Strategies for Improving Students’ Language Skills by Using Information Technology in Higher Education English Education

English as a language subject, for Chinese students whose mother tongue is Chinese, there is a certain degree of learning difficulty, coupled with the complete lack of language environment, which also hinders the enhancement of students’ learning effect, so it is particularly urgent to explore more efficient English teaching methods [1-2]. The development of information technology and its good application in many fields also provides a new perspective for education and teaching, and its application in high school English teaching has a good value performance.

Specifically, the main value of applying information technology in English teaching in colleges and universities includes the following three points. First, enrich the teaching content, with the help of information technology can be integrated with the teaching resources in the Internet, student learning is no longer purely dependent on the textbook, under the guidance of richer teaching resources, English teaching can have a greater expansion of space, to help improve the effectiveness of teaching [3-6]. Second, stimulate students’ learning enthusiasm, information technology to assist high school English teaching, can make the boring language knowledge to image, interesting way to present, not only easy for students to understand and master, but also can stimulate students’ learning enthusiasm, and effectively cultivate their learning autonomy [7-10]. Third, to improve learning efficiency, the application of information technology in classroom teaching can save teachers’ writing time, and this part of the time can be used by students to learn vocabulary and phrases, but also can watch videos to exercise the English sense of language, and gradually improve students’ learning efficiency [11-14]. In short, the teaching of English education in colleges and universities with the help of information technology to innovate the concept of education, enrich the content of the curriculum, and innovate the teaching mode, so as to comprehensively mobilize the language skills of students in listening, speaking, reading, writing and writing, and to achieve a comprehensive enhancement of the English language skills of the students [15-16].

The rapid development of information technology gives more possibilities for oral teaching in high school English classroom teaching, and students can be completely immersed in the English oral communication environment and have the most authentic English conversations. Mandasari, B. et al. describe the process of social media application in English oral teaching, taking video blogs as an example, students can use Vlogs to create an interesting learning process to support their learning of English, which had a significant effect on the improvement of their oral English proficiency [17]. Wu, W. C. V. et al. created an online English learning community based on a flipped classroom under a mobile platform to study its effect on English learners’ oral proficiency and cognitive level, and the experiment showed that the online learning community promoted both meaningful and positive collaboration as well as significantly improved the learners’ oral proficiency [18]. Hsu, M. H. et al. designed an interactive chatbot system, TPBOT, which builds learning content and provides interactive exercises for English learners, eliminating learners’ fear of opening their mouths to speak English and significantly improving their English speaking skills [19]. Choirunnisa, M. R. et al. used TED Talks source videos as the instructional media for an English speaking class to study its degree of enhancement on learners’ English speaking skills, and the findings showed that TED Talks promoted students’ English expression and also improved their speaking skills, which was beneficial for students to overcome the difficulties in English speaking [20].

The level of students’ writing ability is an important marker for judging English language proficiency, and information technology can support the development of students’ writing ability. Ahmadi, D. M. R. examined the importance of information technology in facilitating second language learning and showed that information technology, as an important language teaching tool, can help teachers to carry out classroom activities in a better way and to improve the learners’ attitudes towards learning, thus strengthening the learners’ English reading and writing skills [21]. Guo, K. et al. explored the potential of ChatGPT in facilitating English writing teaching and learning, and suggested that teachers and ChatGPT work together to cultivate students’ English writing skills by providing a greater number of evaluative feedbacks, a relatively even focus of the feedbacks, and a more diverse range of feedbacks, compared to teachers’ evaluative feedbacks on students’ writing [22]. Zou, D. et al. used interactive video tools and real-time collaborative tools to build a new technology-supported flipped learning model, which enables students to develop better writing skills, writing motivation, and critical thinking through instant instruction and peer coaching [23]. Handayani, E. T. et al. studied the role of WhatsApp, a mobile media, in improving students’ writing skills, and the empirical investigation found that students had a positive attitude towards WhatsApp, indicating that it can bring positive impacts and benefits in the process of teaching and learning writing [24].

