic knowledge, when explaining the reading text, just digging the grammatical content, insufficient attention to the students’ text comprehension, unable to form an effective reading strategy.

3

The Development of Resources and Teaching Models for College English Reading Teaching

Under the background of the development of global cultural deep fusion, new challenges have been put forward to students’ English reading comprehension ability and university English reading teaching, which is a good opportunity to promote the improvement of the quality of university English reading teaching. For this reason, university English should pay attention to summarize the traditional reading teaching experience and combine with information technology to realize the optimization of university English reading teaching mode. It is necessary to create a professional and scientific college English reading teaching classroom for students, so that students can fully feel the effectiveness of the combination of technology and teaching, steadily improve their English reading learning ability, and develop English subject literacy.

3.1

English Reading Teaching Model Development

3.1.1

Framework for a model of reading instruction

Giving full play to the support and service functions of modern information technology for teaching and learning in English courses, teachers are encouraged to make rational use of and innovate in the use of digital technology and online teaching platforms to carry out online and offline integrated teaching, to provide support for meeting the personalized learning needs of students, and to promote the balanced development of compulsory education. Smart classroom refers to the rational use of information technology by teachers to build a classroom that helps to broaden students’ learning horizons and activate students’ thinking, which is characterized by intuition, effectiveness and vividness, and promotes the innovative reform and development of curriculum teaching. Therefore, in college English reading teaching, teachers can flexibly apply the concept of smart classroom teaching to improve the links of pre-course preparation, in-class teaching and post-course review, creating a smart reading teaching environment and laying a solid foundation for the innovation and systematization of English reading teaching.

As shown in Figure 1, the English reading teaching model based on the smart classroom uses the Internet as a platform to engage in equal dialogue and communication with students under the guidance of the “output-driven” and “input-facilitated” “output-oriented method”, and moves towards a free and open interactive state [21]. The whole English reading teaching task is accomplished through the combination of online and offline. Teachers assign tasks to students in a reasonable and scientific manner before, during, and after class, and students are able to complete the tasks assigned by teachers seriously and carefully, and discover their own deficiencies. The whole flow chart is based on students and teachers as the main driving force, and the whole Huixue platform as the link to complete the whole teaching task, through the organic unity of the two, students learn more effectively and teachers teach more efficiently.

3.1.2

Implementation of pedagogical model application

In this paper, we utilize the smart classroom to implement the teaching of college English reading class, which is divided into three stages to be implemented before class, during class, and after class. 1)

Pre-course stage. Before the text teaching, the teacher, through the microclass recording function of the reading software, puts the microclasses and tasks and vocabulary preview tasks of the related materials on the wing class online in advance for the students to make effective previews.

2)

In-class phase. In college English reading classroom teaching, there are generally three major stages: Pre-reading, While-reading and Post-reading. By applying the reading software and the wing class network to the English reading classroom, students can further realize the mastery of the text content, and understand the knowledge of the reading text and the general idea in a more vivid way.

3)

Post-class phase. After class, the teacher assigns personalized and graded homework with the use of Yicai.com according to the situation and results of classroom teaching. For students who have the ability to learn, they can provide improvement and extension training. For students with learning difficulties, they can continue to watch the micro-lessons recorded in real time in class after going home and choose slightly simpler exercises to consolidate. Through targeted, personalized homework assignments, students are able to combine their personal realities with their own needs for targeted learning, achieving the maximum effect of quality learning.

3.2

Recommended Resources for Teaching English Reading

The richness or otherwise of English reading teaching resources will directly affect the effectiveness of reading instruction. In order to design an effective English reading teaching mode, teachers should develop teaching resources with the help of information technology platforms and tools. In the actual reading teaching, in order to give full play to the function of information technology, teachers obtain rich teaching resources through the mining of students’ reading preferences, and then realize the diversification of English reading teaching resources, so as to meet the students’ personalized English reading needs.

3.2.1

Student reading preference acquisition

1)

Students’ reading preference acquisition

The resources chosen by students in the process of online learning have obvious preferences, so website analysis tools are used to record the pages visited, learning resources browsed, videos watched, and interactive behaviors of students. By analyzing the above data, students’ English reading preferences and interests are understood, laying the foundation for the next step of personalized learning resources recommendation. The main settings in English reading acquisition include students’ basic information, learning styles, reading styles, interactive behaviors, and so on. According to the above options, learning resources with higher similarity are selected, and students’ English reading preference model is designed by combining students’ auxiliary reading style information as shown in Figure 2.

