A method for visualizing social media events based on metaphor design

Abstract

The present invention discloses a social media event visualization method based on metaphor design, which corresponds microblog event data to fireworks elements one by one, and a single microblog message is metaphorically regarded as a fireworks particle; a keyword is metaphorically regarded as a fireworks launcher represented by a rectangle; and a number of microblog messages related to the keyword are aggregated as fireworks. The method is specifically as follows: (1) obtaining microblog data of social media events and preprocessing; (2) creating a background grid; (3) calculating the coordinates of fireworks particles; (4) calculating the fireworks coordinates; (5) calculating the fireworks launcher coordinates; (6) connecting the rectangles representing the fireworks launchers in the coordinates of each fireworks launcher with fuses; (7) running a program, igniting the fuse, and as time goes by, when the fuse burns to the rectangle corresponding to the fireworks launcher, the fireworks bloom; through the relationship between fireworks particles, fireworks launchers, and fireworks, the theme development of social media events is intuitively displayed.

Description

Technical Field

[0001] The present invention mainly relates to the field of visualization design. The text information processing process involves some natural language processing algorithms, and specifically relates to a social media event visualization method based on metaphor design. Background Art

[0002] Weibo is a social platform that allows users to freely publish opinions and share their daily life. Since its launch in August 2009, Weibo has gained a group of loyal fans for its information diversity, real-time nature, and rapid dissemination. With the entry of official platforms and public figures, Weibo is no longer just a tree hole application for sharing daily life. Official accounts publish relevant information on current hot topics on Weibo as soon as possible, celebrities and other public figures promote their works through Weibo to expand their popularity, and ordinary people’s Weibo posts at specific times and occasions can also attract widespread attention from the society. Weibo has become the primary source of information for many people. Therefore, studying Weibo data is of great value.

[0003] Visualization is the use of computer graphics and image processing technology to convert data into graphics or images, display them on the screen, and perform interactive processing. With the continuous advancement of social productivity and the iterative update of technology, visualization involves multiple fields such as computer graphics, image processing, computer vision, and computer-aided design, and has become a comprehensive technology for studying a series of problems such as data representation, data processing, and decision analysis.

[0004] Multi-project data streams are being generated in all fields [1,2] Visualization has been used in social media for a long time, and social media visualization analysis is also a hot area for analyzing data streams. [3,4] Analysis of dynamic text streams may focus on topics and sentiment [5,6] Or the evolution of events [7,8] Diakopoulos et al. [9] Multiple word clouds have been used to visualize how events are reflected in social media. Textflow uses the visual metaphor of a river to visualize the evolving themes in a social media event

[10] Leadline summarizes the event definition as "topic, time, person, location" and uses multiple timeline streams to visualize the event.

[11] .

[0005] At the same time, many types of visualizations have been proposed to reveal the propagation patterns in social media. Many of them use improved node-link graphs to directly visualize the forwarding tree of messages.

[12] Four different layouts (tree, circle, sail and curve) are provided to help users analyze the communication structure of microblogs.

[13] Monitor sharing activity on Twitter using 3D visualization. Users can smoothly switch to different views to explore different aspects of the information. Chen et al. [8] Analyze information diffusion with a geographic dimension.

[0006] However, existing visualization works either focus on the propagation pattern of a single piece of information and ignore the overall development of the event, or reveal the propagation process of event information and ignore the detailed propagation pattern of a single message. Therefore, based on the above purposes, while considering the evolution of Weibo event topics, we also focus on the propagation pattern of a single Weibo message and propose a novel Weibo event visualization method based on the fireworks metaphor.

[0007] References:

[0008] [1]Aigner W,Miksch S,Schumann H,et al.Visualization of time-orienteddata[M].Springer Science&Business Media,2011.

[0009] [2] Yuan J, Chen C, Yang W, et al.Asurvey of visual analytics techniques for machine learning[J]. Computational Visual Media, 2021,7(1):3-36.

[0010] [3]Chen S, Lin L, Yuan X.Social media visual analytics[C] / / ComputerGraphics Forum.2017,36(3):563-587.

[0011] [4]Wu Y,Cao N,Gotz D,et al.Asurvey on visual analytics of socialmedia data[J].IEEE Transactions on Multimedia,2016,18(11):2135-2148.