With the application of information technology, there are richer and easier ways to teach English listening, and students can utilize the rich English listening materials provided by the Internet for English listening training.Chamundeshwari, C. et al. examined the effect of technology-enhanced language learning (TELL) on learners’ English listening skills, and the study showed that TELL provides learners with listening strategies to help them achieve better performance in their study courses [25]. Suryana, I. et al. investigated the specific use of AI mobile applications in improving learners’ English listening, and the findings showed that Netflix was the most effective and efficient AI mobile application to improve English listening [26]. Khazinat, K. et al. analyzed students’ perceptions of using multimedia applications to improve their English listening skills, using the eBelajar English listening learning platform as the subject of the study, which was found to be reliable in improving students’ listening skills, but occasionally poses challenges to students when used [27].

Combined with the development and requirements of information technology, this paper proposes to change the way of teaching English in colleges and universities and implement the English teaching strategy of blended learning. With the objective of improving students’ language skills, it encourages them to use information technology to enhance their independent learning abilities. Among them, the multimedia resource retrieval technology for English education combining TF-IDF and VSM is proposed. Break down the automatic classification process of VSM combined with TF-IDF. Enhance the TF-IDF algorithm and conduct comparative tests to confirm the efficacy of the improved TD-IDF feature weighting algorithm in text categorization. Setting up the teaching experiment process, the experimental group adopts the way of blended learning activities and the control group adopts the way of face-to-face learning activities to compare the before and after changes in the English scores of the two classes.

2

Construction of Informatized English Classroom in Higher Education

Informatized foreign language classroom teaching goes beyond classroom teaching and encompasses the entire process of students’ foreign language learning. The time should include the assignment of independent learning tasks by the teacher and the preparation of students’ pre-study before the class, the teacher’s explanation and students’ participatory and interactive discussion during the class, and the internalization of students’ language knowledge and reconstruction of the knowledge system after the class. The location should combine offline teacher-led classroom teaching activities with online independent learning for students outside the classroom. In response to the updated transformation of time and space, this paper proposes the English information-based teaching strategy-blended learning strategy.

2.1

Changes in teaching methods

2.1.1

Blended learning

Blended learning is a combination of face-to-face teaching and online learning, which is a new learning mode in the context of education informatization [28].

The online learning platform used in this study is “English Learning Community”, which not only provides access to multimedia resources, but also provides the functions of posting and discussing, doing homework, recording learning progress, etc. It also supports the recording of short scenarios, allows students to complete multiple exercises with automatic marking, and tracks their learning progress.It also supports the dubbing of short scenarios, allows students to do multiple exercises and review automatically, and provides anonymous mutual assessment. Teachers can use the platform to keep track of students’ learning progress, practice, discussion focus, etc., thus discovering students’ good performance and problems on the platform, and giving feedback to students in class.

The activities in the classroom involve peer collaboration to accomplish tasks, role-playing in contextual simulations, and group competitions, etc.The teacher plays the role of guiding and guiding the students throughout the entire process. Teachers play a guiding and supporting role in the whole process, including giving timely feedback on good performances and problems, guiding classroom activities, correcting phonological and grammatical errors in English expressions, and answering questions and solving problems. Based on this, this study designed the functional allocation of each role in blended learning activities.

The functional allocation of each role in the blended learning activity is shown in Figure 1. It consists of three stages: the stage of previewing new knowledge, the stage of reinforcing key points, and the stage of drilling difficult points.

