Research on Content Design and Intelligent Teaching Strategies of Civic and Political Education for Soil-based Engineering Students

Curriculum Civics is to integrate the elements of ideological and political education, including theoretical knowledge, value concepts and spiritual pursuit of ideological and political education into various courses, subconsciously and favorably influence students’ ideology and behavior, and help students form a correct worldview, outlook on life and values [1-3].

The connotation of Civic Education and the core concept of Civil Engineering Education Accreditation (CEEA) are compatible with each other, and the integration of the two can help to promote the improvement of the quality of talent cultivation [4-5]. Therefore, in the process of training civil engineering students, it is necessary to combine the civil engineering education accreditation standards and the guiding program for the construction of higher education courses on the different aspects of the requirements of college students, the civil engineering construction course combing and the establishment of the three aspects of the cultivation objectives, from the team building, the point of integration of the ideological and political, the teaching means and methods and the evaluation system, and so on, to think about the corresponding reform measures, with a view to effectively improve the course teaching quality and effect, cultivate students’ professionalism and comprehensive ability in all aspects [6-9]. Meanwhile, in the graduation requirements for civil engineering students, the requirements for designing/developing solutions, engineering and society, environment and sustainable development, professional norms and lifelong learning point out that the graduation requirements for students should not only take into account the competence in the professional aspects, but also the competence in students’ thoughts, humanistic and social literacy, and sustainable development, which indicates the importance of the Civic and Political Education in the accreditation of the civil engineering education [10-13].

At present, artificial intelligence has had a profound impact on education and teaching. Teachers of Civics and Political Science in colleges and universities must deeply understand the scientific connotation of intelligent teaching of Civics and Political Science, correctly recognize the value of intelligent teaching of Civics and Political Science in colleges and universities, and actively explore the practical path of intelligent teaching of Civics and Political Science [14-17]. Intelligent teaching of Civics and Politics course means that teaching is learner-centered and changes the teaching environment through the application of modern intelligent technology, i.e., through the collection, analysis, processing and feedback of information data [18-20]. The key to the intelligent teaching of ideological and political courses lies in the dynamic process of intelligent “transformation”, that is, the application of technical means or “intelligence + teaching” to empower the teaching process and realize the “intelligent” education and teaching function [21-23]. Intelligent teaching of ideological and political courses in colleges and universities should ultimately reflect the value of intelligent technical means in educating people, and realize the fundamental task of the goal of establishing morality and educating people [24-25].

In this paper, we first study the composition module and operation process of personalized recommendation system, choose collaborative filtering recommendation algorithm and optimize the algorithm based on the user characteristics and content recommendation similarity calculation improvement, and construct the hybrid recommendation algorithm in this paper. Then, based on the recommendation algorithm constructed in this paper, it puts forward the idea of effectively integrating the civil engineering profession and the ideological education. Finally, the recommendation algorithm of this paper carries out two rounds of validity testing of similar algorithms comparison test and practical application experiment. In the comparison of similar algorithms, the recommendation quantity and quality of the algorithm are evaluated. In the practical application experiment, the teaching effect under the support of recommendation is analyzed.

2

Recommendation system

Due to the explosive growth of information resources, it has become very difficult for ordinary users to find the content they need from a large number of resources, and as an information provider, it has also become very difficult to make their product information stand out from the crowded information base and get attention, which has led to the emergence of recommender systems. Recommender systems need to recommend resources that may be of interest to users according to their needs and interest preferences, and provide them with personalized and differentiated content services.

Figure 1 shows that the personalized recommendation system is generally composed of four parts: user information collection and behavioral record module, user attribute modeling, recommendation algorithm implementation module and content display interface.

User information collection and behavioral records module is responsible for obtaining user attributes, rating information and behavioral records from the database and Web logs to provide a data source for the implementation of the recommendation algorithm; user attribute modeling is responsible for analyzing user attributes and their behaviors, extracting effective information, and establishing a user attribute preference model; recommendation algorithms are the core modules of the recommendation system, which are required to generate the recommendation results, and recommend the contents in the resource library to users; content display interface displays the resource information to users, and continuously collects user feedback during the user browsing process to collect user behavioral information and update user requirements. The recommendation algorithm is the core module of the recommendation system, which needs to generate recommendation results, recommend the content in the resource library to the user, and meet the user’s personalized needs; the content display interface displays the resource information to the user, and in the process of browsing by the user, it constantly collects the user’s feedback, collects the user’s behavioral information, updates the user’s needs, and improves the interactivity and real-time of the system.