[0012] [5]Sun G,Wu Y,Liu S,et al.EvoRiver:Visual analysis of topiccoopetition on social media[J].IEEE transactions on visualization andcomputer graphics,2014,20(12):1753-1762.

[0013] [6]Zhao J,Gou L,Wang F,et al.Pearl:An interactive visual analytictool for understanding personal emotion style derived from social media[C] / / 2014 IEEE Conference on Visual Analytics Science and Technology(VAST).IEEE,2014:203-212.

[0014] [7]Chen S,Chen S,Lin L,et al.E-map:Avisual analytics approach forexploring significant event evolutions in social media[C] / / 2017 IEEEConference on Visual Analytics Science and Technology(VAST).IEEE,2017:36-47.

[0015] [8]Chen S,Li J,Andrienko G,et al.Supporting story synthesis:Bridgingthe gap between visual analytics and storytelling[J].IEEE transactions onvisualization and computer graphics,2018,26(7):2499-2516.

[0016] [9] Diakopoulos N, Naaman M, Yazdani T, et al. Social media visual analytics for events[M] / / Social Media Modeling and Computing. Springer, London, 2011: 189-209.

[0017]

[10] Cui W, Liu S, Tan L, et al. Textflow: Towards better understanding of evolving topics in text[J]. IEEE transactions on visualization and computer graphics, 2011, 17(12): 2412-2421.

[0018]

[11] Dou W, Wang X, Skau D, et al. Leadline: Interactive visual analysis of text data through event identification and exploration[C] / / 2012 IEEE Conference on Visual Analytics Science and Technology(VAST). IEEE, 2012: 93-102.

[0019]

[12] Ren D, Zhang X, Wang Z, et al. Weiboevents: A crowd sourcing weibo visual analytic system[C] / / 2014 IEEE Pacific Visualization Symposium. IEEE, 2014: 330-334.

[0020]

[13] T.N.Y.Times. Cascade. http: / / nytlabs.com / projects / cascade.html, [Online;accessed 2019-03-31]. Summary of the Invention

[0021] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide a social media event visualization method based on metaphor design. While considering the evolution of microblog event topics, the present invention also focuses on the propagation mode of a single microblog message.

[0022] The objective of the present invention is achieved through the following technical solutions:

[0023] A social media event visualization method based on metaphor design, which corresponds Weibo event data to fireworks elements one by one. A single Weibo message is metaphorically represented as a fireworks particle, and the color of the fireworks particle depends on the sentiment score of the corresponding single Weibo. If the sentiment score is greater than or equal to zero, the Weibo sentiment polarity is positive, otherwise it is negative. Keywords are metaphorically represented as fireworks launchers represented by rectangles. Several Weibo messages related to keywords are metaphorically aggregated into fireworks. The details are as follows:

[0024] (1) Obtain microblog data of social media events from microblog websites through web crawler technology, and process the obtained microblog data using the Bosen sentiment dictionary, TF-IDF algorithm, and word semantic similarity algorithm based on HowNet and Cilin to meet the needs of subsequent visualization;

[0025] (2) Use D3.js to create a canvas and create a background grid of the view on the canvas. The background grid determines the position where the fireworks particles can eventually move to;

[0026] (3) Calculate the coordinates of the fireworks particles;

[0027] (4) Calculate the coordinates of fireworks;

[0028] (5) Calculate the coordinates of the fireworks launcher;

[0029] (6) connecting the rectangles representing the fireworks launchers in the coordinates of each fireworks launcher with leads;

[0030] (7) When the program is run, the fuse is ignited. As time goes by, the fuse gradually burns to the rectangle corresponding to the firework launcher, and the firework blooms. The relationship between firework particles, firework launchers, and fireworks is used to intuitively display the theme development of the social media event.

