Method for analyzing digital marketing data and computing apparatus for performing the same

Abstract

A digital marketing data analysis method performed by a computing device according to one embodiment comprises: a step of generating at least one document containing numerical information for calculating advertising performance indicators based on the result of automatically analyzing data attributes of a tabular dataset collected from at least one digital advertising platform for digital marketing performance analysis; a step of vectorizing the generated at least one document and storing it in a vector database to enable RAG-based search; a step of extracting semantically similar reference documents through RAG-based search in the vector database in response to an input natural language-based query; and a step of inputting the query and the reference documents into an artificial neural network model to provide the result of the digital marketing performance analysis and a list of documents referenced in the digital marketing performance analysis as the output of the artificial neural network model.

Description

Technology Field

[0001] The present invention relates to a method for analyzing digital marketing data and a computing device for performing the method. Background Technology

[0002] The main duties of an ad operator include planning and executing ad campaigns, managing and running ads, analyzing ad performance, and analyzing trends and optimizing strategies.

[0003] Among these, advertising performance analysis involves analyzing key performance indicators such as Click-Through Rate (CTR), Conversion Rate (CVR), and Return on Advertising Spend (ROAS) based on data provided by the advertising platform to generate advertising performance reports, derive insights to enhance advertising performance, and establish improvement strategies.

[0004] Generally, advertising performance analysis proceeds through a process in which ad operators aggregate and analyze data collected from various platforms to compile performance reports; however, organizing and analyzing this data manually presents difficulties, such as the time required. Prior art literature

[0005] Registered Patent Publication No. 10-2747558, Method and apparatus for obtaining analysis results for medical data using a large-scale language model, Date of publication January 2, 2025. The problem to be solved

[0006] While it is possible to explore ways to utilize artificial neural network models, such as Large-Scale Language Models (LLM) capable of automatic text generation, for advertising performance analysis, there is a possibility of unfounded or data-independent results due to indiscriminate data input and hallucination issues during text generation.

[0007] According to an embodiment, a digital marketing data analysis method is provided that generates a document containing numerical information calculated based on the result of automatically analyzing data attributes of a tabular dataset collected for digital marketing performance analysis, and utilizes this document for digital marketing performance analysis, and a computing device for performing the same.

[0008] The problems to be solved by the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art to which the present invention pertains from the description below. means of solving the problem

[0009] A digital marketing data analysis method performed by a computing device according to a first aspect comprises: a step of generating at least one document containing numerical information for calculating advertising performance indicators based on the result of automatically analyzing data attributes of a tabular dataset collected from at least one digital advertising platform for digital marketing performance analysis; a step of vectorizing (embedding) the generated at least one document and storing it in a vector database to enable retrieval-augmented generation (RAG) based search; a step of extracting semantically similar reference documents through RAG-based search in the vector database in response to an input natural language-based query; and a step of inputting the query and the reference documents into an artificial neural network model to provide the result of the digital marketing performance analysis and a list of documents referenced in the digital marketing performance analysis as the output of the artificial neural network model.

[0010] Here, when automatically analyzing the above data attributes, the campaign ID and time information can be automatically selected as row classification criteria and column classification criteria, and the above advertising performance indicators can be automatically selected as value assignment elements.

[0011] The digital marketing data analysis method described above may further include: a step of generating at least one reference matrix data through a data structure transformation in which the row classification criteria are arranged along the row axis and the column classification criteria are arranged along the column axis to map a matrix and the value assignment elements are placed at the intersection of the mapping; a step of extracting matrix data from the results of the digital marketing performance analysis; and a step of evaluating the reliability of the results of the digital marketing performance analysis through a verification of consistency between the generated reference matrix data and the extracted matrix data.

[0012] The above consistency verification can be performed by extracting the matrix data from the numeric value within the result of the digital marketing performance analysis and then verifying whether it matches the numeric value corresponding to the document list on the reference matrix data.

[0013] The above reference matrix data may be in the form of a pivot table.

[0014] The step of generating the above document may generate the document by removing some lower values ​​of the decimal values ​​of the value assignment elements to reduce the number of input tokens of the artificial neural network model.