A recommender system is formally defined as follows: in a recommender system, let C denote the set of all registered users and S denote the set of all live products. In the actual online system, the number of users and products is usually larger in size. Let utility function $u ()$ $$u\left( {\ } \right)$$ compute user c’s evaluation of product s, i.e., the degree of favoritism u : C × S → R, where R is a fully ordered set. For each user c ∈ C, the recommender system can help him/her to find the product s′ ∈ S with the largest degree of preference, as in equation (1) (1) $\forall c \in C, \exists s \in S, s' = \arg \max u (c, s)$ $$\forall c \in C,\exists s \in S,s^\prime = \arg \max u(c,s)$$

2.1

Collaborative Filtering Recommendation Algorithm

Collaborative filtering algorithm is a classic and commonly used recommendation algorithm, which has been extended from 1992 to the present, and has been widely used in several recommendation scenarios. The so-called collaborative filtering algorithm, the basic idea is based on the user’s historical behavioral data to mine the user’s interests, based on the different interests of the user to divide and recommend items of similar interest to the user. Collaborative filtering algorithms are mainly divided into two categories: user-based collaborative filtering algorithms (User CF) and item-based collaborative filtering algorithms (Item CF). 1)

User-based collaborative filtering algorithm (User CF)

User-based collaborative filtering algorithm is based on the user’s historical behavioral data, based on the similarity between users, users with the same interests and hobbies are divided into different user preference groups, so as to recommend items based on the behavior of users within the same group and for them. The similarity of users is generally expressed by Pearson’s coefficient, which is calculated by the following formula (2): (2) $l_{u, v} = \frac{\sum_{p e P_{u, v}} (r_{u, p} - {\tilde{r}}_{u}) (r_{v, p} - {\tilde{r}}_{v})}{\sqrt{\sum_{p e P_{v, v}} {(r_{u, p} - {\tilde{r}}_{u})}^{2}} \sqrt{\sum_{p e P_{w, v}} {(r_{v, p} - {\tilde{r}}_{v})}^{2}}}$ $${l_{u,v}} = \frac{{\sum\limits_{pe{P_{u,v}}} {\left( {{r_{u,p}} - {{\tilde r}_u}} \right)} \left( {{r_{v,p}} - {{\tilde r}_v}} \right)}}{{\sqrt {\sum\limits_{pe{P_{v,v}}} {{{\left( {{r_{u,p}} - {{\tilde r}_u}} \right)}^2}} } \sqrt {\sum\limits_{pe{P_{w,v}}} {{{\left( {{r_{v,p}} - {{\tilde r}_v}} \right)}^2}} } }}$$

where r_u,p, r_v,p denote the rating vectors of user u and user v for the set of items P, respectively.

Fig. 2 shows an example of a user-based collaborative filtering algorithm. Taking user A as a designated user, based on his user behavior, the system can derive his preferences for different items and subsequently find his neighbors using the extracted hobby information. In the example, user A likes items A and E. User C is a neighboring user of user A. Therefore, the system reads the behavioral records of user C and finds that user C likes items A, D, and E. Therefore, we recommend user C’s favorite item D to user A.

2)

Item-based collaborative filtering algorithm (ItemCF)

Item-based collaborative filtering is used to get the recommended list for a given user by calculating the similarity between items instead of calculating the similarity between users. In ItemCF algorithm, the similarity between items is not computed using the content features of the items, but mainly relies on the behavioral records of the users. Specifically, the ItemCF algorithm determines that the similarity between item A and item B is high due to the fact that most of the users who like item A also like item B. This leads to the formula for calculating the similarity of items (3): (3) $w_{i j} = \frac{| N (i) \cap N (j) |}{| N (i) |}$ $${w_{ij}} = \frac{{|N(i) \cap N(j)|}}{{|N(i)|}}$$

where |N(i)| is the number of users who like item i, and |N(i) ∩ N(j)| denotes the number of users who like both item i and item j. However, this method of calculating item similarity has some limitations, when item j is very popular, the value of w_ij will be very large, almost converging to 1, which leads to any item will have great similarity with the popular item, which is obviously not in line with the original intention of the recommendation algorithm aimed at discovering the long-tail information. In order to prevent the above situation, the following formula (4) can be used to recalculate the item similarity: (4) $w_{i j} = \frac{| N (i) \cap N (j) |}{\sqrt{| N (i) | | N (j) |}}$ $${w_{ij}} = \frac{{|N(i) \cap N(j)|}}{{\sqrt {|N(i)||N(j)|} }}$$

After improvement, if N(j) is large, the denominator becomes correspondingly large, which is equivalent to penalizing the weights of item j, greatly reducing the probability that a popular item will appear to be extremely similar to a large number of the remaining items.

