Information processing device and program
Patent Information
- Application Number
- JP2025029438
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-02-26
- Publication Date
- 2026-08-27
Smart Images
Figure 2026137626000001_ABST
Abstract
Description
Technical Field
[0005] , , , , , ,
[0001] The present disclosure relates to an information processing apparatus and a program.
Background Art
[0002] In recent marketing, the affinity between brands has been emphasized. The affinity between brands refers to a relationship in which different brands share common values and images, and a synergistic effect is produced by their coincidence. It is expected that by cooperating between brands with high affinity, the reliability and attractiveness of the brands will be enhanced and they will have a stronger influence. Also, brand affinity can be utilized when searching for influencers suitable for one's own brand. It is effective to adopt an influencer related to another brand with high affinity to one's own brand as an influencer of one's own brand.
[0003] In conventional methods for estimating the affinity between brands, the following elements are used.
[0004] · Commonality of brand purchasers · Overlap of the customer bases (age, gender, income, place of residence, values, etc.) of both brands · Commonality of the scenes in which the brands are used · Similarity of brand images · Overlap of SNS followers On the other hand, if a collaboration strategy is forcibly promoted despite low affinity between brands, it may cause confusion among consumers and sales may not increase as expected. Moreover, risks such as damage to the brand image and decline in brand value also occur, so a careful judgment is required.
Summary of the Invention
Problems to be Solved by the Invention
[0005] A new method for estimating with high reliability the affinity between a user and a brand with which the user has not posted is expected. [Means for solving the problem]
[0006] The information processing device according to this embodiment comprises a storage unit for storing a program, a processor for executing the program, and a communication unit for communicating with an SNS server device that provides SNS services via an internet line. The processor, by executing the program, functions as a means for requesting the SNS server device to send a plurality of words attached with a hashtag for each of a plurality of posts by a user using the SNS service, and for receiving such a program; a means for extracting a word representing a first brand (first brand word) from the plurality of words; a means for calculating the ratio of the number of posts in which the first brand word is attached with a hashtag to the total number of posts by the user (first brand posting rate); and a means for calculating an affinity score between the user and the second brand (user / second brand affinity score) by multiplying the first brand posting rate by a score representing the affinity between the first brand and the second brand that is not attached to the user's posts (first / second brand affinity score). [Brief explanation of the drawing]
[0007] [Figure 1] Figure 1 is a diagram showing the overall system configuration including the information processing device according to this embodiment. [Figure 2] Figure 2 shows the physical configuration of the information processing device shown in Figure 1. [Figure 3] Figure 3 shows the functional configuration of the processor in Figure 2. [Figure 4] Figure 4 shows an example of a user database of SNS users stored in the memory unit shown in Figure 2. [Figure 5] Figure 5 shows an example of a brand list stored in the memory unit shown in Figure 2. [Figure 6] Figure 6 shows an example of a profile screen in a social networking service (SNS). [Figure 7] Figure 7 shows an example of a post details screen in a social networking service. [Figure 8] Figure 8 is a flowchart showing the processing procedure performed by the processor in Figure 2. [Figure 9] Figure 9 shows an example of a list of hashtag words collected in step S12 of Figure 8. [Figure 10] Figure 10 shows an example of the frequency of occurrence of co-occurring words aggregated in step S15 of Figure 8. [Figure 11] Figure 11 shows an example of a vector relating to the frequency of occurrence of co-occurring words generated in step S16 of Figure 8. [Figure 12] Figure 12 shows an example of a diagram illustrating the affinity scores between brands calculated in step S16 of Figure 8. [Figure 13] Figure 13 is a supplementary diagram that supplements the processing flow of steps S18, S19, and S20 in Figure 8. [Figure 14] Figure 14 shows an example of a survey report generated in step S21 of Figure 8. [Modes for carrying out the invention]
[0008] Embodiments of the present invention will be described below with reference to the drawings. Figure 1 shows the configuration of the overall system including the information processing device according to this embodiment. The information processing device 1 according to this embodiment is connected to SNS server devices 3-1 and 3-2 that provide SNS services via an internet line 2.
[0009] As shown in Figure 2, the information processing device 1 has a processor 11. RAM 12, ROM 13, storage unit 14, input device 15, display 16, and communication unit 17 are connected to the processor 11 via a system bus 10. The processor 11 is composed of, for example, a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit). The processor 11 executes the "brand affinity score estimation processing program" loaded from the storage unit 14 into the RAM 12.