2.1.2

Analysis of the value of blended learning

With the continuous popularization and application of network technology and electronic information technology in the cause of higher education, more and more online learning platforms have appeared in the teaching of various disciplines in universities. College English classroom teaching can obtain rich teaching resources from it, which provides convenient conditions for online teaching. Teachers specializing in college English can use the teaching platform and network of information resources built by universities to carry out more targeted teaching activities.Students can also communicate with teachers and ask questions in real-time on the platform or app while learning online. Teachers are also able to provide differentiated guidance to students in the class, greatly enhancing the relevance of the teaching process. Compared to previous teaching, online-offline blended teaching better solves the problem of space, time, and other binding factors, which has a positive significance on improving teaching quality.

Online and offline hybrid teaching is a new teaching mode based on the background of the Internet era, which makes it clearer that students are the main focus of classroom teaching. With the help of the network platform constructed by colleges and universities or the platform or App software that teachers require students to use uniformly, students can independently learn the knowledge content of the English subject online, which can better stimulate their learning enthusiasm and initiative.

The online-offline blended teaching mode not only clarifies the students’ main position in college English teaching, but also brings into play the teacher’s leading role, which is more reflected in the teacher’s inspiration and guidance to the students. With the help of big data technology teachers can organize and summarize the learning habits of all kinds of students, and set up personalized learning plans for them, gradually expanding the knowledge field of different students, even in the same context can be more efficient teaching results, truly meet the students’ personalized learning needs of the English subject.

2.2

Utilization of information technology

2.2.1

Students’ capacity for independent learning

Based on the functional allocation of students in blended learning activities, in order to further improve students’ language skills, this paper proposes a strategy to emphasize the cultivation of students’ ability to use information technology for independent learning.

Students are the participants in the informationized foreign language classroom, and they are in a central position in the informationized foreign language classroom. Teachers should properly guide students to identify the use of information technology in foreign language teaching and actively participate in interactive and research-based learning activities. Cultivate students to use informationization to improve their independent learning ability. Language knowledge is the foundation of foreign language learning, and it is more important to use language knowledge to find problems, solve them, and improve language skills.

In the age of informationization, students’ learning ability is no longer limited to mastering language knowledge. It is equally important to use information to improve their independent learning ability.Learning to use a foreign language is the purpose of foreign language teaching.In the face of the great convenience brought by multimedia information means, students need to improve their own information literacy skills and make good use of multimedia information means to learn foreign language knowledge.Actively participate in quality online classes, micro-courses, catechism classes, and other teaching activities on the Internet, and collect the strengths of all for their own use.Collect, organize, and summarize foreign language learning materials, share the learning content with classmates, and participate in classroom learning of foreign language information. Accumulate language knowledge while emphasizing the improvement of language skills and humanistic qualities.

2.2.2

Search technology for multimedia resources in English education

The following two algorithms are used in this design, TF-IDF weighting algorithm, spatial vector model (VSM).TF-IDF and VSM are combined for text content based multimedia resource classification [29-30].

The principle of text-based automatic multimedia classification algorithm is to use text to represent the content of multimedia, and then the multimedia is processed through a succession of text classifications.

Text classification refers to the use of computers to automatically classify and label text sets according to a certain classification system or standard, which generally includes organizing the training set, preprocessing, statistics, feature extraction, training classifiers, evaluation and other steps.

1)

VSM algorithm

In the vector space model, each document is viewed as a collection of words, which is then represented as a vector of word weights: (1) $d_{i} = (w_{i 1}, w_{i 2}, \dots, w_{i n})$ where d denotes a document and n denotes the dimension of the lexical space. The weight of each lexical entry represents the importance of that lexical entry in the document. The weights of the lexical entries are usually computed using the TF-IDF method or some of its variations. The similarity of two documents is expressed as the cosine of the angle between their corresponding vectors. I.e: (2) $S i m (d_{i}, d_{j}) = \frac{\sum_{k = 1}^{n} w_{i k} \times w_{j k}}{\sqrt{(\sum_{k = 1}^{n} w_{i k}^{2}) (\sum_{k = 1}^{n} w_{j k}^{2})}}$

2)