Figure 3 shows an example of the item-based collaborative filtering algorithm. According to the preference records of all users, if most of the users who like item B also like item D, i.e., item B’s neighbor is item D, and the specified user A likes the item, we can predict that user A will also like item D.

2.2

Improvement of similarity calculation

2.2.1

Similarity improvement based on user recommendation

In order to improve the recommendation accuracy of the algorithm, this paper introduces two penalty factors on the original Pearson similarity calculation, which are popular resource penalty factor and time decay penalty factor, and the steps for similarity calculation improvement are as follows:

The first step is to construct the user-resource scoring matrix. The construction of the matrix is completed according to the number of users and resources in the dataset and the rating information of users on resources.

In the second step, the Pearson correlation coefficient is used to calculate the similarity between the Civic resources, as shown in equation (5). In the formula, U_mn denotes the set of users who have rated both resource m and resource n, r_um and r_un denote the ratings of resource m and resource n by user u, and $\bar{r_{m}}$ $$\overline {{r_m}}$$ and $\bar{r_{n}}$ $$\overline {{r_n}}$$ denote the mean value of the scores of resource m and resource n, respectively, in a certain time period. (5) $s i m (m, n) = \frac{\sum_{u \in U_{m n}} (r_{u m} - \bar{r_{m}}) (r_{u n} - \bar{T_{n}})}{\sqrt{\sum_{u \in U_{m n}} {(r_{u m} - \bar{T_{m}})}^{2}} \times \sqrt{\sum_{u \in U_{m n}} {(r_{u n} - \bar{T_{n}})}^{2}}}$ $$sim(m,n) = \frac{{\sum\limits_{u \in {U_{mn}}} {\left( {{r_{um}} - \overline {{r_m}} } \right)} \left( {{r_{un}} - \overline {{T_n}} } \right)}}{{\sqrt {\sum\limits_{u \in {U_{mn}}} {{{\left( {{r_{um}} - \overline {{T_m}} } \right)}^2}} } \times \sqrt {\sum\limits_{u \in {U_{mn}}} {{{\left( {{r_{un}} - \overline {{T_n}} } \right)}^2}} } }}$$

In the third step, the weights of the popular resources penalty factors are added. In this paper, the course Civics learning system aims to personalize the recommendation of high-quality Civics resources that students are interested in, so it is necessary to reduce the influence of popular resources on the similarity, and appropriately increase the weight of the ratings of nonpopular resources to achieve accurate recommendation. The introduced weights are shown in Equation (6), where I_uv represents the set of resources that have been rated by user u and user v together, r_ui and r_vi are the ratings of user u and v on resource i, respectively, and r_u and r_v represent the sum of ratings of users u and v who have rated their own resources. This weighting improves the similarity calculation method by calculating the ratio of the sum of users’ jointly rated resource scores to the sum of users’ respective ratings as a penalty factor, eliminating the drawbacks of popular resources that affect the accuracy of recommendations. (6) $θ = \sqrt{\frac{\sum_{i \in I_{u p}} r_{u i} \times r_{v i}}{\sum_{i \in I_{u}} r_{u} \times \sum_{i \in I_{v}} r_{v}}}$ $$\theta = \sqrt {\frac{{\sum\limits_{i \in {I_{up}}} {{r_{ui}}} \times {r_{vi}}}}{{\sum\limits_{i \in {I_u}} {{r_u}} \times \sum\limits_{i \in {I_v}} {{r_v}} }}}$$

The fourth step, adding a time decay penalty factor, mainly for the user to browse the resources with the change of time and change the situation, to the database course of the course of the Civics as an example, the teaching progress with the advancement of time, the students learn the content and technical points are also in-depth, and each chapter of the course of the Civics elements and the specific knowledge is closely related, for example, at the beginning of the semester to learn the overview of databases and the foundation of relational databases, etc. knowledge, mid-semester study of SQL language and database table operations, and the end of the semester is a hands-on session on database design. The Civics resources browsed and rated by students in these phases are quite different, so a penalty factor should be added to these resources with too long a time interval between ratings. The time decay weights are shown in equation (7), where t_um and t_un denote the rating time of user u for resources m and n. It can be seen that the weight function takes values in the range of 0-1 and shows a nonlinear increasing trend with the increment of the rating time difference, highlighting the degree of influence of the time decay of user preferences, improving the inadequacy of the traditional similarity calculation that relies only on user ratings, and can more objectively and accurately recommend the desired resources for users. (7) $f (t) = \frac{e^{t_{u m} - t_{u n}}}{1 + e^{t_{u m} - t_{u n}}}$ $$f(t) = \frac{{{e^{{t_{um}} - {t_{un}}}}}}{{1 + {e^{{t_{um}} - {t_{un}}}}}}$$