[0010] RAM 12 functions as the main memory, work area, etc., of the processor 11. ROM 13 or storage unit 14 stores the BIOS (Basic Input Output System) executed by the processor 11, the operating system program (OS), a user database (see Figure 4) for multiple users of the SNS service, a "brand affinity score estimation processing program," programs for realizing various other functions, and various data required for these processes. Input devices 15 consist of a keyboard (KB), a mouse, a touch panel, or other pointing devices. The display 16 is typically implemented as an LCD (Liquid Crystal Display).
[0011] Figure 3 shows the functional configuration of the processor 11. By executing the "brand affinity score estimation processing program," the processor 11 functions as a control unit 21, a memory management unit 22 that manages writing and reading from the memory unit 14, a user extraction unit 23, a hashtag word collection unit 24, a brand word extraction unit 25, a co-occurring word extraction unit 26, a co-occurring word counting unit 27, a brand affinity score calculation unit 28, a brand posting rate calculation unit 29, a user / brand affinity score calculation unit 30, and a report editing unit 31.
[0012] The user extraction unit 23 extracts multiple users to be applied to the brand affinity score estimation process from a user database of multiple users of the SNS service stored in the storage unit 14, based on extraction conditions specified by the researcher, such as demographic information such as gender and age (age group), as well as items such as hobbies and preferences. As illustrated in Figure 4, the user database consists of predetermined items such as accounts, number of posts, demographic information (gender, age, place of residence), hobbies, and preferences for each of the multiple users of the SNS service. The processor 11 periodically requests the SNS server devices 3-1 and 3-2 to provide information on the above items for multiple users of the SNS service. The storage management unit 22 updates the user database in the storage unit 14 based on the information on the above items received from the SNS server devices 3-1 and 3-2. The processor 11 may also send a user extraction request to the SNS server devices 3-1 and 3-2, along with extraction conditions for arbitrary items such as demographic information such as gender and age (age group), as well as hobbies and preferences, specified by the researcher, when starting the brand affinity score estimation process, and receive user information.
[0013] Figure 6 shows a profile screen including a list of posts by a user, and Figure 7 shows a detail screen of a post. For each post, multiple words are set, each preceded by the hashtag "#". The hashtag word collection unit 24 uses the accounts of multiple extracted users to request and receive from SNS server devices 3-1 and 3-2 the multiple words (hereinafter referred to as hashtag words) that have been attached to each of the multiple posts made by each of those users, along with the hashtag "#". The memory management unit 22 stores the received hashtag words in the memory unit 14, associating them with the user account, post ID, and post date (see Figure 9).
[0014] The memory unit 14 stores a brand list illustrated in FIG. 5. The brand list is a list that tabulates the names of a plurality of brands (strings written in their original official notations). The names of the brands are associated with strings written in various notations different from the original notations, and strings representing general names or aliases (common names) naturally used among ordinary people.
[0015] The brand word extraction unit 25 extracts a hash tag word (referred to as a brand word) representing any brand from the hash tag words by querying the brand list. Since the brand list associates the original notation (brand name) with notations different from the original notation and common names, even if the user writes the hash tag words in arbitrary notations respectively, the brand words can be extracted almost without omission from the hash tag words. Further, the brand word extraction unit 25 counts the number of occurrences of the brand words for each user.
[0016] The co-occurring word extraction unit 26 extracts, as co-occurring words, hash tag words (non-brand words) used together with the brand words, in other words, hash tag words (non-brand words) attached with hash tags for the same content as the brand words, and associates them with the relevant brand. The non-brand words represent the image or awareness that the user who posts the brand words has towards the brand, the products or services considered to be related to the brand, the target layer or usage scenario that the user has in mind, and so on.
[0017] The co-occurring word counting unit 27 classifies all the non-brand words posted by the plurality of extracted users for each brand word (brand), and counts the number of occurrences of each non-brand word (see FIG. 10). In other words, it obtains the distribution of the number of occurrences of non-brand words as co-occurring words for each brand.
[0018] The inter-brand affinity score calculation unit 28 calculates an affinity score representing the affinity between brands based on the frequency of occurrence of non-brand words. Specifically, it defines the number of types of non-brand words that appear across brand words used by all users in the survey as the degree, generates a vector whose element components are the frequency of occurrence of non-brand words counted for each brand word (see Figure 11), and calculates the affinity score as the cosine similarity between the vectors of any two brands. If the vector of a certain brand a is denoted as A and the vector of another brand b is denoted as B, the cosine similarity CS(AB) is calculated by the following formula: A·B represents the dot product of vector A and vector B. |A| represents the length (norm) of vector A, and |B| represents the length (norm) of vector B.