According to the model of students’ English reading preferences, it is possible to gain insights into individual student differences and preferences. Data on students’ English reading preferences are collected, including their learning time, learning styles, preferred types of learning resources, etc., so as to obtain feedback and evaluation on each knowledge point. After obtaining students’ feedback and evaluations, these data can be analyzed to determine the difficulty level of each knowledge point.

2)

Classification of students’ reading preferences

Word Frequency-Inverse Document Frequency (TF-IDF) is a common technique used in information retrieval and text mining to measure the importance of a word in a text [22]. The main idea of the method is that if a word occurs more frequently in a text and the text containing that word occurs less frequently in the whole collection of documents, it means that the word is more important. When the method is applied to describe the preference of students u for learning resource categorical attribute a, TF_uz indicates the frequency of occurrence of categorical attribute a in learning resources in which students u have produced behaviors, which is calculated by the following formula: 1 $T F_{u a} = \frac{\sum_{i = 1}^{n} s_{u i} \times f_{i a}}{\sum_{i = 1}^{n} \sum_{a = 1}^{k} s_{u i} \times f_{i a}}$

Where $\sum_{i - 1}^{n} s_{u i} \times f_{i a}$ denotes the total number of times the categorical attribute a appears in the learning resources in which students u have produced behaviors, and $\sum_{i - 1}^{n} \sum_{a - 1}^{k} s_{u i} \times f_{i a}$ denotes the total number of all categorical attributes in the learning resources in which students u have produced behaviors.

IDF_a denotes the inverse of the frequency of occurrence of categorization attribute a in the whole collection of learning resources, and the smaller IDF_a is, the less the contribution of categorization attribute a to distinguish different student preferences, so the interference of popular categorization attributes in the calculation of student preferences can be effectively excluded. The specific calculation method of IDF_a is as follows: 2 $I D F_{a} = \log (\frac{\sum_{i - 1}^{n} \sum_{a - 1}^{k} f_{i a}}{\sum_{i - 1}^{n} f_{i a}})$

Where $\sum_{i = 1}^{n} f_{i a}$ denotes the total number of categorization attributes a in the collection of learning resources and $\sum_{i = 1}^{n} \sum_{a = 1}^{k} f_{i a}$ denotes the total number of all categorization attributes in the collection of learning resources.

According to the above two equations, it can be seen that when the frequency of categorical attribute a in the learning resources in which students have produced behaviors is greater and the categorical attribute a has better category differentiation ability, i.e., the larger TF_w and IDF_a are, the greater the students’ preference for categorical attribute a is. According to the principle of TF-IDF method, the model of students’ categorical attribute preference is established as follows: 3 $t_{u a} = T F_{u a} \times I D F_{a}$ where t_uu denotes the preference of student u for categorical attribute a, i.e., the rating of student u for categorical attribute a needed in the recommendation algorithm for reading instructional resources.

3.2.2

Similarity of knowledge structures

After clarifying students’ English reading preferences, English reading teaching in a smart classroom can help students better access the knowledge related to English reading of similar users, thus better helping them understand the knowledge structure situation. The calculation of the similarity of knowledge structure between Student A and Student B in the English reading process is illustrated as an example.

Let the set of reading resources for student A be Book_A = {b_A₁,b_A₂,⋯,b_{A_nA}}, where n_A denotes the number of reading resources for A, and the set of reading resources for student B be Book_B = {b_B₁,b_B₂,⋯,b_{B_{n_B}}}, n_B denotes the number of reading resources for student B, then the intersection of reading resources for student A and student B be Book_common = {b₁,b₂,⋯,b_{n_common}}, and n_common denotes the number of resources accessed by student A and student B together [23]. Let the student’s knowledge structure similarity (KSS) in reading resource b_k be KSS^b_k (A,B), then the knowledge structure similarity between student A and student B is: 4 $K S S (A, B) = \frac{\sum_{k = 1}^{n_{c o m m o n}} K S S^{b_{k}} (A, B)}{n_{A} + n_{B} - n_{c o m m o n}}$ where $\sum_{k} = 1^{n_{c o m m o n}} K S S^{b_{t}} (A, B)$ is the total knowledge structure similarity between students A and students B in reading n_common resources, and n_A + n_B – n_common is the total number of resources read by students A and students B.

