Multi-topic opinion evolution method based on confidence interval

Abstract

The invention provides a multi-topic opinion evolution method based on a confidence interval. The multi-topic opinion evolution method comprises the steps that an intelligent agent i collects and stores neighbor information Ni (k); the intelligent agent i performs distance calculation on the collected neighbor information and own opinions so as to classify neighbors to obtain neighbor categories; according to the neighbor category, obtaining a neighbor Nb and a neighbor Nc which have a large opinion difference; trust discounting is carried out on the neighbor Nb and the neighbor Nc, and the trust degree after discounting and a trust degree matrix after discounting are obtained; a feedback matrix is deduced according to the neighbor category, a diagonal matrix capable of distributing credibility is obtained through the feedback matrix and the trust matrix, an element dii on a diagonal line of the diagonal matrix represents the redistributable credibility of the agent i, and credibility redistribution is carried out on the discounted trust matrix; and completing multi-topic opinion evolution through a dynamic formula. By means of the method, the stable consensus can be achieved for a plurality of topics with the logic dependency relationship.

Description

Technical Field

[0001] The present invention relates to the field of social network opinion evolution, and in particular to a multi-topic opinion evolution method based on confidence intervals. Background Art

[0002] In the digital age of the 21st century, social networking platforms such as Tik Tok, Tieba, and Weibo have become an essential part of daily life and a core channel for information transmission and public opinion construction. These platforms allow users to freely share and express their views and emotions on various events, products, or opinions. The posts published by users on social networks not only reflect their individual positions, but may also have a significant impact on the attitudes and behaviors of other users. Therefore, exploring the changing trends of public opinion in social networks - that is, how opinions are formed and evolved when the public discusses a specific issue - has become an important topic in the fields of social science research and information technology.

[0003] In theory, opinion dynamics is intended to describe the evolution of opinions among a group of interacting individuals. Social networks can be modeled based on the fact that everyone holds their own opinions and exchanges them with each other. "Consensus", "polarization" and "splitting" are the main characteristics of the final stage of group opinion evolution. By modeling social networks, the direction of public opinion can be analyzed, which helps to better guide public opinion, respond to sudden incidents of online public opinion, and prevent the occurrence and spread of rumors.

[0004] Existing social network opinion evolution models mostly focus on the study of a single topic, but such methods are not suitable for discussing a group of topics with logical dependencies. When discussing a group of logically related topics, the affirmation of one topic is often affected by the mixed influence of other topics. And if there are stubborn people with extreme opinions among the discussion group, the opinions of others will be seriously disturbed.

[0005] Therefore, the key issue is how to establish dependencies between topics in a weighted directed topology in discrete time and reduce the interference caused by extreme opinions. Summary of the invention

[0006] The purpose of the present invention is to overcome the problem of existing opinion evolution models that the public is disturbed by extreme opinions under multi-topic opinion evolution, resulting in serious deviation from their own opinions. A multi-topic opinion evolution method based on confidence intervals is provided to classify neighbors who exchange opinions, reduce the trust of neighbors with extremely different opinions, and enhance the trust of neighbors with similar opinions.

[0007] The specific technical solution of the present invention is:

[0008] A multi-topic opinion evolution method based on confidence interval, comprising:

[0009] In the discrete time model with fixed topology, at time k, agent i collects and stores neighbor information N i (k);

[0010] Agent i calculates the distance between the collected neighbor information and its own opinions to classify the neighbors and obtain the neighbor categories;

[0011] According to the neighbor categories, we get N neighbors with large opinion differences. b and neighbor N c ; and to the neighbor N b and neighbor N c Discount the trust to obtain the discounted trust and the discounted trust matrix;

[0012] The feedback matrix is ​​derived according to the neighbor categories, and a diagonal matrix of assignable trust is obtained through the feedback matrix and the trust matrix, and the elements d on the diagonal of the diagonal matrix are ii represents the redistributable trust of agent i, and redistributes the trust of the discounted trust matrix;

[0013] Through the dynamic formula, the agent's opinion information is updated to complete the evolution of multi-topic opinions.