TF-IDF algorithm

TF-IDF is the most commonly used method of calculating weight values to assess the importance of a lexical item for a particular document in the whole document set or corpus. The word frequency tf indicates how often the lexical item appears in the document. The inverse document frequency idf reflects the importance of the lexical item in the document dataset, and the main calculation formulas are shown in Eqs. (3) to (5): (3) $t_{d t} = \frac{m_{d t}}{M_{d}}$ (4) $i d f_{t} = \log (\frac{N}{n_{t} + 0.1})$ (5) $w_{d t} = t_{d t} \times i d f_{t}$

The tf_dt value usually needs to be normalized, typically by the frequency m_dt of occurrence of the feature item t in the text d divided by the total number of words M_d in the text d to prevent it from being biased towards longer documents. idf_t Divide the total number of documents N by the number of documents n_t containing the feature term t (adding 0.1 to avoid a denominator of 0). The resulting quotient is then logarithmically obtained. w_dt represents the weight value of feature item t in the text d.

In order to make the weight value of each feature item is in the interval of [0,1], the cosine normalization is used for normalization, and the new weight calculation formula is obtained as shown in Equation (6), and k is the number of feature items in the text d: (6) $w_{d t} = \frac{t f_{d t} \times i d f_{t}}{\sqrt{\sum_{i = 1}^{k} {(t f_{d k} \times i d f_{k})}^{2}}}$

Information gain (IG), defined as the difference between the information entropy H(Y) of the dataset D and the conditional entropy H(Y|X) of the condition X given the condition Y. The larger the value of information gain, the stronger the ability of feature items to distinguish data categories. The specific formulas are shown in Eq. (7) to Eq. (9): (7) $H (Y) = - \sum_{i = 1}^{n} p_{y i} \log p_{y_{i}}$ (8) $H (Y | X) = - P x_{i} \sum_{i = 1}^{n} p (y_{i} | x_{i}) \log p (y_{i} | x_{i})$ (9) $I G (Y, X) = H (Y) - H (Y | X)$

In the categorization problem, the information gain value is calculated by counting the occurrence or non-occurrence of a certain feature item x_i in category y_i. The formula is shown in Equation (10), where p_i represents the probability that a text of category y_i appears in the corpus, i.e., the number of texts of category y_i divided by the total number of texts. p_xi indicates the probability of the text containing feature item x_i in the corpus, i.e., the number of texts containing feature item x_i divided by the total number of texts. ${\bar{x}}_{i}$ indicates the text that does not contain feature item x_i. p(y_i|x_i) denotes the conditional probability that the text belongs to the category y_i when it contains feature item x_i, i.e., the number of texts containing feature item x_i and belonging to category y_i divided by the number of texts containing feature x_i. That is: (10) $\begin{matrix} I G (Y, X) = - \sum_{i = 1}^{n} p_{y_{i}} \log p_{y_{i}} + P x_{i} \sum_{i = 1}^{n} p (y_{i} | x_{i}) \log p (y_{i} | x_{i}) \\ + P {\bar{x}}_{i} \sum_{i = 1}^{n} p (y_{i} | {\bar{x}}_{i}) \log p (y_{i} | {\bar{x}}_{i}) \end{matrix}$

The improved TF-IDF algorithm incorporates the information extraction result term with the information gain value φ to the text category into the weight calculation formula to increase the feature term’s ability to differentiate between categories, and obtains the improved weight calculation formula as shown in equation (11): (11) ${w^{'}}_{d t} = \frac{t f_{d t} \times i d f_{t}}{\sqrt{\sum_{i = 1}^{k} {[t f_{k k} \times i d f_{k}]}^{2}}} \times (1 + φ)$