KSS^b_k (A,B) is calculated as follows: 1)

Obtain the relevant knowledge point access paths of all students when reading Resource b_k.

2)

Obtain the list of knowledge points in Resource b_k and calculate the correlation between any relevant knowledge points in the resource.

3)

Calculate the dwell time of Student A and Student B at each knowledge point in Resource b_k based on the student knowledge point access sub-paths. The dwell time $t_{A_{i}}^{*}$ of student A at knowledge point i is: 5 $t_{A_{i}}^{*} = t_{A_{i}} + \sum_{j} [C o r (i, j) t_{A_{j}}]$ where j is the relevant knowledge point of knowledge point i, t_{A_i} is the time that the student visits knowledge point i, and t_{A_j} is the time that the student visits knowledge point j.

4)

Let the set of knowledge points visited by student A be K_A = {K_A₁,K_A₂,⋯,K_{A_{m_A}} }, m_A denote the number of knowledge points visited by student A, and the set of knowledge points visited by student B be K_B = {K_B₁,K_B₂,⋯,K_{B_{m_A}}}, m_A denote the number of knowledge points visited by student B. The intersection set of knowledge points accessed by student A and student B is K_common = {K₁,K₂,⋯,K_{m_conmon}}, m_common denotes the number of knowledge points accessed by student A and student B together, then the similarity between student A and student B in looking at a particular reading resource is: 6 $K S S^{b_{k}} (A, B) = \frac{1}{m_{A} + m_{B} - m_{c o m m o n}} \times \sum_{i = 1}^{m_{c o m m o n}} \min (\frac{t_{A_{i}}^{*}}{t_{B_{i}}^{*}}, \frac{t_{B_{i}}^{*}}{t_{A_{i}}^{*}})$ where m_A + m_B – m_common denotes the total number of knowledge points visited by Student A and Student B, and $t_{A_{i}}^{*}$ and $t_{B_{i}}^{*}$ denote the time spent at knowledge point K_i (K_i ∈ K_common) by Student A and Student B, respectively.

3.2.3

Recommended Resources for Teaching Reading

As to English reading resources in the process of recommendation, there is a duplication of recommendation, which not only affects the final recommendation results, but also causes overlapping problems. Therefore, this time, a recommendation matrix is established, setting the recommended object as n, the recommended target as k, and the English reading resources as g. The similarity is calculated by setting the matrix, i.e.: 7 $L = [\begin{matrix} k_{1} & \dots & k_{1 g} \\ ⋮ & ⋱ & ⋮ \\ k_{n 1} & \dots & k_{n g} \\ ⋮ & ⋱ & ⋮ \\ k_{n + 1} & \dots & k_{(n + 1) g} \end{matrix}] * s i m$

Where L denotes the recommendation similarity, k denotes the recommendation target, n denotes the recommendation object, g denotes the English reading recommendation resource, and sim denotes the English reading recommendation similarity threshold [24].

Based on the derived similarity, the current recommendation situation is determined, and the Internet environment is associated with the design of the English reading resource recommendation model. English reading resources mainly come from digital libraries and off-campus resources, and each resource is labeled for subsequent recommendations. Next, the associated links for reading resource recommendations are established and the model is trained. This part can randomly select a number of students for preference integration, and combined with the mined features and preferences classification recommendation, to determine the actual application effect of the current model, to continuously optimize the reading resources recommendation results according to the students’ feedback and behavior, and to strengthen the comprehensive ability of the recommendation model.

3.3

Experimental Design of English Reading Teaching

3.3.1

Research Objects and Tools

1)

Research Subjects

In order to verify the feasibility of college English reading teaching based on intelligent classroom design, this paper takes the first-year non-English major students of a university as the research object. Two classes A and B were randomly selected. The number of students in both classes was 40, and the teaching progress of English reading was equal. Class A used the English reading teaching mode designed in this paper to carry out reading instruction, while Class B used the traditional English reading teaching mode to carry out instruction. Class B was taught by using the traditional English reading teaching mode. In order to control the effects of the remaining variables, the content, the teaching time, and the teaching teacher were the same for both halves. The teaching experiment lasted from August 1, 2023, to January 31, 2024, respectively. Before and after the beginning of the teaching experiment, the reading module of the College English Test Band 4 was used as a test to clarify the differences between the teaching models.