[0031] Furthermore, all microblogs with positive polarity are selected to obtain the maximum and minimum values ​​of the sentiment scores of microblogs with positive polarity; among the microblogs with positive polarity, the color of the fireworks particles corresponding to the microblog with the minimum sentiment score is set to color A, and the color of the fireworks particles corresponding to the microblog with the maximum sentiment score is set to color B, and the color of the fireworks particles corresponding to the microblog with positive sentiment score is proportionally gradient between color A and color B;

[0032] Similarly, all microblogs with negative polarity are selected to obtain the maximum and minimum sentiment scores of microblogs with negative polarity. Among the microblogs with negative polarity, the color of the fireworks particles corresponding to the microblog with the minimum sentiment score is set to color C, and the color of the fireworks particles corresponding to the microblog with the maximum sentiment score is set to color D. The colors of the fireworks particles corresponding to the microblogs with negative sentiment scores are proportionally gradient between color C and color D; when the fireworks particles dissipate in the air, regular hexagons are generated instead, and the color of the regular hexagons is consistent with the color of the corresponding fireworks particles;

[0033] For microblogs whose number of reposts, comments and likes exceeds the set threshold, the corresponding hexagon is highlighted with a color E border. For microblogs with a repost relationship, the corresponding hexagon is displayed with a color F border;

[0034] Calculate the number of fireworks particles contained in the fireworks corresponding to each fireworks launcher. The fireworks launcher with the least number of fireworks particles is represented by a rectangle of color G, and the fireworks launcher with the most particles is represented by a rectangle of color H; the colors of the remaining fireworks launchers are proportionally gradient between colors G and H according to the quantity relationship.

[0035] Furthermore, as the fuses connecting the rectangles corresponding to the firework launcher burn, the fuses change from black solid lines to gray dotted lines.

[0036] Furthermore, step (1) is specifically as follows:

[0037] (101) Obtaining microblogs from microblog websites using web crawler technology; the method includes searching for relevant microblogs using keywords or inputting microblog IDs to obtain microblogs;

[0038] (102) For the acquired microblogs, construct corresponding forwarding relationships based on the forwarding reasons of the microblogs;

[0039] (103) Calculate the sentiment scores of microblogs based on the industry standard sentiment dictionary - Bosen Sentiment Dictionary;

[0040] (104) Extract keywords from the acquired microblogs by segmenting them according to the release time using the TF-IDF algorithm;

[0041] (105) A word semantic similarity algorithm based on HowNet and Cilin is used for each microblog to determine which keyword each microblog belongs to.

[0042] The present invention also provides a social media event visualization device based on metaphor design, comprising:

[0043] The data acquisition and processing module is used to obtain microblog data of related events from the microblog website and process the microblog data to meet the needs of subsequent visualization;

[0044] Background grid layout module, used to create a background grid on the canvas to dynamically display firework particles;

[0045] Fireworks particle coordinate calculation module, used to calculate the coordinates of fireworks particles;

[0046] Fireworks coordinate calculation module, used to calculate the fireworks coordinates;

[0047] A fireworks launcher coordinate calculation module, used to calculate the fireworks launcher coordinates;

[0048] The drawing module is used to realize the interactive display of fireworks, fireworks particles, and fireworks emitters on the canvas.

[0049] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of a social media event visualization method based on metaphor design are implemented.

[0050] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a social media event visualization method based on metaphor design.

[0051] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0052] The present invention has a good visualization effect on real microblog event data sets. Users can analyze the evolution of event topics by keywords, and can also analyze the evolution pattern of a single piece of information in an event through interactive views. According to the feedback from volunteers who participated in the survey, this method is real and effective in analyzing events. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a schematic diagram of the background grid;

[0054] Figure 2 Schematic diagram for adjusting the coordinates of fireworks particles;

[0055] Figure 3 This is a schematic diagram of the fireworks layout. DETAILED DESCRIPTION

[0056] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0057] In the field of social media visualization, existing works either focus on the propagation pattern of a single piece of information and ignore the overall development of the event, or reveal the event information propagation process and ignore the detailed propagation pattern of a single message. This embodiment proposes a novel visualization method for microblog events, which is implemented based on D3.js.

[0058] We use web crawlers to obtain Weibo event data, extract keywords from the obtained Weibo, and calculate sentiment scores. We match Weibo event data with fireworks elements one by one. A single Weibo message is metaphorically a fireworks particle, and keywords are metaphorically a fireworks launcher represented by a rectangle; several Weibo messages related to keywords are aggregated into fireworks;

[0059] The color of the fireworks particles depends on the sentiment score of the corresponding microblog. If the score is greater than or equal to zero, the sentiment polarity of the microblog is positive, otherwise it is negative. Select all microblogs with positive polarity, and obtain the maximum and minimum sentiment scores of the microblogs with positive polarity; among the microblogs with positive polarity, the color of the fireworks particles corresponding to the microblog with the minimum sentiment score is set to color A, and the color of the fireworks particles corresponding to the microblog with the maximum sentiment score is set to color B. The color of the fireworks particles corresponding to the microblogs with positive sentiment scores is proportionally gradient between color A and color B;

