TWM634003U - Multi-dimensional correlated word cloud system - Google Patents

Abstract

This creation discloses a multi-dimensional associative word cloud system. First, a processor receives, analyzes, and extracts words in a data set, and uses artificial intelligence to perform an association analysis method for words to generate a degree of association between each word. Construct a multi-dimensional and multi-dimensional relationship cloud diagram with a visual model of the words and their relevance. Each word is used as a cloud, and the relationship between each word is presented through the size of the cloud, the distance, the thickness of the connecting line, etc. ; In addition, words can be screened based on a keyword, and at least one alternative word can be selected, and the multi-dimensional relationship cloud diagram can be adjusted with the alternative word, so that the multi-dimensional relationship cloud diagram can be adjusted for viewing angle, brightness, or for a specific Clouds are tagged for visual analysis and observation.

Description

Multi-dimensional associated word cloud system

This creation involves a word cloud system, especially a multi-dimensional associative word cloud system that can present the association of each word in a text in a three-dimensional and multi-dimensional manner.

Analyzing the keywords in the text plays a decisive role in the application of various fields and situations. You can use this to find out the key breakthrough points of the problem or sort out the goals; and for the presentation of the analysis results, the word cloud has very good data visualization and intuition Interpretation effect, patent literature on word cloud, such as: Patent No. TWI687823B "Content Recommendation Method and Device", which is a representation of word cloud in the form of a tree diagram after text analysis, and other prior technologies for reference as follows: (1) TWI526962B "Method, device, and computer-readable storage medium for user-specific text content advertisements that provide user-associated words to select advertisements"; (2) CN107850993A "Methods for gathering and synergizing search results"; (3) ) US20080072145A1 "METHOD AND APPARATUS FOR CUSTOMIZING THE DISPLAY OF MULTIDIMENSIONAL DATA".

Although the above-mentioned literature can generate word clouds based on texts, there are still the following problems that need to be improved, such as: when the number of words to be analyzed in the data set and texts is huge, it will be difficult to present word clouds in a two-dimensional way It also satisfies the huge amount of data and the intuitive performance of word association; or when it is impossible to analyze specific keywords after the word cloud is established; and the association between words cannot be immediately interpreted intuitively; accordingly, how It is a problem that needs to be solved to make the word cloud image have a better presentation method to meet the complicated texts and data sets to be analyzed today, and to achieve a more intuitive interpretation effect.

In view of the above problems, the author of this article improved the word cloud based on years of experience in related industries; therefore, the main purpose of this creation is to provide a multi-dimensional word cloud map that can be used to input keywords according to needs Screening is performed to adjust the presented multi-dimensional word cloud. In addition, the angle of view can be adjusted according to needs to best present the word cloud. It is a multi-dimensional associated word cloud system for users to interpret more intuitively.

In order to achieve the above-mentioned purpose, the multi-dimensional associative word cloud system of this creation includes a processor, a memory unit, an artificial intelligence unit, a visualization unit, and a screening unit; during implementation, a receiving step is performed first , to obtain the data set or text to be analyzed, and then perform an association step to perform association analysis on each word in the data set or text to generate an association degree, and then perform a cloud image building step, using each word and the association degree as the Enter the data and start with The visualization model performs calculations, and outputs a multi-dimensional relationship cloud image that can present word correlations; in addition, a virtual camera unit can be included to adjust the viewing angle of the multi-dimensional relationship cloud image.

In order to enable your review committee to clearly understand the purpose, technical features and effects of this creation, the following descriptions are provided with illustrations, please refer to it.

1: Multi-dimensional associated word cloud system

11: Processor

12: Memory unit

13: Artificial intelligence unit

14:Visualization unit

15: Screening unit

16: Virtual camera unit

17: Display device

G: dataset

W: word

S: Provider

P: Multi-dimensional relationship cloud diagram

L: connection line

K: Keyword

C: Alternative words

S1: Receiving step

S2: Association step

S3: Steps to create a cloud map

S4: screening step

Figure 1 is the system architecture diagram of this creation.

Fig. 2 is a flow chart of the method of this creation.