Combining the above two penalty factors on the Pearson correlation coefficient not only reduces the scoring ratio of active resources, but also reflects the transfer of user preferences at different times. The similarity calculation formula after combining the two penalty factors is shown in equation (8). (8) $s i m_{1} (m, n) = θ \times f (t) \times \frac{\sum_{u \in U_{m n}} (r_{u m} - \bar{r_{m}}) (r_{u n} - \bar{r_{n}})}{\sqrt{\sum_{u \in U_{m n}} {(r_{u m} - \bar{r_{m}})}^{2}} \times \sqrt{\sum_{u \in U_{m n}} {(r_{u n} - \bar{r_{n}})}^{2}}}$ $$si{m_1}(m,n) = \theta \times f(t) \times \frac{{\sum\limits_{u \in {U_{mn}}} {\left( {{r_{um}} - \overline {{r_m}} } \right)} \left( {{r_{un}} - \overline {{r_n}} } \right)}}{{\sqrt {\sum\limits_{u \in {U_{mn}}} {{{\left( {{r_{um}} - \overline {{r_m}} } \right)}^2}} } \times \sqrt {\sum\limits_{u \in {U_{mn}}} {{{\left( {{r_{un}} - \overline {{r_n}} } \right)}^2}} } }}$$

2.2.2

Similarity improvement based on content recommendation

In order to solve the cold-start problem that exists when a new user enters the system and produce accurate recommendation results, it is also necessary to quantify the resource content information. Label is used to describe the content and form of the resource in general terms, giving a label to the resource is conducive to the user to quickly filter and browse the resources they want, but also facilitates the system to mine and analyze the resource data.

Firstly, the weight value of the label to the resource is defined and the resource-label matrix is constructed as shown in equation (9). Where a denotes the number of resources, b denotes the number of resource label contents, w_ab denotes the bth label attribute of the ath resource, and w_ab is 1 if the resource a has this label attribute, otherwise w_ab is 0. (9) $R = [\begin{matrix} w_{11} & w_{12} & \dots & w_{1 b} \\ w_{21} & w_{22} & \dots & w_{2 b} \\ ⋮ & ⋮ & ⋱ & ⋮ \\ w_{a 1} & w_{a 2} & \dots & w_{a b} \end{matrix}]$ $$R = \left[ {\begin{array}{*{20}{c}} {{w_{11}}}&{ {w_{12}}}& \cdots &{ {w_{1b}}} \\ {{w_{21}}}&{ {w_{22}}}& \cdots &{ {w_{2b}}} \\ \vdots & \vdots & \ddots & \vdots \\ {{w_{a1}}}&{ {w_{a2}}}& \cdots &{ {w_{ab}}} \end{array}} \right]$$

After constructing the resource-label matrix, the weight value of the label to the user is calculated as shown in equation (10). Where $r (t_{m})$ $$r\left( {{t_m}} \right)$$ denotes the sum of user’s ratings for resources with t_m labels, $r (u_{i})$ $$r\left( {{u_i}} \right)$$ denotes the sum of user’s ratings for all resources, a denotes the total number of resources, and b denotes the number of resources with t_m labels. (10) $w = \frac{r (t_{m})}{r (u_{i})} \times \frac{b}{a}$ $$w = \frac{{r\left( {{t_m}} \right)}}{{r\left( {{u_i}} \right)}} \times \frac{b}{a}$$

Then the user’s interest features and the label attributes of the resources are vectorized using the method of spatial vector representation, where the user’s interest features can be represented as $U_{m} = {(t_{1}, w_{1}), (t_{2}, w_{2}), \dots (t_{m}, w_{m})}$ $${U_m} = \left\{ {\left( {{t_1},{w_1}} \right),\left( {{t_2},{w_2}} \right), \ldots \left( {{t_m},{w_m}} \right)} \right\}$$, and the feature vector of the resources can be represented as $V_{n} = {(t_{1}, w_{1}), (t_{2}, w_{2}), \dots (t_{m}, w_{n})}, t$ $${V_n} = \left\{ {\left( {{t_1},{w_1}} \right),\left( {{t_2},{w_2}} \right), \ldots \left( {{t_m},{w_n}} \right)} \right\},t$$ for the label attributes in the resources, and w for the weights occupied by the label attributes. The similarity between resources is calculated by comparing the content attributes of the resources, as shown in equation (11). (11) $s i m_{2} (m, n) = \frac{U_{m} \times V_{n}}{\sqrt{| U_{m} | \times | V_{n} |}}$ $$si{m_2}(m,n) = \frac{{{U_m} \times {V_n}}}{{\sqrt {\left| {{U_m}} \right| \times \left| {{V_n}} \right|} }}$$