[0060] Similarly, all microblogs with negative polarity are selected to obtain the maximum and minimum sentiment scores of microblogs with negative polarity. Among the microblogs with negative polarity, the color of the fireworks particles corresponding to the microblog with the minimum sentiment score is set to color C, and the color of the fireworks particles corresponding to the microblog with the maximum sentiment score is set to color D. The color of the fireworks particles corresponding to the microblog with negative sentiment score gradually changes between color C and color D in equal proportion.

[0061] When the fireworks particles dissipate in the air, regular hexagons are generated instead, and the color of the regular hexagons is consistent with the color of the corresponding fireworks particles; for microblogs whose number of reposts, comments and likes exceeds the set threshold, the corresponding hexagons are highlighted with color E borders, and for microblogs with reposts, the corresponding hexagons are highlighted with color F borders. The number of fireworks particles contained in the fireworks corresponding to each fireworks launcher is calculated. The fireworks launcher with the least number of fireworks particles is represented by a rectangle of color G, and the fireworks launcher with the most particles is represented by a rectangle of color H. The colors of the remaining fireworks launchers are proportionally gradient between colors G and H according to the quantity relationship. The rectangles are connected by fuses, and as they burn, the fuses change from black solid lines to gray dashed lines. When the fuses burn to the rectangles corresponding to the fireworks launchers, the fireworks bloom in the air. By interactively exploring the relationship between fireworks particles, fireworks launchers, and fireworks, users can not only intuitively observe the development of the theme of the entire event, but also explore the propagation mode of a single message in the event and the impact it has on the event.

[0062] Specifically, the visualization method can be divided into: data acquisition and processing stage, background grid layout stage, fireworks particle coordinate calculation stage, fireworks coordinate calculation stage, fireworks launcher coordinate relationship calculation stage, and drawing stage.

[0063] 1. Data acquisition and processing

[0064] This visualization method requires real microblog event data as input. First, we use web crawler technology to obtain microblog data of related events from the microblog website, and use natural language processing algorithms on the obtained microblogs to make them meet the needs of subsequent visualization. The main tasks of the data acquisition and processing stage are:

[0065] (101) Obtaining microblogs from microblog websites using web crawler technology. Relevant microblogs can be retrieved by keyword or microblog IDs can be entered to obtain microblogs;

[0066] (102) For the acquired microblogs, construct corresponding forwarding relationships based on the forwarding reasons of the microblogs;

[0067] (103) Calculate the sentiment scores of microblogs based on the industry standard sentiment dictionary - Bosen Sentiment Dictionary;

[0068] (104) Extract keywords from the acquired microblogs by segmenting them according to the release time using the TF-IDF algorithm;

[0069] (105) Each microblog is subjected to a word semantic similarity algorithm based on CNKI and Cilin to determine which keyword each microblog belongs to. For microblogs that cannot be calculated due to the training set problem, the keywords to which they belong are divided according to the time proximity.

[0070] 2. Background Grid Layout

[0071] Use D3.js to create a canvas and create a background grid for the view on the canvas. The background grid determines where the fireworks particles can eventually move to. In order to avoid visual confusion in the generated effect, this visualization method uses a compact layout method and chooses regular hexagons to construct the background grid. The canvas is compactly covered with regular hexagons, such as Figure 1 As shown. According to the relationship between the side lengths of a regular hexagon, there is the following formula:

[0072]

[0073] Where r represents the radius of the regular hexagon, that is, the distance from the center to any vertex. l represents the distance between the centers of two horizontally adjacent regular hexagons, that is Figure 1 OO', interval v It represents the distance between the centers of two vertically adjacent regular hexagons after projection in the vertical direction, that is, Figure 1 OB. Then the calculation of the center coordinates of the regular hexagonal background grid in the i-th row and j-th column is as follows:

[0074]

[0075] In order to achieve a compact arrangement in space, the regular hexagonal background grid of even rows needs to be translated along the positive x-axis by interval l Half of the length. According to the above formula, calculate the position coordinates of the center of all regular hexagonal background grids in space.