[0019] CS(AB) = A·B / |A||B| Cosine similarity is an index that measures how similar two vectors are. A vector relating to a brand has the frequency of occurrence of co-occurring words of the brand as its elements. Therefore, the cosine similarity CS(AB) between a vector A relating to brand a and a vector B relating to brand b represents the degree of similarity and commonality between brands a and b in terms of the image and perception that general users have of brands a and b, the products and services related to brands a and b, and the target audience and usage scenarios that users perceive for brands a and b. In other words, it represents the degree of affinity, the degree to which they are easily linked, compatible, and familiar with each other. The value of cosine similarity CS(AB) can be calculated within a range from a minimum value of "-1" to a maximum value of "+1". The affinity between brands is lowest at "-1" and higher as it approaches the maximum value of "+1".
[0020] While cosine similarity is the optimal affinity score, it is not limited to this method; for example, the Euclidean distance between vectors for each brand word could also be used.
[0021] The inter-brand affinity score calculation unit 28 can calculate multiple affinity scores for two different combinations of brands (brand pairs). The inter-brand affinity score calculation unit 28 outputs multiple affinity scores for multiple brand pairs in a unified manner, for example, as a correlation diagram (see Figure 12).
[0022] Furthermore, by setting multiple conditions for extracting users, user groups with different attributes can be extracted. For example, a user group of men in their 40s and a user group of women in their 20s can be extracted, and the inter-brand affinity score calculation unit 28 can calculate affinity scores for the same two brands, one from the user group of men in their 40s and the other from the user group of women in their 20s. By outputting these affinity scores, one from the user group of men in their 40s and the other from the user group of women in their 20s, side by side, it is possible to recognize the change in affinity scores depending on the user group.
[0023] The calculation (estimation) of affinity scores between brands can be extended to the calculation (estimation) of affinity scores between users and brands. The calculation (estimation) of affinity scores between users and brands is realized by the brand posting rate calculation unit 29 and the user / brand affinity score calculation unit 30. For example, if a user is an influencer for a certain brand, it is possible to estimate the degree of affinity that user has with other brands (such as the company's own brand) with which they have not had any previous involvement. This can be used, for example, to discover new influencers.
[0024] The brand posting rate calculation unit 29 calculates for each user the percentage of posts in which each brand word is attached with a hashtag (brand posting rate) out of the total number of posts made by that user. The brand posting rate indicates the proportion of posts made by each user for each brand. The user / brand affinity score calculation unit 30 calculates (estimates) the affinity score between the user and non-posted brand words by multiplying the affinity score (already calculated by the inter-brand affinity score calculation unit 28) between the brand words attached to the user's own posts (posted brand words) and brand words attached to other users' posts that are not attached to the user's posts (non-posted brand words) by the brand posting rate for the posted brand words.
[0025] The report editing department 31 creates a brand affinity survey report that includes the affinity scores between brands calculated by the inter-brand affinity score calculation unit 28 and the affinity scores between users and brands calculated by the user / brand affinity score calculation unit 30 (see Figure 14). A template is provided for this report. According to the template, the affinity scores between brands calculated by the inter-brand affinity score calculation unit 28 are laid out according to the template, along with interpretations that have been created in advance and stored in the storage unit 14, corresponding to the score value range. Furthermore, a correlation diagram is created and laid out that makes the brand names easy to understand visually, by placing brands with high affinity closer together and brands with low affinity further apart, based on the reciprocal of the affinity scores of multiple different combinations of two brands. In addition, for a specific user, the affinity score between that user and brands that the user has never posted to and has had little to no interaction with before is laid out according to the template.
[0026] Figure 8 shows the processing procedure by the processor 11. The researcher specifies user extraction conditions for items such as the number of posts, gender, age (age group), hobbies, and preferences. The user extraction unit 23 searches the user database exemplified in Figure 4 according to the user extraction conditions, and multiple users that match the user extraction conditions are extracted as users to be applied to the brand affinity score estimation process (S11). The user database is constructed based on information on items such as the number of posts, gender, age (age group), hobbies, and preferences of multiple users using the SNS service obtained from SNS server devices 3-1 and 3-2, and is also updated periodically.
[0027] Using the accounts of the extracted multiple users, the hashtag word collection unit 24 collects multiple words (hashtag words) attached to each of the multiple posts made by each of those users, along with the hashtag "#", from the SNS server devices 3-1 and 3-2 (S12). As illustrated in Figure 9, the hashtag words are stored in the storage unit 14, associated with the user account, post ID, and posting date.