2.3

Hybrid recommendation algorithm score prediction

Combining the improved Pearson similarity sim₁(m, n) using the penalty factor and the similarity sim₂(m, n) calculated based on the content attributes to get the resource similarity of the hybrid recommendation algorithm, as shown in equation (12). Where α represents the weight accounted for by the similarity calculation method, the value range is 0-1, according to the differences in the resources can take different weights, the similarity obtained is also different. (12) $s i m (m, n) = α \times s i m_{1} (m, n) + (1 - α) \times s i m_{2} (m, n)$ $$sim(m,n) = \alpha \times si{m_1}(m,n) + (1 - \alpha ) \times si{m_2}(m,n)$$

According to the improved hybrid algorithm similarity calculation to find the set of nearest neighbors of the target resource, the predicted score of user u for resource m is shown in equation (13). Where p_um represents the predicted score of user u on resource m, S(m, K) represents the set of similar neighbors of resource m, where the first K resources are taken as the number of resources in the list to be recommended, $\bar{r_{m}}$ $$\overline {{r_m}}$$ represents the mean value of the score of resource m, r_un represents the score of user u on resource n, and $\bar{r_{n}}$ $$\overline {{r_n}}$$ represents the mean value of the score of resource n. (13) $p_{u m} = \bar{r_{m}} + \frac{\sum_{n \in S (m, K} s i m (m, n) \times (r_{u n} - \bar{r_{n}})}{\sum_{n \in S (m, K)} | s i m (m, n) |}$ $${p_{um}} = \overline {{r_m}} + \frac{{{\sum_{n \in S(m,K}}sim(m,n) \times \left( {{r_{un}} - \overline {{r_n}} } \right)}}{{\sum\limits_{n \in S(m,K)} {\left| {sim(m,n)} \right|} }}$$

Finally, for the target user, traverse all the itemsets, select the set of resources for which no behavioral records have been recorded yet, and generate the to-be-recommended list by sorting the resources in descending order of their predicted scores. Finally, the top N resources in the to-be-recommended list are selected as the Top-N recommended list.

3

Effective Integration of Civil Engineering Specialization and Civics Teaching

The effective integration of civil engineering majors and Civic and political education cannot be separated from rich teaching content design and efficient intelligent teaching. For the implementation of the educational content design and intelligent teaching strategy for civil engineering students’ civic politics, based on the recommendation algorithm system constructed in this paper, the following contents are developed on the two levels of the database of civic politics elements and the invisible integration of civic politics elements.

3.1

Building a database of Civics elements

Civics elements should be mined scientifically and reasonably in combination with the characteristics of civil engineering majors and their students, in order to enrich the database of the recommendation system and meet the diversified content design needs of different teachers for Civics teaching. 1)

Mining of Civic and Political Elements

First of all, outstanding people within the profession and their spiritual qualities can be used as elements of civics. Some outstanding figures and deeds in the field of civil engineering, such as Mao Yisheng, Zhan Tianyou and other people’s “patriotic spirit”, “scientific spirit”, “spirit of struggle” and so on can be used as elements of political thinking for in-depth excavation. Excavation. Secondly, classic quotes and stories can also be mined as elements of political thinking. For example, in the lectures of “foundation engineering” of civil engineering majors, classic quotes such as “a tree is born from the end of a thread; a nine-story platform is built on a base” can be introduced to explore the civic elements, so as to let the students know the importance of the foundation engineering, and at the same time cultivate the moral cultivation of the students to pursue the value of life unremittingly. At the same time, it cultivates students’ moral cultivation of relentlessly pursuing and striving to realize the value of life. Finally, current political hotspots with educational significance can also be mined as elements of ideology and politics. In the Internet era, college students are keen on network things and often check all kinds of hot news on the Internet, so scientifically and reasonably digging out the elements of ideology and politics in social hotspots and integrating them into course teaching can not only help students distinguish right from wrong, but also stimulate students’ interest in learning.

2)

Construction of database

At present, different professional courses contain rich elements of ideology and politics, but one of the key points to scientifically and reasonably use these elements in course teaching is to enrich, classify and sort out the elements of ideology and politics of different courses in the recommendation system. Firstly, according to the characteristics of civil engineering majors, the elements of ideology and politics should be planned and sorted out in a targeted way, and entered into the database of ideology and politics elements. Secondly, the construction of the Civic and Political Element Database is strengthened to classify and organize the possible Civic and Political Elements, and recommend appropriate Civic and Political Resources to teachers of different courses.