Figure 3 is a schematic diagram of the implementation of this creation (1).

Fig. 4 is the implementation schematic diagram (2) of this creation.

Figure 5 is a schematic diagram of the implementation of this creation (3).

Fig. 6 is the implementation schematic diagram (4) of this creation.

Fig. 7 is another embodiment (1) of this creation.

Fig. 8 is another embodiment (2) of this creation.

Fig. 9 is another embodiment (3) of this creation.

Fig. 10 is another embodiment (four) of this creation.

Figure 11 is a schematic diagram of the situation of this creation (1).

Figure 12 is a schematic diagram of the situation of this creation (2).

Figure 13 is a schematic diagram of the situation of this creation (3).

Please refer to "Fig. 1". Fig. 1 is a system architecture diagram of this creation. As shown in the figure, the multi-dimensional associated word cloud system 1 of this creation includes a processor 11, a memory unit 12, an artificial intelligence unit 13, A visualization unit 14, a screening unit 15, and a display device 17, the processor 11 is connected with the memory unit 12, the artificial intelligence unit 13, the visualization unit 14, the screening unit 15, and a display device 17 respectively, as follows Example for each component:

(1) Processor 11: The processor 11 can be, for example, one of a central processing unit (Central Processing Unit, CPU) and a graphics processing unit (Graphics Processing Unit, GPU) or a combination thereof; the processor 11 can be, for example, via Connect with another information device in a wired and wireless manner and perform data transmission, browser search and download, crawler (Web Crawler), etc. to obtain at least one data set G, and the processor 11 can be used to process a plurality of data sets G The word W is extracted, and extracted.

(2) memory unit 12: the memory unit 12 can be a solid state disk (Solid State Disk or Solid State Drive, SSD), a hard disk (Hard Disk Drive, HDD), a static memory (Static Random Access Memory, SRAM ), or a random access memory (Random Access Memory, DRAM), etc., the memory unit 12 stores a data set G and a plurality of words W in the data set G, for when the multi-dimensional relationship cloud map P is established, it can be Data acquisition and processing.

(3) Artificial intelligence unit 13: for generating a degree of association between each of the words W based on a correlation analysis method based on a plurality of words W in the data set G; wherein, the degree of association analysis can be, for example: After the word W is segmented, fill in a matrix respectively, and then perform the co-word analysis (co-word analysis) of the word W by calculating the correlation between the matrices; or, for example, use machine learning to use Apriori Algorithms are used to obtain the degree of relevance; optionally, the artificial intelligence unit 13 can perform text mining (Text mining) on the data set G first, so as to filter out more representative or high-value keywords K or words W, Relevance analysis is then carried out, wherein, text mining (Text mining) can be, for example, the Jieba suite using the Python language, tm, tidytext, jiebaR, Rwordseg, tmcn and other suites in the R language, the above are just examples and are not limited to this .

(4) Visualization unit 14: the visualization unit 14 can be, for example, a graphics processing unit (Graphics Processing, GPU), which provides each word W and degree of association as input data and parameters, performs calculations with a visualization model, and outputs Present a multi-dimensional relationship cloud map P with the degree of correlation; wherein, the visual model can be, for example, using packages such as matplotlib, seaborn, Pandas in the programming language Python, packages such as Aaron and ggplot in the R language, or tools such as Visual Studio For construction, the above is only an example and not limited thereto; the multi-dimensional relationship cloud map P contains various words W, and the associated words W are connected by a connecting line L, optional , each word W can be covered by a three-dimensional pattern such as a sphere, a cylinder, a cone, and an irregular three-dimensional.

(5) Screening unit 15: for screening complex words W based on a keyword K, selecting at least one candidate word C, and using the candidate word C as input data to readjust the multi-dimensional relationship cloud map P; Among them, the keyword K can be the word W screened out by the aforementioned text mining (Text mining); and the adjustment method can be, for example, using the word W in the multi-dimensional relationship cloud image P that is not the candidate word C as Deletion, lightening, dotted lines, etc. are presented, so that the adjusted multi-dimensional relationship cloud map P can be clearly presented with the combination of alternative words C.