[0076] 3. Calculation of fireworks particle coordinates

[0077] From the moment of ignition, the fireworks rise from the fireworks launcher into the air and explode into gorgeous fireworks. The number of fireworks particles contained in the fireworks is consistent with the number of Weibo contained in the corresponding keywords. One fireworks particle represents one Weibo. Each fireworks particle moves at a specific speed and along a specific angle from the moment it blooms, shrinking until it dissipates in the air.

[0078] First, the position (x0, y0) where the fireworks bloom is used as the origin coordinate. The angle and speed of movement are determined based on the properties of the corresponding microblog of the fireworks particle. The microblogs to which the keywords belong are sorted by release time, and the number of positive and negative microblogs P and N are counted respectively. The rank of each microblog in the microblog set of its corresponding polarity is recorded in turn. i . The horizontal line of the blooming origin is used as the dividing line. The upper half is used as the moving area for negative fireworks particles, and the lower half is used as the moving area for positive fireworks particles. The corresponding semicircle is divided equally according to the number of particles, and equal angles are allocated according to the order of Weibo posting time. The formula for calculating the particle movement angle is as follows:

[0079]

[0080] This method can ensure the gradual change of the release time between particles of the same polarity but different polarities, and distinguish particles by polarity for easy observation and improved visualization. The moving speed of the particles depends on the absolute value of the sentiment score of the microblog. The larger the absolute value of the sentiment score of the microblog, the faster the corresponding particle speed. The initial size of the particles is the same, and the reduction ratio when the particles move is set. When the particle size is less than the set threshold, the particle is deemed to no longer exist. Record the life cycle of the particle and the position coordinates (x finali ,y finali ). However, the fireworks particles moving in the above manner will eventually collide, stick and aggregate. To improve this situation, this visualization method adds a second round of construction, such as Figure 2 shown.

[0081] According to the position coordinates of the particles when they finally dissipate in the first round of construction ( Figure 2 Point B in the figure) and the regular hexagonal background grid position coordinates calculated in the previous section, find the end point coordinates that meet the requirements. For each firework particle, find the regular hexagon center coordinates with the closest Euclidean distance. That is:

[0082]

[0083] where x j ,y j The center coordinates of the jth unoccupied background grid are respectively represented by the center coordinates of the jth unoccupied background grid. Once a grid is occupied, the center coordinates of the grid are deleted from the set. The center coordinates of the background grid with the smallest Euclidean distance to the coordinates when the fireworks particles dissipate (x finali’ ,y finali’ ) is the actual final coordinate of the firework particle. Figure 2 The final coordinates of the fireworks particle at point B are the coordinates of point A. Update the coordinates of the fireworks particle and delete the center coordinates of this grid from the collection. The moving angle of the particle can be obtained from the coordinates of the origin of the fireworks and the final coordinates of the fireworks particle. The distance between the two points divided by the life cycle of the particle is the initial velocity. The formula is as follows:

[0084]

[0085] Assigning the angle and speed calculated in the second round of construction to the fireworks particles can avoid collisions, adhesions, and aggregations.

[0086] 4. Fireworks coordinate calculation

[0087] After preprocessing the text data obtained by the web crawler, the top n keywords with the highest scores are selected for visualization, where n is manually input. Correspondingly, there are n fireworks in the space, and the radius of each firework is R. i , R i It depends on the fireworks particle farthest from the center of the fireworks. The distance between the fireworks particle and the center is the radius R. i . Calculate the average publishing time t of microblogs containing each keyword i . i Sorting can get keyword k i The order in which the fireworks appear in the selected event information.

[0088] Read the preset canvas length width, subtract the fireworks radius of the keyword corresponding to t1 and t n The corresponding keyword's fireworks radius. Calculate the time difference t between the subsequent keyword and the first keyword respectively i1 , divide the remaining length of the canvas proportionally according to the difference.

[0089]

[0090] Set the horizontal coordinate value of the first firework position to R1, that is, x1 = R1. Ensure that all particles of the first firework can appear on the canvas. Then the horizontal coordinate value x of subsequent fireworks i =R1+l i . Make sure that the horizontal distribution of fireworks in space is strictly in the order of time. The horizontal interval of fireworks in space represents the difference in time. Set the vertical coordinate value of the first firework to y1=height / 2, where height is the width of the canvas. The height is initially large enough. Draw a horizontal ray starting from (x1,y1). According to the horizontal coordinate value, determine the initial position of the remaining fireworks on the ray as (x i ,y1).