[0015] A computing device according to the second aspect comprises: a memory in which a computer program including at least one instruction is stored; and a processor that executes said instruction; wherein the processor, by executing said instruction, generates at least one document containing numerical information that calculates advertising performance indicators based on the result of automatically analyzing data attributes of a tabular dataset collected for digital marketing performance analysis from at least one digital advertising platform; vectorizes said at least one document and stores it in a vector database to enable RAG-based search; extracts semantically similar reference documents through said RAG-based search in said vector database in response to an input natural language-based query; and inputs said query and said reference documents into an artificial neural network model to provide the result of said digital marketing performance analysis and a list of documents referenced in said digital marketing performance analysis as the output of said artificial neural network model.

[0016] According to a third perspective, a computer program stored on a computer-readable recording medium includes instructions for the computer program to cause a processor to perform the digital marketing data analysis method.

[0017] According to the fourth aspect, a computer-readable recording medium in which a computer program is stored includes instructions for the computer program to cause a processor to perform the digital marketing data analysis method. Effects of the invention

[0018] According to an embodiment, a document containing numerical information calculated based on the results of automatically analyzing data attributes of a tabular dataset collected for digital marketing performance analysis is generated and utilized for digital marketing performance analysis, thereby resolving the hallucination problem caused by artificial neural network models such as large-scale language models.

[0019] In addition, according to an embodiment, a document containing the numerical information is vectorized and processed to enable RAG-based search, and a reference document corresponding to the query is extracted through the search and utilized for digital marketing performance analysis, thereby allowing the reference document to be used as a basis for the results of the digital marketing performance analysis.

[0020] In addition, according to the embodiment, there is an effect of enabling the evaluation of the reliability of the results of the digital marketing performance analysis by generating reference matrix data based on the results of the automatic analysis of the data attributes and verifying the consistency by comparing it with matrix data extracted from the results of the digital marketing performance analysis. Brief explanation of the drawing

[0021] FIG. 1 is a configuration diagram of a computing device for performing a digital marketing data analysis method according to an embodiment of the present invention. FIG. 2 is a configuration diagram showing the functional classification of computer programs loaded by a processor of a computing device according to an embodiment of the present invention. FIG. 3 is a flowchart illustrating a digital marketing data analysis method according to an embodiment of the present invention. Specific details for implementing the invention

[0022] The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims.

[0023] The terms used in this specification will be briefly explained, and the invention will be described in detail.

[0024] The terms used in this invention have been selected based on currently widely used general terms, taking into account their functions within the invention; however, these terms may vary depending on the intent of those skilled in the art, case law, the emergence of new technologies, etc. Additionally, in specific cases, terms have been arbitrarily selected by the applicant, and in such cases, their meanings will be described in detail in the relevant description of the invention. Therefore, the terms used in this invention should be defined not merely by their names, but based on their meanings and the overall content of the invention.

[0025] When a part of a specification is described as 'comprising' a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0026] Additionally, the term "part" as used in the specification refers to software or hardware components, such as FPGAs or ASICs, and the "part" performs certain roles. However, the meaning of "part" is not limited to software or hardware. The "part" may be configured to reside in an addressable storage medium or configured to operate one or more processor parts. Thus, by example, the "part" includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functions provided within the components and "parts" may be combined into a smaller number of components and "parts" or further separated into additional components and "parts."

[0027] Below, embodiments of the present invention are described in detail with reference to the attached drawings so that those skilled in the art can easily implement the invention. Additionally, parts of the drawings that are irrelevant to the description are omitted to clearly explain the invention.

[0028] FIG. 1 is a configuration diagram of a computing device for performing a digital marketing data analysis method according to an embodiment of the present invention, and FIG. 2 is a configuration diagram functionally classifying a computer program loaded by a processor of a computing device according to an embodiment of the present invention.

[0029] Referring to FIGS. 1 and 2, a computing device (100) according to an embodiment includes a memory (110) in which a computer program (111) is stored and a processor (120), and may further include an input unit (130) and / or an output unit (140).

[0030] A computer program (111) stored in the memory (110) of such a computing device (100) includes at least one instruction that can be executed by a processor (120). By executing the instruction of such a computer program (111) by the processor (120), the computer program (111) is loaded so that a digital marketing data analysis method according to an embodiment of the present invention can be performed. In addition, various data necessary for the processor (120) to perform a digital marketing data analysis method according to an embodiment of the present invention may be stored in the memory (110).