(6) Display device 17: The display device 17 is used to display the multi-dimensional relationship cloud map P, and the display device 17 can be, for example, a liquid crystal screen, an electronic paper, a head-mounted display, a stereoscopic projector, or a head-up display; The device 17 enables the multi-dimensional relationship cloud map P generated by the visualization unit 14 to be presented, and can be combined with virtual reality (Virtual Reality, VR), augmented reality (Augmented Reality, AR), mixed reality (Mix Reality, MR ) and other technologies, so that the multi-dimensional relationship cloud map P is clearly and concretely displayed.

Please refer to "Figure 2", which is the flow chart of the method of this creation. As shown in the figure, the implementation method of the multi-dimensional associated word cloud of this creation is as follows: (1) A receiving step S1: Please refer to "Figure 3", for this creation Implementation diagram (1), a multi-dimensional associated word cloud system 1 receives a data set G and a plurality of words W in the data set G, and its provider S can, for example, communicate with another information device in a wired or wireless manner Serial connection and data transmission, browser search and download, crawler (Web Crawler), etc.; (2) An association step S2: Please refer to "Fig. 4", which is the implementation schematic diagram (2) of this creation. The multi-dimensional association word cloud system 1 generates the relationship between the plurality of words W based on an association analysis method. A correlation degree among them; its correlation analysis method can be for example: fill in a matrix respectively after each word W is segmented, and carry out co-word analysis (co-word analysis) of word W by calculating the correlation degree between the matrices ), it can further use the item-by-item comparison method or Gulin method (KLEE); also can for example use the machine learning method (Machine Learning) to obtain the degree of correlation with the Apriori algorithm; (3) a cloud map establishment step S3: see "Figure 5" is the schematic diagram of the implementation of this creation (3). The multi-dimensional associated word cloud system 1 uses each word W and the degree of association as input data, and performs calculations with a visual model, and the output can show word associations A multi-dimensional relationship cloud map P; (4) a screening step S4, please refer to "Figure 6", which is the implementation schematic diagram of this creation (4), the multi-dimensional associated word cloud system 1 is based on a keyword K pair of complex words W screens, selects at least one candidate word C, and uses the candidate word C as input data to readjust the multi-dimensional relationship cloud map P, and the word W that is not the candidate word C is deleted, diluted, dashed, etc. Presentation in such a manner that the adjusted multi-dimensional relationship cloud image P can be clearly presented with a combination of alternative words C; preferably, the keyword K can be the word W screened out by the aforementioned text mining.

Please refer to "Fig. 7", which is another embodiment (1) of this creation. The visualization unit 14 can use the degree of relevance of each word W as an adjustment parameter to make the presentation In the current multi-dimensional relationship cloud map P, the shape of the connecting line L is used to represent the degree of association of the word W in the data set G. For example, the thickness of the connecting line L between the words W in the multi-dimensional relationship cloud map P represents The degree of association between the words W in the data set G, the higher the degree of association, the thicker the connecting line L; it can also be the length of the connecting line L between the words W in the multi-dimensional relationship cloud map P, representing the data set G The degree of association between each word W in the word W, the higher the degree of association, the shorter the connecting line L.

Please refer to "Fig. 8", which is another embodiment (2) of this creation. The visualization unit 14 can use the correlation degree of each word W as an adjustment parameter, so that in the presented multi-dimensional relationship cloud map P, each word The appearance of the three-dimensional pattern of W is used to represent the number of occurrences of the word W in the data set G, for example: the volume of each word W and the three-dimensional pattern in the multi-dimensional relationship cloud map P represents the occurrence of each word W in the data set G The number of times, the higher the number of occurrences, the larger the word W and the three-dimensional pattern; the position of each word W and three-dimensional pattern in the multidimensional relationship cloud map P can also represent the number of occurrences of each word W in the data set G, and the number of occurrences The higher the value, the word W and the three-dimensional pattern appear at the higher position of the multi-dimensional relationship cloud map P.