[0028] The hashtag words are queried against the brand list in Figure 5. This extracts hashtag words (brand words) that represent any of the brands (S13). The brand list includes not only the original spelling but also various spellings and colloquialisms, so it is possible to extract almost all brand words. The frequency of occurrence of brand words is counted for each user by the brand word extraction unit 25.
[0029] Next, hashtag words (non-brand words), that is, hashtag words (non-brand words) that are attached with the hashtag # to the same content (posts) as the extracted brand words, are extracted as co-occurring words associated with the brand (S14). For example, in the example in Figure 7, the non-brand words "Sports," "Running," and "socker" are extracted along with the brand word "Niko." The words "Sports," "Running," and "socker" are co-occurring words with the brand "Niko," and the user perceives them as images of the brand "Niko."
[0030] As illustrated in Figure 10, the co-occurrence word counting unit 27 classifies all non-brand words (co-occurring words) related to all extracted user posts by brand word (brand), and counts the number of occurrences of each non-brand word (co-occurring word) (S15).
[0031] The inter-brand affinity score calculation unit 28 generates a vector, as illustrated in Figure 11, for each of the extracted brand words (brands), with the frequency of occurrence of non-brand words as an element (S16). The extracted brand words (brands) are combined in pairs, and an affinity score representing the affinity between the two brands is calculated based on the two vectors relating to those two brand words (S17). The affinity score is calculated as the cosine similarity between the two vectors.
[0032] As is well known, cosine similarity is an index that measures how similar two vectors are. Since a vector relating to a particular brand has the frequency of occurrence of co-occurring words related to the brand as an element, the cosine similarity between a vector relating to one brand and a vector relating to another brand can serve as an index to measure whether the two brands have similar or distant perceptions of the image and perceptions that general users have of the brands, the products and services related to brands a and b, and the target audience and usage scenarios that users perceive for the brands.
[0033] Next, using the affinity scores between brands, it is possible to calculate (estimate) the affinity score between a specific user and brands that the user is not currently interested in.
[0034] The brand posting rate calculation unit 29 calculates the brand posting rate for a given brand as the ratio of the number of posts made by each extracted user regarding a particular brand to the total number of posts made by that user (S18). A specific example is shown in Figure 13. Assume that user A is an influencer for a certain brand and has experience posting about brands 1, 2, and 3. Also assume that the brands extracted from the posts of all extracted users are 1, 2, 3, ..., 6. In other words, user A has never posted about brands 4, 5, and 6; in other words, user A has had no prior involvement with brands 4, 5, and 6.
[0035] What we want to estimate here is the affinity score between User A and each of the brands 4, 5, and 6. The user / brand affinity score calculation unit 30 extracts the brand affinity scores for two combinations of brands 1, 2, 3, ..., 6: brands 1, 2, 3 posted by user A and brands 4, 5, 6 not posted by user A. For each of brands 4, 5, and 6, the highest brand affinity score is extracted (S19). In this example, the highest brand affinity score for brand 4 is with brand 3. For brand 5, the highest brand affinity score is with brand 2. For brand 6, the highest brand affinity score is with brand 3.
[0036] The user / brand affinity score calculation unit 30 calculates the affinity score between a user and brands that user A has not posted about by multiplying the affinity score between the brands related to user A's posts and brands that user A has not posted about by the brand posting rate for the brands related to the posts (S20).
[0037] In the example in Figure 13, the affinity score between User A and Brand 4 (which User A has not posted about) is estimated by multiplying User A's posting rate for Brand 3 by the affinity score between Brand 3 (which User A has posted about) and Brand 4 (which User A has not posted about). In other words, User A's posting rate for Brand 3 can be said to represent the degree of User A's interest in Brand 3, so the affinity score between User A and Brand 4 can be indirectly estimated by discounting the affinity score between Brand 3 (which User A has posted about) and Brand 4 (which User A has not posted about) by User A's posting rate for Brand 3. Similarly, the affinity score between User A and Brand 5 (which User A has not posted about) is estimated by multiplying User A's posting rate for Brand 2 (which User A has posted about) by the affinity score between Brand 2 (which User A has posted about) and Brand 5 (which User A has not posted about). The affinity score between User A and Brand 6 (which User A has not posted about) is estimated by multiplying User A's posting rate for Brand 3 by the affinity score between Brand 3 and Brand 6.