3.2

Implicit Integration of Civics Elements

A major key to the development of intelligent teaching strategies for the Civics course of engineering majors lies in the implicit integration of the Civics elements of the recommended algorithms into the professional curriculum in combination with the teaching content and student characteristics. First of all, the implicit integration of Civic and Political elements should grasp the principle of moderation. In the teaching process of professional courses, teachers should add the Civic and Political elements as needed in the process of teaching content design, and consciously display them in the course of teaching to stimulate the enthusiasm of students, inspire their thinking, and guide them to further exploration. Secondly, the implicit integration of Civic and Political elements should grasp the principle of personalization. In the teaching process of professional courses, it is necessary to pay attention to the individual differences of students at different stages of the recommendation algorithm system, so that the Civic and political education can naturally enter the spiritual level of each student, so as to improve the ideological and moral cultivation of students. Finally, the implicit integration of Civic and Political elements should grasp the principle of humanization. Students, as individuals who are taught knowledge, have feelings and consciousness. Therefore, when the lecturer is carrying out the intelligent teaching of the Civics course, he should reasonably use the Civics elements under the guidance of the algorithm of the recommended system in this paper, in order to trigger the students’ ideological identity and emotional resonance, so that the students can feel the glory of humanity while receiving professional knowledge. The effective integration of civil engineering professional courses and the implementation of the path of Civic and Political Education is shown in Figure 4.

4

Analysis of Recommended Effectiveness and Teaching Applications

In order to test the effectiveness of this paper’s recommendation algorithm in the actual Civics teaching effect, here we respectively launch the comparison test between this paper’s recommendation algorithm and other recommendation algorithms, and the experiment of this paper’s recommendation algorithm in the actual Civics teaching.

4.1

Comparison of this paper’s algorithm with similar algorithms

In order to verify the efficiency of the personalized recommendation algorithm studied in this paper, the algorithm is tested and experimented with two similar algorithms: collaborative filtering algorithm and similar algorithm. In the experiment, the test of recommending the Civics teaching resources in the Civics education of civil engineering majors is selected, and 150 texts in this field are extracted as test data. The distribution of these texts is relatively uniform, and the category characteristics are obvious. This experiment evaluates the recommendation quality of the system from the aspect of accuracy, which is judged by the user browsing time index, and when the length of its browsing resources is higher than 30 s, the recommendation is judged to be accurate. Based on this, subsequent experiments are conducted. The test results of the number of recommendations of the collaborative filtering algorithm, the number of recommendations of similar algorithms and the number of recommendations of this paper’s algorithm are shown in Table 1, and it can be seen that the number of recommendations of this paper is 30, 59, 89, 119, 147, and its accuracy is higher than that of the other two algorithms of the same kind.

Table 1.

Experimental result

Recommended number of resource texts/count	Collaborative Filtering Recommendation number/count	Similar algorithm recommendation number/count	The recommended number of methods in this paper/count
30	28	29	30
60	55	58	59
90	89	88	89
120	118	117	119
150	145	145	147

In order to analyze the practical application effect of this paper’s recommendation algorithm in personalized recommendation of Civics and Political Science teaching resources in civil engineering majors, this paper launches the comparison test of similar algorithms on the recommended objects. A total of three indicators are used for comparison, where H indicates the hit rate, M indicates the average inverse ranking, and D indicates the normalized discount cumulative gain. The test results of the number of collaborative filtering recommendation algorithms, the number of similar algorithms recommendation and the number of algorithms recommendation in this paper are shown in Table 2, and the test results of the three different recommendation algorithms present different characteristics. In the test results of collaborative filtering algorithm recommendation, H, M and D all show relatively stable characteristics, in which H is stable at 16.21-17.94, M is stable at 1.12-1.41, and D is stable at 4.16-4.35, and there is room for further improvement of the overall recommendation effect. In the test results recommended by the similarity algorithm, the degree of fluctuation of H, M and D increases significantly, in which the maximum value of H is 19.30, but the downward fluctuation amplitude reaches 4.08; the maximum value of M is 1.29, and the downward fluctuation amplitude reaches 0.19; the maximum value of D is 5.25, and the downward fluctuation amplitude reaches 1.95. In contrast, in the test results recommended by the algorithm of this paper, the fluctuation degree of H, M and D is significantly increased. not only show high stability, but also always at a high level, the corresponding interval ranges of H, M and D are 18.45-21.88, 1.85-2.01 and 4.40-4.85, respectively.

Table 2.