[0091] Perform collision detection on the circular area formed by each firework in sequence, and calculate the vertical coordinate value of each firework, such as Figure 3 As shown. The coordinate starting point of each firework is (x i ,y1), perform collision detection on the circular area formed after each firework blooms from left to right in chronological order. If there is no collision with all the previous circles, the starting coordinates are the final coordinates of the firework ( Figure 3 If there is a collision, keep its horizontal coordinate x i The circular area moves vertically to both ends until it does not collide with any of the previous circles. The end with the shorter moving distance is selected as the fireworks coordinate. If the moving distances in both directions are the same, you can select either end ( Figure 3 The firework corresponding to x2 in the figure is chosen to move in the positive direction). However, when the moving distance is consistent again, in order to ensure the symmetry and coordination of the visualization effect, the opposite direction is chosen. When the firework at the end of the time sorting completes the process, the ordinate values ​​of all fireworks are sorted and the minimum ordinate value y is taken. j The y coordinate values ​​of all fireworks are subtracted from j , this is to ensure that the fireworks view is exactly in the upper left corner of the canvas to save space.

[0092] 5. Calculation of fireworks launcher coordinate relationship

[0093] The horizontal and vertical coordinates of each firework in space are obtained through the proportional scaling algorithm and collision detection algorithm in the previous section. Fireworks and firework launchers have a one-to-one correspondence, so the horizontal coordinate value of the firework also applies to the firework launcher. The relative position of the firework launcher in space should be consistent with the firework, so the vertical coordinate of the firework is proportionally reduced and then moved along the vertical axis by the same number of pixels in the positive direction. The result is the coordinate of the firework launcher in space.

[0094] If a Weibo user forwards a message belonging to keyword k i The microblogs and comments published at the same time were judged to belong to the keyword k j , then there exists k i to k j The forwarding relationship ij . Count all the forwarding behaviors between microblogs and get the forwarding relationship set E = {e 11 ,e 12 ,...,e ij ,...}. Remove duplicates from set E and delete the forwarding relationship from itself to itself, i.e., e ii Combining the keyword set K with the forwarding relationship set E, we can get a directed graph of related keywords.

[0095] Input the time set T, keyword set K, and forwarding relationship set E, and start from the keyword k with the latest release time. i Find it. j is sorted by time in k i The previous keyword, if there is k in the set K j to k i If there is a forwarding relationship between k and k, then the forwarding relationship is retained. j Previous keyword k k , to determine whether there is k k to k j If it does not exist, put the node in the queue, skip the node, and find the previous node of the node. Iterate the above process until the keyword with the earliest release time is found. Then for the keyword in the sequence, find the node that has a forwarding relationship and is closest to the keyword that was published before it, keep the forwarding relationship and remove the node from the queue, and repeat the process until the queue is empty. At this point, the spanning tree of the relevant keywords is obtained, that is, the layout of the leads between the fireworks.

[0096] 6. Drawing stage

[0097] The positions and colors of the fireworks, fireworks particles, and fireworks launchers on the canvas are obtained through the above calculations. Based on D3.js, SVG is used to draw the corresponding graphics on the canvas, and JavaScript events such as mouseover, mousemove, mouseout, and click are used to implement the interaction of mouse movement and clicking.

[0098] The present invention is not limited to the embodiments described above. The above description of the specific embodiments is intended to describe and illustrate the technical solution of the present invention. The above specific embodiments are merely illustrative and not restrictive. Without departing from the scope of the present invention and the scope of protection of the claims, a person of ordinary skill in the art can also make many forms of specific changes under the guidance of the present invention, which all fall within the scope of protection of the present invention.