[0031] The processor (120) performs a digital marketing data analysis method according to the embodiment by executing instructions of a computer program (111) stored in memory (110) and loading the computer program (111). The functions of the data attribute analysis unit (210), reference generation unit (220), document vectorization processing unit (230), database search unit (240), performance analysis unit (250), and consistency verification unit (260) based on the computer program (111) thus loaded will be explained again below.

[0032] The input unit (130) receives various data necessary for the processor (120) of the computing device (100) to perform a digital marketing data analysis method according to an embodiment and provides it to the processor (120). For example, the input unit (130) may provide a table-shaped dataset collected for digital marketing performance analysis from at least one digital advertising platform to the processor (120). Additionally, the input unit (130) may receive a natural language-based query and provide it to the processor (120).

[0033] The output unit (140) can provide various data obtained and / or generated as the processor (120) of the computing device (100) performs the digital marketing data analysis method according to the embodiment to the outside. For example, the output unit (140) can provide the results of the digital marketing performance analysis and a list of documents referenced in the digital marketing performance analysis. Additionally, the output unit (140) can provide the results of the reliability evaluation of the results of the digital marketing performance analysis. For example, the output unit (140) providing various data, a list of documents, or reliability evaluation results may include outputting to a peripheral device connected via a serial interface, etc., or transmitting via a communication channel.

[0034] When a computer program (111) loaded by a processor (120) performs the function of a data attribute analysis unit (210), it generates at least one document containing numerical information that calculates advertising performance indicators based on the result of automatically analyzing data attributes of a tabular dataset collected for digital marketing performance analysis from at least one digital advertising platform. Here, when the data attribute unit (210) automatically analyzes data attributes, it can automatically select campaign ID and time information as row classification criteria and column classification criteria, and can automatically select advertising performance indicators as value assignment elements. Additionally, when generating a document, the data attribute analysis unit (210) can generate the document by removing some lower values ​​of the decimal values ​​of the value assignment elements to reduce the number of input tokens of the artificial neural network model.

[0035] When a computer program (111) loaded by a processor (120) performs the function of a reference generation unit (220), it generates at least one reference matrix data through data structure transformation in which row classification criteria are arranged along the row axis and column classification criteria are arranged along the column axis, a matrix is ​​mapped, and value assignment elements are placed at the intersection of the mappings. Here, the reference matrix data may be in the form of a pivot table.

[0036] When a computer program (111) loaded by a processor (120) performs the function of a document vectorization processing unit (230), at least one document generated by a data attribute analysis unit (210) is vectorized (embedding) and stored in a vector database to enable retrieval-augmented generation (RAG) based search.

[0037] When a computer program (111) loaded by a processor (120) performs the function of a database search unit (240), it extracts semantically similar reference documents through a RAG-based search in a vector database in response to a natural language-based query input through an input unit (130).

[0038] When a computer program (111) loaded by a processor (120) performs the function of a performance analysis unit (250), it inputs reference documents extracted by a query and a database search unit (240) into an artificial neural network model and provides the result of digital marketing performance analysis and a list of documents referenced in the digital marketing performance analysis as the output of the artificial neural network model.

[0039] When a computer program (111) loaded by a processor (120) performs the function of a consistency verification unit (260), matrix data is extracted from the results of digital marketing performance analysis, and then the reliability of the results of digital marketing performance analysis is evaluated through consistency verification between the reference matrix data generated by the reference generation unit (220) and the previously extracted matrix data. For example, when performing consistency verification, the consistency verification unit (260) can extract matrix data from numeric values ​​within the results of digital marketing performance analysis and then check whether it matches the numeric values ​​corresponding to the document list on the reference matrix data generated by the reference generation unit (220).

[0040] FIG. 3 is a flowchart illustrating a digital marketing data analysis method according to an embodiment of the present invention.

[0041] Hereinafter, a digital marketing data analysis method performed by a computing device (100) according to an embodiment will be described in detail with reference to FIGS. 1 to 3.

[0042] First, the processor (120) of the computing device (100) can perform the digital marketing data analysis method according to the embodiment by executing instructions of the computer program (111) stored in memory (110) and loading the computer program (111). The computer program (111) thus loaded performs the functions of a data attribute analysis unit (210), a reference generation unit (220), a document vectorization processing unit (230), a database search unit (240), a performance analysis unit (250), and a consistency verification unit (260).