2)

Questionnaire Design

After conducting English reading instruction, a survey is conducted to assess students’ English reading comprehension abilities and the situation related to deep learning. The questionnaire mainly includes four dimensions, namely cognitive, behavioral, affective, and evaluative. Each dimension contains a number of different types of questions, so as to understand students’ English reading comprehension ability. In the content of the questionnaire, where 1~5 means not at all, generally, relatively, and completely, respectively. The higher the score of the questionnaire, the higher the students’ agreement. Data were collected online through the Questionnaire Star platform so as to understand the current situation of students’ deep learning in English reading teaching. To ensure the validity and authenticity of the study, the questionnaire was anonymized. The data received from the survey were mainly quantitatively and statistically analyzed through SPSS software and Excel table processing software, striving to obtain objective and authentic research results.

3.3.2

Flow of teaching experiments

Before conducting the English reading teaching experiment, the teacher introduced the students to the knowledge about the smart classroom English reading teaching and analyzed the importance of the smart classroom for college English reading learning so that the students would have a general understanding of the smart classroom. The teachers followed the design of the smart classroom reading teaching strategies and applied the specific teaching strategies in college English reading teaching from August 2023 to January 2024.

Figure 3 depicts the progression of the college English reading teaching experiment. In the teaching experiment, the teacher evaluates the performance of students’ learning behaviors in the English reading class through classroom observation based on the theory of expressive evaluation, and observes the changes in students’ reading ability as well as the changes in students’ participation in the classroom to analyze the impact of the university English reading teaching strategies on students’ reading learning in the smart classroom. At the same time, the classroom observation found that the teachers need to improve their teaching strategies to improve reading skills.

4

Example Analysis of the Application of College English Reading Teaching Models

The teaching of English reading aspect is the key point of modern teaching and at the same time it is also a difficult point, so teachers should learn to guide students to strengthen the practice of English reading aspect. College students have a certain English foundation, so teachers should find a suitable method for college students to improve their English reading ability. Teachers should focus on student-centered teaching and fully mobilize students’ learning enthusiasm. Teachers’ teaching should be combined with students’ learning. Teachers teach students the correct way of reading and learning, and at the same time, they also need students to understand and find their own way of doing the questions, so as to realize the cultivation of students’ independent learning ability. The improvement of English reading comprehension ability needs to be accumulated continuously, through the correct guidance of teachers and the use of appropriate pedagogical methods to promote the improvement of students’ English reading comprehension ability.

4.1

Recommended Performance of Reading Instructional Resources

4.1.1

Comparison of Teaching Resource Recommendations

Based on the logs of students’ interaction behaviors in the smart classroom, the reading preference model between students and English reading resources is constructed, and then the similarity of students’ knowledge structure is calculated, and then the recommendation of English reading resources is realized. User-based collaborative filtering recommendation algorithm (User-CF), itembased collaborative filtering recommendation algorithm (Item-CF) and K-Means clustering-based collaborative filtering recommendation algorithm (K-Means) are adopted as the comparison methods of recommendation algorithms in this paper, and the set of recommended English reading resources items are calculated respectively. The recommendation performance of each algorithm is evaluated from the three aspects of the recommendation algorithm evaluation metrics: accuracy, recall, and F-score value, respectively. Figure 4 shows the comparison results of different algorithms for recommending English reading teaching resources.

From the results in the figure, it can be concluded that in the comparison of the accuracy of the recommendation results the recommendation algorithm for English reading teaching resources proposed in this paper has a more obvious advantage, and the accuracy rate is about 3%~40% higher than the User-CF and Item-CF algorithms when the number of students reaches a certain level. This method draws on and combines the recommendation ideas of the two basic methods, so its performance will be better than the traditional methods. Meanwhile, with the increase in the number of students, the accuracy of the recommendation results gradually smooths out and eventually reaches over 44%. The recall rates of all four algorithms gradually increase with the increase in the number of students, and the recall rate of this paper’s recommendation algorithm is about 3% to 30% higher than the recall rate of the traditional collaborative filtering recommendation algorithm. The average F-score of this paper’s recommendation algorithm in recommending English reading teaching resources is 0.365, which is about 0.052, 0.073, and 0.02 higher than User-CF, Item-CF, and K-Means algorithms, respectively.