Claims

1. A social media event visualization method based on metaphor design, characterized in that: The microblog event data is mapped to the fireworks element one by one. A single microblog message is metaphorically represented as a fireworks particle. The color of the fireworks particle depends on the sentiment score of the corresponding microblog. If the sentiment score is greater than or equal to zero, the sentiment polarity of the microblog is positive, otherwise it is negative; the keyword is metaphorically represented by a rectangle representing a firework launcher; several microblog messages related to the keyword are gathered together to represent fireworks; the details are as follows: (1) Obtain microblog data of social media events from microblog websites through web crawler technology, and process the obtained microblog data using the Bosen sentiment dictionary, TF-IDF algorithm, and word semantic similarity algorithm based on HowNet and Cilin to meet the needs of subsequent visualization; (2) Use D3.js to create a canvas and create a background grid of the view on the canvas. The background grid determines the position where the fireworks particles can eventually move to; (3) Calculate the coordinates of the fireworks particles; (4) Calculate the coordinates of fireworks; (5) Calculate the coordinates of the fireworks launcher; (6) connecting the rectangles representing the fireworks launchers in the coordinates of each fireworks launcher with leads; (7) When the program is run, the fuse is ignited. As time goes by, the fuse gradually burns to the rectangle corresponding to the firework launcher, and the firework blooms. The relationship between firework particles, firework launchers, and fireworks is used to intuitively display the theme development of the social media event.

2. The method for visualizing social media events based on metaphor design according to claim 1, characterized in that: Select all microblogs with positive polarity, and obtain the maximum and minimum sentiment scores of microblogs with positive polarity; among the microblogs with positive polarity, the color of the fireworks particles corresponding to the microblog with the minimum sentiment score is set to color A, and the color of the fireworks particles corresponding to the microblog with the maximum sentiment score is set to color B. The color of the fireworks particles corresponding to the microblog with positive sentiment score is proportionally gradient between color A and color B; Similarly, all microblogs with negative polarity are selected to obtain the maximum and minimum sentiment scores of microblogs with negative polarity. Among the microblogs with negative polarity, the color of the fireworks particles corresponding to the microblog with the minimum sentiment score is set to color C, and the color of the fireworks particles corresponding to the microblog with the maximum sentiment score is set to color D. The colors of the fireworks particles corresponding to the microblogs with negative sentiment scores are proportionally gradient between color C and color D; when the fireworks particles dissipate in the air, regular hexagons are generated instead, and the color of the regular hexagons is consistent with the color of the corresponding fireworks particles; For microblogs whose number of reposts, comments and likes exceeds the set threshold, the corresponding hexagon is highlighted with a color E border. For microblogs with a repost relationship, the corresponding hexagon is displayed with a color F border; Calculate the number of fireworks particles contained in the fireworks corresponding to each fireworks launcher. The fireworks launcher with the least number of fireworks particles is represented by a rectangle of color G, and the fireworks launcher with the most particles is represented by a rectangle of color H; the colors of the remaining fireworks launchers are proportionally gradient between colors G and H according to the quantity relationship.

3. The method for visualizing social media events based on metaphor design according to claim 1, characterized in that: As the fuses connecting the rectangles corresponding to the firework launchers burn, they change from solid black lines to dotted gray lines.

4. The method for visualizing social media events based on metaphor design according to claim 1, characterized in that: Step (1) is as follows: (101) Obtaining microblogs from microblog websites using web crawler technology; the method includes searching for relevant microblogs using keywords or inputting microblog IDs to obtain microblogs; (102) For the acquired microblogs, construct corresponding forwarding relationships based on the forwarding reasons of the microblogs; (103) Calculate the sentiment scores of microblogs based on the industry standard sentiment dictionary - Bosen Sentiment Dictionary; (104) Extract keywords from the acquired microblogs by segmenting them according to the release time using the TF-IDF algorithm; (105) A word semantic similarity algorithm based on HowNet and Cilin is used for each microblog to determine which keyword each microblog belongs to.

5. A social media event visualization device based on metaphor design, based on the social media event visualization method according to any one of claims 1 to 4, characterized in that: include: The data acquisition and processing module is used to obtain microblog data of related events from the microblog website and process the microblog data to meet the needs of subsequent visualization; Background grid layout module, used to create a background grid on the canvas to dynamically display firework particles; Fireworks particle coordinate calculation module, used to calculate the coordinates of fireworks particles; Fireworks coordinate calculation module, used to calculate the fireworks coordinates; A fireworks launcher coordinate calculation module, used to calculate the fireworks launcher coordinates; The drawing module is used to realize the interactive display of fireworks, fireworks particles, and fireworks emitters on the canvas.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the social media event visualization method based on metaphor design according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the social media event visualization method based on metaphor design as described in any one of claims 1 to 4 are implemented.

Images