Please refer to "Fig. 9" and "Fig. 10". The figure shows the embodiment (3)~(4) of this creation. The multi-dimensional associated word cloud system 1 of this creation further includes a virtual computer connected to the processor 11 by information. The camera unit 16 and the virtual camera unit 16 are used to adjust the presentation angle of the multi-dimensional relationship cloud image P according to user needs, or adjust the presentation angle of the multi-dimensional relationship cloud image P based on a display rule, so that each word W can not block each other, Or make the keyword K, the alternative word C and the connecting line L of the multi-dimensional relationship cloud map P not blocked by other words W; virtual The camera unit 16 can be, for example, one of a fixed camera system (fixed camera system), a tracking camera system (tracking camera system), or an interactive camera system (interactive camera system) or a combination thereof; In the screening step S4, when the multi-dimensional relationship cloud image P is established, or the multi-dimensional relationship cloud image P is readjusted, the virtual camera unit 16 adjusts the presentation angle of the multi-dimensional relationship cloud image P based on the keyword K and the alternative word C, In order to achieve the best presentation mode; specifically, the display rule is to take a plane through which the most words W pass in the multi-dimensional relationship cloud image P, and the virtual camera unit 16 displays the multi-dimensional relationship cloud image P from a perspective parallel to this plane; and, when When there are keywords K and alternative words C to be selected, take the keyword K as the center, take a plane through which the most alternative words C pass, and make the virtual camera unit 16 display the multi-dimensional relationship cloud image P from a perspective parallel to this plane.

As mentioned above, this creation can further include a hand-held wearable device, which is informationally connected with the processor 11 and the virtual camera unit 16 for adjusting the viewing angle of the multi-dimensional relationship cloud map P. The hand-held wearable device can be, for example, a handle, a watch, glasses and other electronic devices The device carrier, when implemented, the handheld wearable device allows the user to interact with the multi-dimensional relationship cloud map P presented by the display device 17, such as adjusting the presentation angle of the multi-dimensional relationship cloud map P, zooming in and out, active marking, and text marking Wait.

This embodiment aims at "understanding dietary trends", please refer to "Figure 11" ~ "Figure 13" together; (1) First, execute the receiving step S1, the multi-dimensional associated word cloud system 1 Collect and receive all kinds of data sets G about food notes, restaurant introductions, etc. through crawlers; (2) carry out the association step S2 again, and the multi-dimensional association word cloud system 1 will segment each word W in the data set G, and then fill them in A matrix, and then use the correlation degree calculation between the matrices to carry out the correlation analysis of the word W, and further derive the correlation degree from another data set, such as "Japanese style", "curry", "Thailand", "raw Words such as "fish fillets", "wagyu", and "snacks" and the degree of association between words W, after the degree of association is obtained, the multi-dimensional relationship cloud map P can be constructed; (3) Cloud map creation step S3, multi-dimensional association The word cloud system 1 takes the above-mentioned words W and their degree of association as input data, performs calculations with a visual model, and outputs a multi-dimensional relationship cloud map P showing the degree of association. Please refer to "Figure 10", which is a schematic diagram of the situation created in this paper (a ), as shown in the figure, the multi-dimensional relationship cloud map P not only presents the aforementioned words W such as "Japanese style", "curry", "Thailand", "sashimi", "wagyu" and "snacks", but also uses words The volume represents the appearance density of the word W in the data set G, and the degree of relevance between the words W is represented by the thickness of the connecting line between the words W; (4) Screening step S4, the user can input keywords K, the multi-dimensional associated word cloud system 1 screens the complex word W based on the keyword K, selects at least one candidate word C, and adjusts the multi-dimensional relationship cloud map P with the candidate word C, please refer to "Figure 11 ", is the scenario diagram (2) of this creation, as shown in the figure, the multi-dimensional associated word cloud system 1 uses "curry" as the keyword K, and selects "Japanese style", "Thailand", "and Cattle" and other alternative words C, and the rest of the non-alternative words C are presented in a dotted line, etc., so that the adjusted multi-dimensional relationship cloud map P can be clearly presented with the combination of alternative words C; please refer to " Figure 12” is a schematic diagram of the situation of this creation (3). As shown in the figure, the multi-dimensional relational word cloud system 1 adjusts the presentation angle of the multi-dimensional relational cloud map P, so that "Japanese style", "Thailand", "Wagyu" and so on Alternative word C has the best presentation style.