[0014] The agent i calculates the distance between the collected neighbor information and its own opinions to classify the neighbors and obtain neighbor categories, including:

[0015] Agent i calculates the distance between the collected neighbor information and its own opinions. And, each agent has a confidence interval in and They are respectively called the lower confidence limit and upper confidence limit of agent i;

[0016] Classification is performed based on whether the opinions on different topics from those of neighbors are within their own confidence intervals.

[0017] The classification is performed based on whether the opinions on different topics from neighbors are within their own confidence intervals, including:

[0018] Class A neighbor N a are neighbors whose distances to all topic opinions are within their own confidence intervals;

[0019] Class B neighbor N b There are neighbors whose distances of some topic opinions are within their own confidence intervals, while some are not within their own confidence intervals;

[0020] Neighbors of type C care neighbors whose distances to all topic opinions are outside their own confidence intervals.

[0021] The neighbor N b and neighbor N c Conduct trust discounts, including:

[0022] Set coefficients α, β and α>β∈(0,1), corresponding to the neighbors N with large opinion differences b and neighbor N c , and discount its trust, that is, or in is the discounted trust; w ij It’s the original level of trust.

[0023] The step of deriving a feedback matrix according to neighbor categories includes:

[0024] The feedback matrix F∈R is derived based on the neighbor categories. n×n :

[0025]

[0026] where f ij It represents the element in the i-th row and j-th column in the feedback matrix, that is, the feedback coefficient of individual i to neighbor j.

[0027] The diagonal matrix of assignable trust is obtained by using the feedback matrix and the trust matrix, and the elements d on the diagonal of the diagonal matrix are ii Represents the trust that can be redistributed by agent i, and redistributes the trust after the discount trust matrix, including:

[0028] After obtaining the feedback matrix, a diagonal matrix D = diag (W × F T ), the elements on its diagonal d ii represents the trust that can be redistributed by agent i, and finally the discounted trust matrix Redistribute trust, that is

[0029] Where W is the trust matrix; F T is the transpose of the feedback matrix.

[0030] The above method updates the agent opinion information through the dynamics formula to complete the multi-topic opinion evolution, including:

[0031] A dynamic formula for the logical dependency relationship between topics is established, and opinions are iteratively evolved through the dynamic formula to update the agent's opinion information and complete the evolution task according to conditions.

[0032] The kinetic formula is:

[0033]

[0034] The conditions for completing the evolution task are:

[0035] For agent i, if the condition lim is satisfied k→∞ ||x i (k+1)-x i (k)||=0, the evolution task is completed;

[0036] Among them, x i (k+1) is the opinion of individual i at time k+1; x i (k) is the opinion of individual i at the kth moment. When the difference between the two opinions is infinitely close to 0, it means that the evolution task is completed.

[0037] Beneficial effects:

[0038] The beneficial effects of the present invention include at least:

[0039] The method of the present invention first collects and stores neighbor information N i (k); and calculate the distance between the collected neighbor information and its own opinions to classify the neighbors and obtain neighbor categories; then, according to the neighbor categories, obtain the neighbors N with large opinion differences b and neighbor N c ; and to the neighbor N b and neighbor N c Trust discount is performed to obtain the discounted trust and the discounted trust matrix; then, a feedback matrix is ​​derived according to the neighbor category, and a diagonal matrix of assignable trust is obtained through the feedback matrix and the trust matrix, and the elements d on the diagonal of the diagonal matrix are ii Represents the trust that can be redistributed by agent i, and redistributes the trust after discounting the trust matrix; finally, the agent opinion information is updated through the dynamic formula to complete the multi-topic opinion evolution. The present invention classifies neighbors through confidence intervals and adjusts the trust of neighbors, thereby weakening the trust of neighbors with extreme opinions, thereby reducing the interference of extreme opinions on the agent's own opinions. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a flow chart of the method of the present invention.

[0041] Figure 2 It is the simulation result of the multi-topic opinion evolution model with confidence intervals in a custom network.

[0042] Figure 3It is the simulation result of the multi-topic opinion evolution model with confidence interval in the ER network. DETAILED DESCRIPTION

[0043] The present application is further described below in conjunction with the accompanying drawings.

[0044] In order to explain the present application in more detail, it is necessary to explain the prior art in detail.