In the text representation process, the keywords in the text that are directly related to the information extraction results are identified, and if the keywords corresponding to the information extraction results are true, then the improved weight calculation formula ${w^{'}}_{d t}$ is used to calculate the value of the feature item weight, otherwise it is calculated according to the traditional weight calculation formula w_d. The text feature item weight assignment is shown in equation (12). t indicates the feature items in the text. T indicates the set of text keywords directly related to the extraction results. R indicates that the corresponding information extraction result is true. That is: (12) $W_{d t} = {\begin{array}{l} w_{d t}, t \in T \cap R \\ w_{d t} \end{array}$

3)

The process of automatic classification using VSM combined with TF-IDF

Text categorization generally includes the process of text expression, classifier selection and training, evaluation and feedback of classification results, among which texta can be subdivided into text preprocessing, indexing and statistics, feature extraction and other steps. The overall block of the text classification system is: 1)

Organize the training set: organizing the training set is to collect documents according to the known category system that meet the criteria of different categories under the system. The documents in the training set must represent the category, otherwise the classifier trained according to the training set will not be able to obtain good classification results.

2)

Preprocessing: this step is mainly to read the content of the text document and make a word separation process on its content. In the process of word separation, you can add the words of a specialized field or words without different meaning (such as “?”, “ah”, etc.) in the dictionary, in order to improve the accuracy of word separation.

3)

Statistics: This step mainly utilizes the formula to calculate the weights of all the words appearing in the results of the previous step. It should be noted that in the calculation process, each word only needs to be calculated once. In order to facilitate the subsequent feature extraction and training of the classifier, it is recommended to store the computed results and words as key-value pairs in relevant files, such as databases, txt files, xml files, etc. in this step.

4)

Feature extraction: extract features from the document that reflect the theme of the document, the significance of this step is to extract feature items that can reflect the theme of a certain type of document from the separated words according to the calculation results of the previous step. The specific extraction method is to extract a number of words in accordance with the order of weight from large to small as a feature item of the category.The problem that may arise at this time is that the same word may have different weights in different documents. The paper’s approach is to take the maximum value.

5)

Classifier: the training of the classifier, according to the feature items extracted from a category in the previous step to calculate the weight of each feature item in different documents under the category, to get the vector model of all the training documents under the category, and finally calculate the average of these vectors that is the vector model of the category.

6)

Evaluation: the test results of the classifier are analyzed, and the evaluation is to measure the classification accuracy of the classifier obtained in the fifth step.

Such tests are divided into two categories: closed tests and open tests. A closed test is when the classifier is tested directly using the training set as the test set. On the contrary, closed testing is to re-gather the data set to test the classifier. If the results of the test meet the expected classification requirements, the classifier can be used in practice. Otherwise, the classifier should be retrained from the first step.

3

Evaluation of text classification performance

3.1

Evaluation indicators

The evaluation metrics for text classification performance are checking completeness, checking accuracy, correctness, macro-averaging, micro-averaging, and so on.

The formula for the classifier’s checking rate on class c_i is as follows: (13) $Re c a l l_{i} = \frac{T P_{i}}{T P_{i} + F N_{i}}$

The formula for the classifier’s checking accuracy on class c_i is as follows: (14) $\Pr e c i s i o n_{i} = \frac{T P_{i}}{T P_{i} + F P_{i}}$

Macro average check rate: (15) $M a c A v g_Re c a l l = \frac{\sum_{i = 1}^{| C |} Re c a l l_{i}}{| C |}$

Macro average checking accuracy: (16) $M a c A v g_p r e c i s i o n = \frac{\sum_{i = 1}^{| C |} p r e c i s i o n_{i}}{| C |}$

F -Measurement:

Recall and Precision are two conflicting performance metrics, so in many cases the two are considered together. The most commonly used method is F-Measurement, the formula is as follows: (17) $F_{B} (P, R) = \frac{(β^{2} + 1) P R}{β^{2} P + R}$ Where β is the adjustment factor, which is used to assign different weights to Recall and Precision. When β is 1, i.e. Recall and Precision are given the same weight, then F-measurement is called F₁-indicator and the formula is as follows: (18) $F_{1} (P, R) = \frac{2 P R}{P + R}$

3.2

Experimental simulation results and analysis

For the feasibility of the improved algorithm proposed in this paper, experimental simulation is carried out to analyse.