In summary, the recommendation algorithms proposed in this paper for teaching English reading that combine students’ reading preferences and knowledge structure similarity are superior to traditional methods, with improved accuracy, recall, and F-score. It indicates that the recommendation algorithm in this paper can relatively accurately predict students’ English reading preferences when predicting students’ English reading preference resources, and provide data references for teachers to clarify students’ English reading levels.

The model of students’ English reading preference contains reading ability, cognitive style, learning goals, volume size, etc. The reading resource recommendation mainly includes resource type, topic, and difficulty, and the results of this recommendation are used to assist in analyzing students’ English reading preferences. And the correlation analysis was carried out between students’ English reading preference and the correctness and self-assessment difficulty of doing the questions. Table 1 shows the results of the correlation analysis between reading preference and correctness and error. Among them, the symbols * and ** represent significance at the 5% and 1% levels, respectively.

Table 1.

Resource recommendation and learning effect

-		Reading ability	Cognitive style	Learning goal	Volume size
Right and wrong	Correlation coefficient	0.151**	-0.093*	0.184**	0.027
Right and wrong	Sig (2-tailed)	0.002	0.016	0.001	0.535
Self-evaluation difficulty	Correlation coefficient	-0.158**	-0.014	-0.125**	-0.086*
Self-evaluation difficulty	Sig (2-tailed)	0.003	0.715	0.002	0.023

The correlation coefficients of reading ability, learning goals and correct errors were 0.151 and 0.184 by Spearman’s rank correlation coefficient test, indicating that there is a highly significant correlation between correct errors in doing reading comprehension questions and students’ reading ability and whether they have set learning goals or not. If students with low reading ability levels do better in reading comprehension questions, it shows that the recommended resources are more reasonable and can improve the learning effect of students with low reading ability, and if the English reading resources involved are related to the content of the fourth level of college English, the learning effect of learners with high levels is not obvious. It is generally believed that students with learning goals are more motivated to learn and have higher learning effects. However, it is just the opposite, so learning goals need to be further tested. The correlation coefficient between students’ cognitive styles and positive errors is -0.093 with a significant level of 0.016<0.05, which indicates that students’ cognitive styles and positive errors are significantly negatively correlated. It indicates that the field-independent learners among the selected students did better on the reading comprehension questions than the field-dependent learners. However, this result is not necessarily generalizable and may be related to the selected students. Moreover, the way of obtaining the decibel value for the degree of volume size is not perfect enough, and it can only be obtained by manual selection and only at one login, which may lead to the push learning resources not being sensitive to the changes in the learning environment, so that there is no correlation between the effect of the noisy degree of volume on the learning effect.

The correlation coefficients of students’ self-assessed difficulty with reading ability, learning objectives, and volume were -0.158, -0.125, and -0.086, respectively, indicating that the positive and cognitive styles, and the self-assessed difficulty were significantly negatively correlated with the reading ability, learning objectives, and volume level, respectively. When making recommendations for reading resources, the volume of certain English reading resources could not satisfy the students’ needs, thus making them unable to clarify their specific content when doing reading comprehension questions, leading to the occurrence of reading comprehension errors. On the whole, through the recommendation of English reading teaching resources, to a certain extent, it can assist students to improve their own reading ability and better achieve their English reading learning goals.

4.2

Analysis of the effectiveness of the teaching model

4.2.1

Comparison of reading comprehension scores

1)

Comparison of reading comprehension scores before and after the test

Before and after the teaching experiment, the English reading comprehension of the students in Class A and Class B was tested using the test paper of College English Band 4. The reading comprehension scores of students of different genders in different classes were summarized, entered into SPSS software and subjected to independent samples t-test, and the test results were shown in Table 2.