[0045] The multi-topic opinion evolution method based on confidence interval proposed in the present invention uses the confidence interval neighbor classification algorithm to hierarchically divide neighbors to update the trust, such as Figure 1 , the specific steps are as follows:

[0046] Step 1: Under the fixed topology of the discrete time model, at time k, individual i collects and stores neighbor information N i (k);

[0047] Step 2: Individual i calculates the distance between the collected neighbor information and its own opinion:

[0048] And, each individual has a confidence interval in and They are called the lower and upper confidence limits of individual i respectively. According to whether the opinions on different topics with neighbors are within their own confidence intervals, neighbors are divided into three categories: Class A N a It is a neighbor whose distance of all topic opinions is within its own confidence interval; Class B N b There are neighbors whose distances of some topic opinions are within their own confidence intervals, and some are not within their own confidence intervals; Class C N c are neighbors whose distances to all topic opinions are outside their own confidence intervals.

[0049] Step 3: After completing the neighbor classification, set the coefficients α, β and α>β∈(0,1), corresponding to the neighbors N with large opinion differences. b and neighbor N c , and discount its trust. or in It is the discounted trust level.

[0050] Step 4: In order to achieve trust redistribution, it is necessary to derive the feedback matrix F∈R according to the neighbor categories n×n :

[0051]

[0052] Step 5: After obtaining the feedback matrix, we can obtain a diagonal matrix D = diag (W × F T ), the elements on its diagonal d ii Represents the trust that can be redistributed by individual i. Finally, the discounted trust matrix Redistribute trust, that is

[0053]

[0054] Step 6: The updated trust matrix can be substituted into the dynamic formula to update the individual opinion information. Specific embodiment 1:

[0056] The present invention provides an embodiment:

[0057] At the initial state of opinion evolution, people’s opinions on the topic are collected and discretized into values ​​between [-1, 1], where [-1, 0] represents an opposing attitude and (0, 1] represents an approving attitude. X(k) is the opinion matrix of all individuals at the kth moment. in represents the attitude of individual i towards the mth topic at the kth moment. At the same time, the trust relationship between people is mathematically modeled and represented by the trust matrix W, W = [w ij ]∈R n×n is a weight matrix whose elements w ij is the weight of edge (i, j), representing the degree of trust of individual i on individual j. Therefore, for a given matrix W, it uniquely determines a network, denoted as G(W). The neighbor set of individual i is defined as N i ={j|w ij ≠0,j≠i}. When j=i, w ij is regarded as the degree of adherence to one's own opinion. And the matrix W is a row random matrix, that is, For All of them hold true. Through the multi-question dependency structure matrix Establish dependencies between topics.

[0058] After the initial state is set, the opinions are iteratively evolved through the dynamics formula:

[0059]

[0060] Among them, λ i It represents the sensitivity of individual i to external influences, with a value range of [0,1], usually λ i =1-w ii , that is, the self-weight w iiIt is considered to be an indicator of the sensitivity or closedness of the ith individual to interpersonal influence. i It is an individual's obsession, usually taking the form of u i =x i (0), which is the individual’s initial opinion. The specific input and output of the model are as follows:

[0061]

[0062]

[0063] Compared with existing research algorithms, such as Figure 2 ,3,The advantage of the present invention is that the neighbors are classified through confidence intervals, the trust of neighbors with extremely different opinions is reduced, the trust of neighbors with similar opinions is enhanced, and the resistance of individuals to extreme opinions is improved. Specific embodiment 2:

[0065] This application also provides an embodiment:

[0066] An electronic device comprises: a storage medium and a processing unit; wherein the storage medium is used to store a computer program, and the processing unit exchanges data with the storage medium, and is used to execute the computer program through the processing unit when performing multi-topic opinion evolution, and perform the steps of the multi-topic opinion evolution method based on confidence intervals as described in specific embodiment 1. Specific embodiment 3:

[0068] A computer-readable storage medium having a computer program stored therein; when the computer program is run, the steps of the confidence interval-based multi-topic opinion evolution method described in specific embodiment 1 are executed.

[0069] In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0070] The above disclosures are only several specific implementation scenarios of the present application, but the present application is not limited thereto, and any changes that can be thought of by a person skilled in the art should fall within the protection scope of the present application. The above application numbers are only for description and do not represent the advantages and disadvantages of the implementation scenarios.