The experiment was carried out under Windows 10 system, hardware environment: intel core i7-2350M, 2.4GHZ, memory: 8GB, software environment Python3.0.

The text set used in this experiment is from the Natural Language Processing Group of the International Database Centre, Department of Computer Information Technology, Fudan University, and some documents from 10 classes are selected to be combined into a training set and a test set. There are 1800 documents in the training sample and 950 documents in the test sample. The experimental text categories include education, transport, art, environment, computer, economy, agriculture, politics, history, and sports.

3.2.1

Completeness rate

The KNN algorithm with better classification results is used in the experimental simulation to compare the check-perfect, check-accurate, macro-averaged check-perfect, macro-averaged check-accurate, and F1 obtained using the TF algorithm, the TF-IDF algorithm, the TF-IDF-CHI algorithm, the TF-IDF-IG algorithm, and the TF-IDF algorithm improved in this paper.

The check accuracy rates (%) obtained by different feature weighting algorithms are shown in Fig. 2. It can be seen that the TF algorithm appears to have the lowest check-percentage on the experimental art text, which is 51.63%.The improved TF-IDF algorithm in this paper achieves the highest search rate in a total of six experimental texts, namely, transport, art, computer, economy, agriculture, and history. All other text categories are above average.

3.2.2

Accuracy

Similarly, the check accuracy is calculated for each algorithm. The checking accuracy (%) obtained by different feature weighting algorithms is shown in Fig. 3. The improved TF-IDF algorithm in this paper has the highest mean value of 79.562% in each experimental text category when compared to the check accuracy rate of each algorithm.

3.2.3

Overall evaluation of classification effectiveness

The check all rate and check accuracy rate are used to calculate the overall evaluation of the classification effect, and the overall evaluation of the classification effect (%) under different feature weights is depicted in Fig. 4.

The values of macro-averaged check completeness, macro-averaged check accuracy and F1 obtained by the improved TF-IDF feature weighting algorithm in this paper are higher than the classification indexes under the TF algorithm, TF-IDF algorithm, TF-IDF-CHI algorithm and TF-IDF-IG algorithm.

The improved algorithm in this paper is higher than other algorithms in MacAvg_Recall, MacAvg_Precision and F1 by 1.209%-9.648%, 4.324%-10.641% and 3.67%-9.76% respectively.

In summary, the improved TF-IDF feature weight algorithm in this paper, based on the inheritance of the advantages of the TF-IDF algorithm, is able to measure the weights of feature words better than the TF-IDF algorithm, thus obtaining better classification performance. The experiment indicates that the improved TF-IDF features in this paper are feasible.

4

Testing the implementation of English language skills enhancement strategies

4.1

Experimental flow of blended learning activities

The experimental flow of the blended learning activity is shown in Figure 5. The study was carried out in the autumn semester of 2023, and after that, the formal experiment was carried out after adjusting the problem accordingly. The time of the formal experiment is from March 2024 to July 2024. The experimental subjects were 85 middle-level students, including 43 male students and 42 female students, aged between 20 and 22 years old in the English major 2022 class of Shanghai S-school. The experimental group had 45 students while the control group had 40 students.

The selected course was ‘College English’. Before the official class, both the experimental and control groups took the first general examination and filled out the learning attitude questionnaire.Afterwards, the experimental group used blended learning activities while the control group used face-to-face learning activities to study the course. At the end of the course, both experimental and control groups took the second general examination, and the experimental group completed the course satisfaction questionnaire. Finally, both the experimental and control groups took the oral examination of the Professional English Level Examination (PELE).