[0043] The input unit (130) of the computing device (100) collects a table-shaped dataset from at least one digital advertising platform for digital marketing performance analysis and provides it to the processor (120). Of course, the table-shaped dataset required for digital marketing performance analysis may be collected in advance and stored in memory (110), or the processor (120) may extract the table-shaped dataset from memory (110). For example, the digital advertising platform may include Google Ads, which displays advertisements through search networks and display networks, or Meta Ads, which displays advertisements through social services. For example, the table-shaped dataset may be composed of rows and columns, each row may represent an ad impression, click, or campaign ID, and the columns may contain attributes (time information, features, metrics, etc.) for that.

[0044] And, when the computer program (111) performs the function of the data attribute analysis unit (210), it automatically analyzes the data attributes of a table-shaped dataset collected for digital marketing performance analysis from at least one digital advertising platform. Here, when the data attribute unit (210) automatically analyzes the data attributes, it can automatically select the campaign ID and time information as row classification criteria and column classification criteria, and can automatically select advertising performance indicators such as the click-through rate (CTR), conversion rate (CVR), and return on advertising spend (ROAS) as value assignment elements (S310).

[0045] In addition, when a computer program (111) loaded by a processor (120) performs the function of a reference generation unit (220), it generates at least one reference matrix data through data structure transformation, which arranges row classification criteria along the row axis, arranges column classification criteria along the column axis to map the matrix, and places value assignment elements at the intersection of the mapping. Here, the reference matrix data may be in the form of a pivot table. Such reference matrix data may be used for reliability evaluation by step S370, which will be described below. For example, the reference generation unit (220) may generate reference matrix data in the form of a pivot table containing daily ad click-through rates (CTR) by campaign (S320).

[0046] And, the data attribute analysis unit (210) generates at least one document containing numerical information that calculates advertising performance indicators such as click-through rate (CTR), conversion rate (CVR), and return on advertising spend (ROAS) based on the results of automatic analysis in step S310. When analyzing digital marketing performance using an artificial neural network model in step S360, which will be described below, the generated document generates analysis results based on numerical information that calculates advertising performance indicators such as click-through rate (CTR), conversion rate (CVR), and return on advertising spend (ROAS), thereby eliminating the possibility of unfounded results or results not based on data due to hallucination problems.

[0047] Additionally, the data attribute analysis unit (210) can generate a document by removing some lower values ​​of the decimal values ​​of the value assignment elements to reduce the number of input tokens of the artificial neural network model when generating the document (S330).

[0048] And, when the computer program (111) performs the function of the document vectorization processing unit (230), it vectorizes at least one document generated by the data attribute analysis unit (210) and stores it in a vector database to enable RAG-based searching. For example, the vector database may be located in memory (110) (S340).

[0049] Next, when the computer program (111) performs the function of the database search unit (240), it extracts semantically similar reference documents through a RAG-based search in the vector database in response to a natural language-based query input through the input unit (130) (S350).

[0050] When the computer program (111) performs the function of the performance analysis unit (250), the reference documents extracted by the query and database search unit (240) are input into an artificial neural network model to provide the result of the digital marketing performance analysis and a list of documents referenced in the digital marketing performance analysis as the output of the artificial neural network model. For example, the performance analysis unit (250) can provide comment sentences as the result of the digital marketing performance analysis and can present the basis for the creation of the comment sentence by providing a list of documents used together (S360).

[0051] Additionally, when the computer program (111) performs the function of the consistency verification unit (260), it extracts matrix data from the results of the digital marketing performance analysis and evaluates the reliability of the results of the digital marketing performance analysis through consistency verification between the reference matrix data generated by the reference generation unit (220) and the previously extracted matrix data. For example, when performing consistency verification, the consistency verification unit (260) can extract matrix data from numerical values ​​within the results of the digital marketing performance analysis and then check whether it matches numerical values ​​corresponding to the document list on the reference matrix data generated by the reference generation unit (220) (S370).

[0052] Meanwhile, according to an embodiment, a computer program stored on a computer-readable recording medium includes at least one instruction for a processor of a computing device to perform each step included in the digital marketing data analysis method according to the aforementioned embodiments.