From the table, it can be seen that the pre-test reading comprehension scores of girls in class A and class B were 22.38±1.94 and 21.95±1.97 respectively, and the scores of boys in both classes were 14.15±2.13 and 14.29±2.06. This indicates that the average pre-test reading comprehension scores of girls in Class A and B were higher than those of boys and that there was a large difference between the reading levels of men and women. The results of independent samples t-test showed that there was a significant difference (p<0.01) between male and female students in pre-test reading comprehension scores, and that girls’ reading comprehension scores were significantly higher than those of boys in both Class A and Class B. This indicates that overall, girls have a higher level of reading comprehension than boys in the pre-test of reading achievement. At the end of the teaching experiment, the post-test reading comprehension scores of the girls in class A reached 26.73±5.64 points, and the overall reading comprehension scores of the boys also improved to a certain extent, with the reading comprehension scores of the girls in class A higher than those of the boys in class A by 8.49 points, and the independent samples test showed that there was a significant difference between the two (P<0.01). The girls’ posttest reading comprehension scores in class B were only 0.54 points higher than those in the pre-test, but there was still a significant difference between them and the boys’ posttest scores in class B (P<0.01). This also indicates to a certain extent that the college English reading teaching mode supported by the smart classroom can improve students’ English reading comprehension scores.

2)

Paired-samples t-test for pre and post-test of reading comprehension scores

After comparing the differences in English reading comprehension scores between students of different genders in Classes A and B, this paper conducts a paired-samples t-test on the pre and post-test scores of Classes A and B to further validate the effectiveness of the English reading teaching model designed in this paper in improving students’ reading comprehension. The results of the paired samples t-test for the pre- and post-tests are displayed in Table 3.

According to Table 3 we can clearly see the difference between the two classes, before the experiment the two classes only differed by 0.49 points in terms of average scores. However, after the experiment, there is a clear difference in the rate of improvement between the two classes, and the average grade of Class A has increased by 8.44 points. Compared with the results of the pre-test, the English reading comprehension scores of the students in class A possessed a significant difference (p<0.01). In addition, according to its standard deviation, we can also see that the standard deviation of class A is 4.29 before the experiment and 2.64 after the experiment, which can effectively reflect the relationship between each student’s performance and the average performance, the smaller the standard deviation means that the performance of the students in the class is more balanced, while the larger standard deviation means that the differentiation of the students’ reading performance scores is greater, which also shows that the students’ learning performance is not the same as that of the average. Through the changes in the standard deviation of the two classes, we can see that the differentiation of class B is further widened after the experiment, while the performance of the students in class A is more balanced, indicating that the overall reading level of the students is relatively high.

Table 2.

Comparison of reading comprehension results before and after

Test	Class	Sex	M±SD	T	Sig
Pre-test	A	Male	14.15±2.13	6.432***	0.001
	A	Female	22.38±1.94	6.432***	0.001
	B	Male	14.29±2.06	3.985***	0.000
	B	Female	21.95±1.97	3.985***	0.000
Post-test	A	Male	18.24±3.05	5.794***	0.005
	A	Female	26.73±5.64	5.794***	0.005
	B	Male	15.06±2.73	4.571***	0.003
	B	Female	22.49±2.51	4.571***	0.003

Table 3.

Test results of the matching sample of the pair

Pair	-	M±SD	T	Sig.	Mean difference
Pair 1	A-Pre-test	36.53±4.29	6.847***	0.005	8.44
Pair 1	A-Post-test	44.97±2.64	6.847***	0.005	8.44
Pair 2	B-Pre-test	36.04±4.72	0.158	0.172	1.51
Pair 2	B-Post-test	37.55±5.03	0.158	0.172	1.51

Through the table, we can see that the effect of this experiment is very obvious. Under the influence of this paper’s English reading teaching mode based on the smart classroom, the students’ learning efficiency and learning performance have been significantly improved, while the reading effect of the students who use the traditional teaching mode is not so obvious, and even the reading level between the students varies from high to low. The experiment provides us with a deeper understanding of English reading teaching and further confirms the effectiveness of integrating information technology and college English reading instruction. Therefore, it is very important to apply modern information technology to college English reading teaching, which plays a decisive role in increasing students’ interest in reading, stimulating students’ enthusiasm for reading, and improving students’ reading comprehension scores.