4.2

Statistical tests and results of experimental data

The data on English composite scores (including listening scores), speaking scores, and students’ attitudinal scales obtained from the pre-test and post-test for both the experimental and control classes. Next, further analyses of the two classes’ achievements and attitudes using statistical analysis with significance of difference test are required separately. To verify whether using IT-supported blended learning teaching strategies for teaching English has a significant impact on the students in the two classes.

4.2.1

Combined results

In this study, the written tests have been used to obtain the English composite scores of the pre-test and post-test of the experimental and control classes, which are large samples because the number of both experimental and control classes is greater than 30. Therefore, the method of Z-test was chosen to detect the significant difference between the two independent samples.

The significance test for the difference between the pre-test composite scores of the experimental and control classes is shown in Table 1.

Table 1.

The difference significance test of the previous survey

Class	Cross-reference class	Laboratory class
Average	66.19160844	62.05796512
Known covariance	521.42569315	557.33210894
Observed value	40	45
Hypothesis mean difference	0
z	0.29936654
P(Z⇐z) Double tail	0.78954276
z Double tail critical	1.89571243

In the test of the significance of the difference between the pre-test composite scores of the experimental and control classes, the Z-value was compared with the critical value, Z = 0.29, Z < 1.90, P > 0.05.

According to the test statistical inference rule, it was concluded that the difference between the English composite scores of the experimental class and the control class was not significant prior to the use of IT-supported blended learning instructional strategies for teaching and learning.

The significance test of the difference between the post-test composite scores of the experimental and control classes is shown in Table 2.

Table 2.

The difference significance test of the post-test comprehensive achievement

Class	Cross-reference class	Laboratory class
Average	68.94251039	75.89694414
Known covariance	510.33690457	489.60090915
Observed value	40	45
Hypothesis mean difference	0
z	1.99365115
P(Z⇐z) Double tail	0.04172116
z Double tail critical	1.96336405

In the test of significance of the difference between the post-test composite achievement of the experimental and control classes, comparing the Z value with the critical value, Z = 1.99, Z > 1.96, P < 0.05. According to the test statistical inference rule, it is concluded that the difference between the English composite achievement of the experimental class and the control class is significant after teaching with the IT-supported blended learning teaching strategy.

The significance test of the difference between the pre-test and post-test composite scores of the experimental class is shown in Table 3.

Table 3.

The difference significance test of the previous survey and the post-test results

Class	Posttest	Premeasurement
Average	75.89694414	62.05796512
Known covariance	489.60090915	557.33210894
Observed value	45	45
Hypothesis mean difference	0
z	3.36602499
P(Z⇐z) Double tail	0.03578167
z Double tail critical	1.91232459

In the test of the significance of the difference between the pre-test and post-test composite scores of the experimental class, comparing the Z value with the critical value, Z = 3.37, Z > 1.91, P < 0.05.

According to the test statistical inference rule, it is concluded that there is a significant difference in the English comprehensive achievement of the experimental class before and after teaching using IT-supported blended learning teaching strategies.

From the above test results, it can be concluded that before using IT-supported blended learning teaching strategies for teaching, the English composite scores of the students in the experimental class and the control class are basically the same. After using IT-supported blended learning teaching strategies for teaching, the English comprehensive achievement of the students in the experimental class was significantly improved. Therefore, the use of IT-supported blended learning teaching strategies can significantly improve students’ overall English achievement.

4.2.2

Listening performance

Similarly, pre and post-test difference tests were conducted on the listening scores. The significance of the difference between pre-test and post-test listening scores of the experimental class is shown in Table 4.

Table 4.