[0053] Additionally, according to an embodiment, a computer-readable recording medium storing a computer program includes at least one instruction in the computer program for causing a processor of a computing device to perform each step included in the digital marketing data analysis method according to the aforementioned embodiments.

[0054] As explained above, according to an embodiment of the present invention, a document containing numerical information for calculating advertising performance indicators based on the results of automatically analyzing data attributes of a tabular dataset collected for digital marketing performance analysis is generated and utilized for digital marketing performance analysis, thereby resolving the hallucination problem caused by artificial neural network models such as large-scale language models.

[0055] In addition, according to an embodiment, a document containing the numerical information is vectorized and processed to enable RAG-based search, and a reference document corresponding to the query is extracted through the search and utilized for digital marketing performance analysis, thereby allowing the reference document to be used as a basis for the results of the digital marketing performance analysis.

[0056] In addition, according to the embodiment, there is an effect of enabling the evaluation of the reliability of the results of the digital marketing performance analysis by generating reference matrix data based on the results of the automatic analysis of the data attributes and verifying the consistency by comparing it with matrix data extracted from the results of the digital marketing performance analysis.

[0057] Combinations of each step of each flowchart attached to the present invention may be performed by computer program instructions. Since these computer program instructions may be loaded into the processor of a general-purpose computer, a special-purpose computer, or other programmable data processing equipment, the instructions performed through the processor of the computer or other programmable data processing equipment create means for performing the functions described in each step of the flowchart. Since these computer program instructions may also be stored in a computer-available or computer-readable recording medium that can be directed toward the computer or other programmable data processing equipment to implement the function in a specific manner, the instructions stored in the computer-available or computer-readable recording medium may also produce a manufactured item containing instruction means for performing the function described in each step of the flowchart. Since computer program instructions can be loaded onto a computer or other programmable data processing equipment, instructions that execute a computer or other programmable data processing equipment by performing a series of operation steps on the computer or other programmable data processing equipment to create a process executed by the computer can also provide steps for executing the functions described in each step of the flowchart.

[0058] Additionally, each step may represent a module, segment, or part of code containing one or more executable instructions for executing a specified logical function(s). Also, it should be noted that in some alternative embodiments, the functions mentioned in the steps may occur out of order. For example, two steps described in succession may actually be performed substantially simultaneously, or the steps may sometimes be performed in reverse order according to the corresponding function.

[0059] The above description is merely an illustrative explanation of the technical concept of the present invention, and those skilled in the art to which the present invention pertains will be able to make various modifications and variations within the scope of the essential quality of the present invention. Accordingly, the embodiments disclosed in the present invention are intended to explain, not limit, the technical concept of the present invention, and the scope of the technical concept of the present invention is not limited by such embodiments. The scope of protection of the present invention shall be interpreted by the claims below, and all technical concepts within the equivalent scope shall be interpreted as being included within the scope of rights of the present invention. Explanation of the symbols

[0060] 100: Computing device 110: Memory 111: Computer program 120: Processor 130: Input section 140: Output section 210 : Data Attribute Analysis Department 220: Reference creation section 230: Document Vectorization Processing Unit 240: Database Search Section 250: Performance Analysis Department 260: Consistency Verification Unit