Comparison table of test results of different methods

Recommended algorithm	Evaluation index	Recommended object
Recommended algorithm	Evaluation index	1	2	3	4	5	6	7
Collaborative filtering recommended number	H	17.94	17.39	16.48	17.74	17.45	16.21	16.54
	M	1.12	1.38	1.16	1.13	1.35	1.29	1.41
	D	4.28	4.35	4.25	4.16	4.27	4.22	4.28
Recommended number of similar algorithms	H	15.22	19.01	19.30	15.74	18.18	17.97	15.6
	M	1.23	1.20	1.02	1.29	1.19	1.08	1.11
	D	3.30	4.56	4.79	5.25	5.13	5.06	3.87
The number of recommended design methods in this paper	H	21.88	18.45	20.63	19.65	19.42	20.03	18.57
	M	1.88	1.98	2.01	1.86	1.94	1.96	1.85
	D	4.66	4.40	4.59	4.78	4.53	4.69	4.85

The comprehensive test results can be concluded that the hybrid recommender system designed in this paper can realize accurate screening and effective recommendation of resources.

4.2

Instructional Design Experiments Based on Recommendation Algorithms

In order to verify the practicality of this paper’s recommendation algorithm, this paper applies the algorithm to the Civics and Political Science course for civil engineering majors in a university. From the patriotic sentiment, scientific spirit, humanistic spirit, aesthetic consciousness, network morality, cooperative spirit of the six dimensions of the use of this paper’s algorithm to carry out data organization and analysis, in order to assess the practical application of this paper’s recommended algorithm effect. Patriotic sentiment refers to the love and maintenance of the country, scientific spirit refers to the spiritual qualities embodied in the practice of natural science, humanistic spirit refers to the maintenance, pursuit and concern for human dignity, value and destiny, aesthetic consciousness refers to the dynamic reflection of the aesthetic object in the aesthetic activities, network morality refers to the basic behavior in the network of the social and ethical relations, and cooperative spirit refers to the conscious participation in multi-party collaboration to achieve common goals. The spirit of cooperation refers to the conscious participation in multi-party collaboration to achieve common goals.

4.2.1

Experimental Objects

This experiment was conducted in a university, combined with the teaching case of the implementation of Civics and Politics courses in civil engineering majors, two classes with the same basic ability were selected for this experiment, and the average grades of students in the two classes were basically the same at the time of enrollment. The teaching experiment method of this study adopts the quasi-experimental method, and the irrelevant variables are treated in the following ways: first, to ensure that the Civics and Political Science knowledge and ability of the students in the two selected classes are basically the same; second, the number of students in the experimental class and the control class is basically the same, and the ratio of male and female students is basically the same; third, the teachers of the subjects in the experimental class and the control class are all the same, including Civics and Political Science disciplines and other disciplines; fourth, the clear questionnaires of the experimental and control classes are consistent.

4.2.2

Analysis of experimental results

Table 3 shows the statistics of paired sample groups for each dimension in the pre-test and post-test of the values of the students in the experimental class. It can be seen that the average value of the total score of each dimension in the pre-test is between 19.000 and 21.000, and the average value of the total score of each dimension in the post-test reaches between 25.000 and 30.000, which indicates the effectiveness of this paper’s recommended algorithms in the design of the content of Civic and Political Education and intelligent teaching in general.

Table 3.

Values before and after test sample group statistics

dimensionality	class	number of people	Dimensional total mean	standard deviation
Patriotism	Experiment before class test	45	19.062	2.893
Patriotism	Experiment after class test	45	29.483	1.295
Scientific spirit	Experiment before class test	45	20.703	0.442
Scientific spirit	Experiment after class test	45	26.914	1.376
humanistic spirit	Experiment before class test	45	20.955	2.39
humanistic spirit	Experiment after class test	45	25.689	2.747
Aesthetic consciousness	Experiment before class test	45	21.223	1.791
Aesthetic consciousness	Experiment after class test	45	27.314	0.861
Internet morality	Experiment before class test	45	21.448	1.621
Internet morality	Experiment after class test	45	26.566	1.688
Team spirit	Experiment before class test	45	21.412	1.518
Team spirit	Experiment after class test	45	25.027	0.582

The results of the t-test of paired samples before and after the experimental class are shown in Table 4, and the P values of the six sub-items of patriotic feelings, scientific spirit, humanistic spirit, aesthetic awareness, network morality, and cooperative spirit are 0.003, 0.000, 0.000, 0.000, 0.000, 0.000, and 0.000, and the significance levels are all P<0.05, reaching a significant difference.

Table 4.