4.2.2

Students’ reading comprehension level

After the teaching experiment was carried out by using the university English reading teaching model supported by the smart classroom, the questionnaire designed in the previous section was utilized to investigate the students’ English reading comprehension level. The data related to students’ English reading comprehension level were obtained through the questionnaire and statistically analyzed using STATA software. Figure 5 shows the normal distribution graph of students’ English reading comprehension levels, and Table 4 shows the levels of each dimension of students’ English reading comprehension.

Table 4.

Students’ English reading understands each dimension level

-	Cognition	Behavior	Affections	Evaluation
Means	3.842	3.173	3.726	3.598
Std.	0.418	0.415	0.421	0.422
Variance	0.206	0.203	0.239	0.341
Skewness	0.179	0.117	0.178	-0.015
Kurtosis	0.283	0.209	0.215	0.174

The overall level mean of students’ English reading learning is 3.697, which indicates that the degree of achievement of students’ reading comprehension learning in English reading classroom is between average and more compliant. The standard deviation is 0.415, which is a small value, indicating that there is not much difference between the students’ own knowledge and actual actions in English reading comprehension learning. The skewness is 0.243, indicating that the normal distribution of the data is right skewed, i.e., there are more values on the right side of the normal distribution. It indicates that the students’ choices of most of the question items are above the middle value of 3 (general), which can be concluded that the students’ overall reading comprehension learning in the English reading classroom is realized to a higher degree, which also confirms that the English reading teaching mode supported by the smart classroom can significantly improve the English reading comprehension ability of the students.

In addition, regarding the levels of the dimensions of students’ English reading learning, the cognitive level average is the highest (3.842), and their affective level is the second highest (3.726). It can be seen that students have a better grasp of memorizing and understanding basic knowledge, extracting and analyzing text information in English reading learning, and students can basically achieve different degrees of emotional sublimation in reading learning. And the skewness is positive, indicating that the normal distribution is right-skewed, and the number of people choosing more than 3 is higher, from which it can be seen that in the cognitive and affective dimensions most of the students’ reading comprehension learning is still in a strong state. Evaluative and behavioral dimensions are lower, with 3.598 and 3.173 respectively. From the data, it is concluded that the students’ reading comprehension learning behavior is obviously insufficient. It shows that the students’ ability in reading learning in terms of creative application of what they have learned to solve real-world problems is lacking, and their learning is in the stage of weak initiative. In terms of standard deviation and variance, the evaluation dimension has the largest value (0.341), which indicates that students’ choices in this dimension are more varied and scattered, although it also reflects that students have different opinions.

4.3

Strategies for Improving Reading Comprehension

Combining the smart classroom with college English reading teaching can significantly enhance students’ English reading comprehension ability, fully enhance students’ reading communication ability, and also provide a reliable teaching mode support for cultivating intercultural communicative talents. On this basis, in order to further enhance students’ reading comprehension ability, this paper proposes enhancement strategies from the following dimensions.

4.3.1

Full integration of information technology tools

In the process of social modernization and development, multimedia, as an important carrier of information dissemination, integrates rich information content into teaching activities, so it is rapidly favored by the majority of teachers and students, and its value is mainly reflected in the creation of classroom teaching atmosphere. Due to their own characteristics, college students are interested in a vibrant and interesting classroom atmosphere, while the traditional teaching mode is difficult to stimulate their learning enthusiasm. Therefore, when organizing reading teaching, college English teachers can fully utilize the advantages of multimedia equipment to create a relaxed and active reading atmosphere. With the help of multimedia resources related to the theme of the reading text, students’ interest in reading exploration is stimulated, which is conducive to their active participation in reading learning.

For example, in the reading teaching of “Weather” as the theme, activate the pre-course activities with the help of daily topics, and stimulate students’ desire to read by introducing topics that students can participate in. At the same time, with the help of multimedia equipment, play different weather pictures for students to create a relaxing atmosphere and stimulate their visual senses, so as to pave the way for efficient reading teaching. Teachers can also use animated videos on related topics to visualize the content related to reading and help students imagine the scenes during reading. The design of a reading teaching mode assists them in understanding the perception of experiential reading.

4.3.2

Setting up a model for improving reading instruction

The teaching context of the reading classroom will deepen students’ emotional awareness and allow them to develop their thinking and abilities. The teaching context will pave the way for the teacher to explain the key knowledge. Therefore, teachers in the classroom teaching process, through the reading classroom teaching context setting, realize the teaching objectives, strengthen the students’ reading learning thinking, and promote the continuous development of students, so that students in the learning process to make comprehensive progress.