Claims

Claim 1 A method for analyzing digital marketing data performed by a computing device, comprising: a step of automatically analyzing data attributes of a tabular dataset collected from at least one digital advertising platform for digital marketing performance analysis and automatically selecting row classification criteria, column classification criteria, and value assignment elements; a step of generating at least one reference matrix data through data structure transformation in which the row classification criteria are arranged along the row axis and the column classification criteria are arranged along the column axis to map a matrix and the value assignment elements are placed at the intersection of the mapping; a step of generating at least one document containing numerical information in which advertising performance indicators are calculated by the value assignment elements based on the result of the automatic analysis; a step of vectorizing at least one document generated based on the result of the automatic analysis of data attributes and storing it in a vector database; a step of extracting a reference document from the vector database in which the document containing the numerical information is vectorized and stored in response to an input natural language-based query; and a step of inputting the query and the reference document into an artificial neural network model and, based on the document containing the numerical information, as the output of the artificial neural network model, the digital marketing performance A step of providing the results of the analysis and a list of documents referenced in the digital marketing performance analysis; a step of extracting matrix data from the results of the digital marketing performance analysis and then evaluating the reliability of the results of the digital marketing performance analysis through verification of consistency with the generated reference matrix data;A digital marketing data analysis method comprising the step of outputting at least one of the result of the digital marketing performance analysis, the list of documents referenced in the digital marketing performance analysis, and the result of evaluating reliability, wherein the consistency verification involves extracting the matrix data from the numeric value within the result of the digital marketing performance analysis output by the artificial neural network model in response to the query, and then verifying whether it matches the numeric value corresponding to the list of documents on the reference matrix data. Claim 2 A digital marketing data analysis method according to claim 1, wherein the automatic selection step comprises automatically selecting campaign ID and time information among data attributes included in the table-shaped dataset based on the row classification criteria and the column classification criteria, and automatically selecting at least one of the ad click-through rate (CTR), conversion rate (CVR), and return on advertising spend (ROAS) as the value assignment factor. Claim 3 A digital marketing data analysis method according to claim 1, wherein the step of storing in the vector database is processed to enable retrieval-augmented generation (RAG) based search, and the step of extracting the reference document is to extract the semantically similar reference document through the RAG-based search. Claim 4 A digital marketing data analysis method according to claim 1, wherein the reference matrix data is in the form of a pivot table. Claim 5 A digital marketing data analysis method according to claim 1, wherein the step of generating the document is to generate the document by removing some lower values ​​of the decimal values ​​of the value assignment elements to reduce the number of input tokens of the artificial neural network model. Claim 6 Memory storing a computer program including at least one instruction; a processor executing the instruction; The processor includes an output unit, and by executing the command, the processor automatically analyzes the data attributes of a tabular dataset collected for digital marketing performance analysis from at least one digital advertising platform and automatically selects row classification criteria, column classification criteria, and value assignment elements; generates at least one reference matrix data through data structure transformation in which the row classification criteria are arranged along the row axis and the column classification criteria are arranged along the column axis to map a matrix and the value assignment elements are placed at the intersection of the mapping; generates at least one document containing numerical information in which advertising performance indicators are calculated by the value assignment elements based on the result of the automatic analysis; vectorizes the at least one document generated based on the result of the automatic analysis of the data attributes and stores it in a vector database; in response to an input natural language-based query, extracts a reference document from the vector database in which the document containing the numerical information is vectorized and stored; inputs the query and the reference document into an artificial neural network model, and based on the document containing the numerical information, the result of the digital marketing performance analysis and as the output of the artificial neural network model Provides a list of documents referenced in the digital marketing performance analysis, extracts matrix data from the results of the digital marketing performance analysis, and evaluates the reliability of the results of the digital marketing performance analysis through consistency verification with the generated reference matrix data, and outputs at least one of the results of the digital marketing performance analysis, the list of documents referenced in the digital marketing performance analysis, and the result of evaluating reliability through the output unit.A computing device that performs the above consistency verification by extracting the matrix data from the numeric value within the result of the digital marketing performance analysis output by the artificial neural network model in response to the above query, and then verifying whether it matches the numeric value corresponding to the document list on the reference matrix data. Claim 7 In claim 6, the processor automatically selects campaign ID and time information among the data attributes included in the table-shaped dataset based on the row classification criteria and the column classification criteria, and automatically selects at least one of the ad click-through rate (CTR), conversion rate (CVR), and return on advertising spend (ROAS) as the value assignment factor. Claim 8 In claim 6, the processor processes the at least one document to enable retrieval-augmented generation (RAG)-based search when storing it in the vector database, and extracts semantically similar reference documents through the RAG-based search when extracting the reference documents. Claim 9 In claim 6, the reference matrix data is in the form of a pivot table in a computing device. Claim 10 In claim 6, the processor is a computing device that generates the document by removing a lower part of the decimal value of the value assignment element to reduce the number of input tokens of the artificial neural network model when generating the document. Claim 11 A computer program stored on a computer-readable recording medium, wherein the computer program comprises instructions for the processor to perform a digital marketing data analysis method of any one of claims 1 to 5 by the processor executing at least one instruction. Claim 12 A computer-readable recording medium storing a computer program, wherein the computer program includes instructions for the processor to perform a digital marketing data analysis method of any one of claims 1 to 5 by the processor executing at least one instruction.

Images