Values before and after test sample t test analysis

Dimensionality	Data	Mean Difference	Standard deviation	t	df	P value	conclusion
Patriotism	Forward test. - Back side	-2.532	4.485	-3.282	44	0.004	outstanding
Scientific spirit	Forward test. - Back side	-0.853	4.176	-4.976	44	0.000	outstanding
humanistic spirit	Forward test. - Back side	-2.692	1.925	-4.172	44	0.000	outstanding
Aesthetic consciousness	Forward test. - Back side	-2.305	2.923	-4.825	44	0.000	outstanding
Internet morality	Forward test. - Back side	-2.848	2.716	-4.001	44	0.000	outstanding
Team spirit	Forward test. - Back side	-1.56	2.966	-3.289	44	0.000	outstanding

Table 5 shows the analysis of the pre and post-test data and the results of independent samples t-test for the values of the students in the experimental and control classes. As can be seen from Table 5, the independent samples t-test results for the experimental pre-test data of the experimental and control classes were t=0.028, P=0.978, P-value=0.978>0.05, which did not reach a significant difference. The P=0.02, P-value=0.02<0.05 for the post-test results of values of the students in the experimental and control classes reached a significant difference. Therefore, the results of this data as a whole proved that the values of the students in the experimental class have changed significantly after the teaching experiment implemented by Curriculum Civics.

Table 5.

Class values pre-test and post-test statistical processing results

Class		Number of people	Average value	Standard deviation	t	df	P value	conclusion
control class	before test	45	135.636	13.981	0.027	95	0.968	non-significant
experimental class	after test	45	136.536	13.939	0.027	95	0.968	non-significant
control class	before test	45	140.149	12.116	-2.367	84	0.021	Outstanding
experimental class	before test	45	144.223	11.407	-2.367	84	0.021	Outstanding

In summary, the recommendation algorithm in this paper not only performs well among the existing recommendation algorithms and can realize the effective recommendation of the Civics resources in a relatively stable way, but also achieves obvious results in the practical application of the content design and intelligent teaching of Civics courses in civil engineering majors.

5

Conclusion

The research content of this paper is as follows: 1)

Aiming at the current problem of the hit rate of the recommendation algorithm of the civic politics teaching resources, this paper first analyzes the collaborative filtering algorithm, the content-based recommendation algorithm and the hybrid recommendation algorithm, and for the problems of the collaborative filtering algorithm, such as cold start, sparse resource matrix and inaccurate similarity computation, the similarity computation method based on the optimization of the content-based recommendation improves the recommendation effect, and constructs the hybrid recommendation algorithm of this paper.

2)

Based on the recommendation algorithm of this paper, it is proposed to implement the effective integration of civil engineering and civic education through the construction of perfect civic elements and the invisible integration of civic elements into the classroom, so as to realize the effective content design and intelligent teaching of civic education for civil engineering students.

3)

The algorithm of this paper is compared with similar algorithms, and the number of resource recommendations of this paper’s algorithm in the first test is significantly higher than the rest of similar algorithms, which indicates that this paper’s algorithm has a high degree of accuracy. In the second test, the H, M and D values of this paper’s algorithm correspond to the interval range of 18.45-21.88, 1.85-2.01, and 4.40-4.85, respectively, which not only shows high stability compared to other algorithms, but also remains at a high level.

4)

Finally, the algorithm of this paper is applied to actual teaching to analyze the application effect. The P=0.02<0.05 of the post-test results of the student values test shows a significant difference. It shows that the algorithm in this paper is feasible in optimizing the content design and intelligent teaching strategy of Civics education.

In summary, the recommendation algorithm constructed in this paper has significant advantages in optimizing the content design of Civic and Political Education for Civil Engineering majors and providing intelligent teaching strategies accurately.

Język:: Angielski

Częstotliwość wydawania:: 1 razy w roku
Dziedziny czasopisma:: Nauki biologiczne, Nauki biologiczne, inne, Matematyka, Matematyka stosowana, Matematyka ogólna, Fizyka, Fizyka, inne

Kanał RSS czasopisma

Research on Content Design and Intelligent Teaching Strategies of Civic and Political Education for Soil-based Engineering Students

Zhaochao Li

Xiya Tang

Jie He

Yanling Liu

Hui Zheng

Ling Shen

Data publikacji: 29 wrz 2025

Otrzymano: 27 gru 2024

Przyjęty: 20 kwi 2025

DOI: https://doi.org/10.2478/amns-2025-1118

Słowa kluczoweCivil engineering, Civic education, Similarity improvement, Hybrid recommendation algorithm

© 2025 Zhaochao Li et al., published by Sciendo

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

Słowa kluczowe
Civil engineering, Civic education, Similarity improvement, Hybrid recommendation algorithm