Improving classroom teaching links will ensure that students achieve comprehensive learning outcomes in the classroom learning process. Therefore, in the process of teaching college English in classrooms, teachers should strengthen students’ thinking development abilities through continuous improvement of teaching links. The improvement of teaching links allows students to complete English knowledge learning in a meticulous and detailed classroom teaching atmosphere will accelerate the realization of learning goals, strengthen the development of students’ thinking, and allow students to form the basic ability of comprehensive development.

4.3.3

Creating a digital reading resource bank

In college English reading teaching, the establishment and application of digital resource libraries play a very important role, which not only broadens the sources of teaching materials, but also greatly enriches students’ reading experience. By integrating diverse digital materials, such as e-books, online articles, and interactive learning software, teachers can provide students with more vivid and realistic learning materials. Take the lesson “Living with Technology” as an example. Teachers can select reading materials from digital resources that are closely related to life with technology, including the latest news on technology development, articles on how technology is changing our lifestyles, and even experts’ predictions of future technology trends. Such content can not only stimulate students’ reading interest, but also help them better understand and think about the application and impact of technology in their daily lives.

Additionally, the interactivity of digital resources is one of their major advantages. Many digital platforms allow students to exchange ideas and hold discussions with other students or learners around the world during the reading process, and this kind of communication can greatly increase students’ engagement and motivation to learn, so that they can enhance their critical thinking skills and broaden their global perspective while learning the language. However, in order to fully utilize the potential of the digital resource library, teachers also need to have the ability to screen suitable resources and design teaching methods that can effectively integrate these resources. Relying on the application of digital English reading resource library not only promotes students’ active learning, but also deepens their understanding of English reading learning, and invariably helps to improve students’ comprehensive English proficiency.

5

Conclusion

The article proposes a university English reading teaching model based on smart classroom, and designs an English reading teaching resource recommendation algorithm through students’ reading preference and resource preference mining, combined with the similarity of knowledge structure, so as to enrich university English reading teaching resources. Aiming at the effectiveness of this teaching model, the quantitative analysis of data is carried out through teaching experiments. 1)

When using the English reading teaching resources recommendation algorithm designed in this paper to recommend resources, the accuracy rate is about 3%~40% higher than that of User-CF and Item-CF algorithms when the number of students reaches a certain level. And based on the recommended English reading teaching resources, the correlation coefficients of students’ correctness and error in doing reading comprehension questions with reading ability and learning goals are 0.151 and 0.181 respectively, which indicates that there is an extremely significant correlation between the correctness and error in doing reading comprehension questions and students’ reading ability and the presence of setting learning goals or not.

2)

Before the teaching experiment, there was only a 0.49-point difference between the reading comprehension scores of the students in Class A and Class B. After the teaching experiment, there was a 7.42-point difference between the two classes, and there was a significant difference between the reading comprehension scores before and after the teaching experiment in Class A (P<0.01), but not in Class B. The difference between the two classes was 7.42 points, and there was a significant difference between the reading comprehension scores of Class A and Class B (P<0.01).

3)

The mean value of students ’ English reading learning level at the end of the teaching experiment is 3.697 points, which indicates that the achievement of students ’ reading comprehension learning in the English reading classroom is between general and comparative compliance. The smart classroom-supported university English reading comprehension teaching model can effectively improve students ’ reading comprehension level and guarantee the cultivation of intercultural communicative talents.

Language:: English

Publication timeframe:: 1 times per year
Journal Subjects:: Life Sciences, Life Sciences, other, Mathematics, Applied Mathematics, General Mathematics, Physics, Physics, other

Journal RSS Feed

Strategies for Improving College English Reading Comprehension Based on Text Mining Technology

Weihui Hong

Published Online: Mar 19, 2025

Received: Oct 30, 2024

Accepted: Jan 31, 2025

DOI: https://doi.org/10.2478/amns-2025-0494

KeywordsTF-IDF algorithm, Reading preference model, Knowledge structure similarity, Resource recommendation, English reading

© 2025 Weihui Hong, published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Keywords
TF-IDF algorithm, Reading preference model, Knowledge structure similarity, Resource recommendation, English reading