The significance test of the previous test and the performance of the following

Class	Premeasurement		Posttest
Class	Cross-reference class	Laboratory class	Cross-reference class	Laboratory class
Average	8.96536145	8.23669876	9.01219556	12.36151196
Known covariance	15.00248691	14.79003125	12.03650491	14.60611757
Observed value	40	45	40	45
Hypothesis mean difference	0		0
z	0.49360654		4.59683615
P(Z⇐z) Double tail	0.68917153		0.03992736
z Double tail critical	0.19833469		2.75915344

In the test of significance of the difference between the post-test listening scores of the experimental class and the control class, comparing the value of Z with the critical value, Z = 4.60, Z > 2.76, P < 0.05. According to the rule of statistical inference of the test, it is concluded that the difference in the English listening scores of the experimental class before and after the use of information technology-supported blended learning instructional strategies for teaching is extremely significant.

The table shows that the English listening scores of the students in the experimental class and the control class were almost identical before using IT-supported blended learning teaching strategies. After using IT-supported blended learning teaching strategies, the English listening performance of students in the experimental class was significantly improved. Therefore, the use of IT-supported blended learning teaching strategies can significantly improve students’ English listening performance.

4.2.3

Oral performance

In this study, the scores of pre-test and post-test speaking tests of the experimental and control classes have been obtained, and as in the case of the analysis of the written test scores, the test was chosen to detect the significant difference between the two independent samples.

The test of significance of difference between the pre-test and post-test speaking scores of the experimental class is shown in Table 5. In the test of significance of difference between the post-test speaking scores, Z=3.68>2.68 and P<0.05. The difference between the experimental class and the control class in English speaking scores is extremely significant.

Table 5.

The difference significance test of oral performance

Class	Premeasurement		Posttest
Class	Cross-reference class	Laboratory class	Cross-reference class	Laboratory class
Average	58.90141791	57.63541299	59.66120431	69.79802286
Known covariance	348.3100525	326.0147391	345.75750563	291.01066302
Observed value	40	45	40	45
Hypothesis mean difference	0		0
z	0.37569033		3.68211902
P(Z⇐z) Double tail	0.82665995		0.00280069
z Double tail critical	1.91131250		2.67936651

5

Conclusion

This paper proposes an implementation strategy for building an English information technology teaching classroom in colleges and universities, integrating the advantages of online learning and face-to-face teaching, and proposing a blended learning strategy. In response to the development requirements of the information technology era, it focuses on cultivating students’ ability to use information technology in order to improve their language skills in colleges and universities.

The combination of TF-IDF and VSM is applied to classify English education multimedia resources based on text content, and the overall performance of English education text resource classification is good. The improved TF-IDF feature weighting algorithm accurately measures the weights of feature words compared to the TF-IDF algorithm, resulting in better classification performance. It is 1.209%-9.648%, 4.324%-10.641%, 3.67%-9.76% higher than TF-IDF algorithm on MacAvg_Recall, MacAvg_Precision and F1 respectively.

The significant difference between the pre and post-test scores was tested by analyzing the combined scores of the pre and post-tests of the experimental and control classes.There is a significant difference between the students’ English comprehensive scores before and after the implementation of blended learning activities.The blended learning English teaching strategy and the cultivation of students’ ability to use information technology can effectively improve students’ English achievement and help improve language skills.

Funding:

This study was funded by the Research and Reforms in Teaching Projects of Panzhihua University (Grant No. JJ24108)

Idioma:: Inglés

Calendario de la edición:: 1 veces al año
Temas de la revista:: Ciencias de la vida, Ciencias de la vida, otros, Matemáticas, Matemáticas aplicadas, Matemáticas generales, Física, Física, otros

RSS Feed de revista

Strategies for Improving Students’ Language Skills by Using Information Technology in Higher Education English Education

Quanquan Zhu

Publicado en línea: 17 mar 2025

Recibido: 04 nov 2024

Aceptado: 12 feb 2025

DOI: https://doi.org/10.2478/amns-2025-0324

Palabras claveTF-IDF algorithm, VSM model, Information technology English classroom, Language skills

© 2025 Quanquan Zhu, published by Sciendo

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

Palabras clave
TF-IDF algorithm, VSM model, Information technology English classroom, Language skills