It can be seen from the above that the multi-dimensional associated word cloud system of this creation includes a processor, a memory unit, an artificial intelligence unit, a visualization unit, a screening unit, and a virtual camera unit, and the artificial intelligence unit Analyze the correlation degree of the text received and stored by the processor and memory unit, and then construct a multi-dimensional relationship cloud map through the visualization unit, and this multi-dimensional relationship cloud map can be further improved through word size, position, and connection line thickness between words, etc. Displays the density of words in the data set and text, the degree of correlation between words, etc.; in addition, the user can input keywords through the screening unit and filter the alternative words related to keywords to Focus on the information that needs to be interpreted; in addition, when there is a need to change the viewing angle, it can be converted through the virtual camera unit, and the virtual camera unit can also automatically adjust the presentation of the multi-dimensional relationship cloud image based on a display rule; thus, After the implementation of this creation, it is indeed possible to provide a multi-dimensional word cloud map that can be presented, which can be filtered according to the needs of input keywords to adjust the presented multi-dimensional word cloud map. In addition, the angle of view can be adjusted according to needs. The best presentation of the word cloud, for the purpose of a multi-dimensional associated word cloud system for users to interpret more intuitively.

Only, the above-mentioned ones are only preferred embodiments of this creation, and are not used to limit the scope of implementation of this creation; anyone who is familiar with this technology can make equal changes and changes without departing from the spirit and scope of this patent. Modifications should all be covered within the patent scope of the present invention.

To sum up, this utility model patent has the patent elements of "industrial applicability", "novelty" and "progressiveness". The applicant should file a utility model patent application with the Jun Bureau in accordance with the provisions of the Patent Law.

1: Multi-dimensional associated word cloud system

11: Processor

12: Memory unit

13: Artificial intelligence unit

14:Visualization unit

15: Screening unit

17: Display device

G: dataset

W: word

Claims (7)

A multi-dimensional associative word cloud system for expressing the association of a word in a data set, including: a processor for retrieving a plurality of words in the data set; a memory unit in information connection with the processor , storing the data set, and a plurality of the words in the data set; an artificial intelligence unit, in information connection with the processor, for generating each of the words based on a correlation analysis method for the plurality of the words A degree of relevance between them; a visualization unit, which is informationally connected to the processor, and uses the words and the degree of relevance as input data, performs calculations with a visual model, and outputs a value representing the degree of relevance Multi-dimensional relationship cloud diagram; a screening unit, which is informationally connected with the processor, for screening plural words based on a keyword, selects at least one alternative word, and uses the alternative word as input data , readjusting the multi-dimensional relationship cloud diagram, so that the visualization unit presents the keyword and each of the candidate words; and a display device, in information connection with the processor, for displaying the multi-dimensional relationship cloud diagram. The multi-dimensional relational word cloud system as described in Claim 1, wherein the multi-dimensional relational cloud image includes at least two words, and the associated words are connected by a connecting line. The multi-dimensional associated word cloud system as described in Claim 2, wherein the degree of association between the two words is represented by the thickness of the connecting line. The multi-dimensional associative word cloud system as described in Claim 2, wherein the appearance density of the word in the data set is represented by the size of the word. The multi-dimensional relational word cloud system as described in Claim 1, further comprising a virtual camera unit, the virtual camera adjusts the viewing angle for presenting the multi-dimensional relational cloud image based on a display rule. The multi-dimensional associated word cloud system as described in Claim 1, wherein the display device is one of a liquid crystal screen, an electronic paper, a head-mounted display, a stereo projector, or a head-up display. The multi-dimensional relational word cloud system as described in claim 1, which includes a hand-held wearable device, which is informationally connected to the processor and the virtual camera unit for adjusting the viewing angle of the multi-dimensional relational cloud image.

Images