Claims

1. A multi-topic opinion evolution method based on confidence interval, characterized in that: include: In the discrete time model with fixed topology, at time k, agent i collects and stores neighbor information N i (k); Agent i calculates the distance between the collected neighbor information and its own opinions to classify the neighbors and obtain the neighbor categories; According to the neighbor categories, we get N neighbors with large opinion differences. b and neighbor N c ; and to the neighbor N b and neighbor N c Discount the trust to obtain the discounted trust and the discounted trust matrix; The feedback matrix is ​​derived according to the neighbor categories, and a diagonal matrix of assignable trust is obtained through the feedback matrix and the trust matrix, and the elements d on the diagonal of the diagonal matrix are ii represents the redistributable trust of agent i, and redistributes the trust of the discounted trust matrix; Through the dynamic formula, the agent's opinion information is updated to complete the evolution of multi-topic opinions.

2. The confidence interval-based multi-topic opinion evolution method according to claim 1, characterized in that: The agent i calculates the distance between the collected neighbor information and its own opinions to classify the neighbors and obtain neighbor categories, including: Agent i calculates the distance between the collected neighbor information and its own opinions. In addition, each agent has a confidence interval - lower i ≤x j -x i ≤upper i ,in and They are respectively called the lower confidence limit and upper confidence limit of agent i; Classification is performed based on whether the opinions on different topics from those of neighbors are within their own confidence intervals.

3. The confidence interval-based multi-topic opinion evolution method according to claim 2, characterized in that: The classification is performed based on whether the opinions on different topics from neighbors are within their own confidence intervals, including: Class A neighbor N a are neighbors whose distances to all topic opinions are within their own confidence intervals; Class B neighbor N b There are neighbors whose distances of some topic opinions are within their own confidence intervals, while some are not within their own confidence intervals; Neighbors of type C c are neighbors whose distances to all topic opinions are outside their own confidence intervals.

4. The confidence interval-based multi-topic opinion evolution method according to claim 1, characterized in that: The neighbor N b and neighbor N c Conduct trust discounts, including: Set coefficients α, β and α>β∈(0,1), corresponding to the neighbors N with large opinion differences b and neighbor N c , and discount its trust, that is, or in is the discounted trust; w ij It’s the original level of trust.

5. The confidence interval-based multi-topic opinion evolution method according to claim 1, characterized in that: The step of deriving a feedback matrix according to neighbor categories includes: The feedback matrix F∈R is derived based on the neighbor categories. n×n : Among them, f ij It represents the element in the i-th row and j-th column in the feedback matrix, that is, the feedback coefficient of individual i to neighbor j.

6. The confidence interval-based multi-topic opinion evolution method according to claim 1, characterized in that: The diagonal matrix of assignable trust is obtained by using the feedback matrix and the trust matrix, and the elements d on the diagonal of the diagonal matrix are ii Represents the trust that can be redistributed by agent i, and redistributes the trust after the discount trust matrix, including: After obtaining the feedback matrix, a diagonal matrix D = diag (W × F T ), the elements on its diagonal d ii represents the trust that can be redistributed by agent i, and finally the discounted trust matrix Redistribute trust, that is Where W is the trust matrix; F T is the transpose of the feedback matrix.

7. The confidence interval-based multi-topic opinion evolution method according to claim 1, characterized in that: The above method updates the agent opinion information through the dynamics formula to complete the multi-topic opinion evolution, including: A dynamic formula for the logical dependency relationship between topics is established, and opinions are iteratively evolved through the dynamic formula to update the agent's opinion information and complete the evolution task according to conditions.

8. The confidence interval-based multi-topic opinion evolution method according to claim 7, characterized in that: The kinetic formula is: Among them, λ i represents the sensitivity of individual i to external influences; u i is an individual's obsession; C is the structural matrix.

9. The confidence interval-based multi-topic opinion evolution method according to claim 7, characterized in that: The conditions for completing the evolution task are: For agent i, if the condition lim is satisfied k→∞ ||x i (k+1)-x i (k)||=0, the evolution task is completed; Among them, x i (k+1) is the opinion of individual i at time k+1; x i (k) is the opinion of individual i at the kth moment. When the difference between the two opinions is infinitely close to 0, it means that the evolution task is completed.

Images