[0038] Finally, the report editorial department 31 prepares the brand affinity survey report (S21). As illustrated in Figure 14, the calculated affinity scores between brands are laid out according to the brand affinity survey report template. An interpretation statement is pre-defined for each of the multiple score ranges, and the interpretation statement corresponding to the calculated affinity score between brands is written alongside the affinity score.
[0039] It is anticipated that a very large number of brands will appear in the posts of multiple users that have been extracted. The combinations of these numerous brands, each pairing, will also be extremely diverse. It is practical to extract a predetermined number of brands with high affinity scores and include only those brand combinations in the report. Furthermore, if there is a brand of particular interest, such as your own brand, it is also appropriate to calculate the affinity score between that brand and each of the other brands, and include only the brand combinations with the highest affinity scores in the report.
[0040] Below the affinity score and its interpretation, a visually easy-to-understand correlation diagram is provided, where brand names are placed at distances corresponding to their affinity scores. Since an affinity score closer to +1.0 indicates higher brand affinity and a score closer to -1.0 indicates lower brand affinity, the distance is set according to the reciprocal of the affinity score.
[0041] Furthermore, below the correlation diagram, affinity scores are laid out between specific users designated by the researchers from among the extracted users—for example, users who are envisioned as influencers for the company's own brand or users who are active as influencers for any brand with a large number of posts—and brands that those users have never posted about and have had little to no prior contact with.
[0042] As explained above, according to this embodiment, brand words included in hashtag words in an SNS service are extracted, hashtag words (non-brand words) attached to the same content along with the brand words are applied as co-occurring words of the brand words, and the degree of similarity (cosine similarity, etc.) between the distributions (vectors) of the frequency of occurrence of co-occurring words between brands is calculated as the affinity between brands, thereby enabling the estimation of the degree of affinity between brands with high reliability. Furthermore, it can be extended to estimate the degree of affinity between any user and brands that the user has not posted to, i.e., brands that are assumed to have little interest in the past. For example, it can be effectively used to search for brands to partner with and to confirm synergistic effects. It can also be used, for example, to search for and confirm influencers suitable for one's own brand.
[0043] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]
[0044] 1... Information processing device, 2... Internet connection, 3-1, 3-2... SNS server device, 21... Control unit, 22... Memory management unit, 23... User extraction unit, 24... Hashtag word collection unit, 25... Brand word extraction unit, 26... Co-occurring word extraction unit, 27... Co-occurring word aggregation unit, 28... Inter-brand affinity score calculation unit, 29... Brand posting rate calculation unit, 30... User / brand affinity score calculation unit, 31... Report editing unit.
Claims
1. It comprises a storage unit for storing a program, a processor for executing the program, and a communication unit for communicating with an SNS server device that provides SNS services via an internet connection. The processor executes the program, A means for requesting and receiving from the SNS server device the transmission of multiple words attached with hashtags for each of multiple posts made by a user using the aforementioned SNS service, A brand word extraction means for extracting a word representing a first brand (first brand word) from the aforementioned plurality of words, A means for calculating the ratio of the number of posts in which the first brand word is attached with a hashtag to the total number of posts by the user (first brand posting rate), An information processing device that functions as a means for calculating an affinity score between a user and the second brand (user / second brand affinity score) by multiplying the first brand posting rate by a score representing the affinity between the first brand and the second brand not attached to the user's posting (first / second brand affinity score).
2. The memory unit stores a brand list in which the name of the first brand is associated with multiple notations different from the name of the first brand and a common name that represents the first brand. The information processing apparatus according to claim 1, wherein the brand word extraction means extracts the first brand word by matching the word with the brand list.
3. The information processing apparatus according to claim 1, wherein the affinity score calculation means functions as a means for calculating an affinity score between the user and the third brand (user / third brand affinity score) by multiplying the first brand posting rate by the affinity score between the first brand and a third brand not attached to the user's posting (first / third brand affinity score).
4. The information processing apparatus according to claim 3, wherein the processor further functions as a means for generating a report that displays the affinity score between the user and the second brand together with the affinity score between the user and the third brand.
5. The affinity score between the first and second brands is the cosine similarity or Euclidean distance between a first vector whose elements are the frequency of occurrence of multiple words (co-occurring words) attached with hashtags to the same content as the first brand word, and a second vector whose elements are the frequency of occurrence of multiple words (co-occurring words) attached with hashtags to the same content as the second brand word. The information processing apparatus according to claim 3, wherein the affinity score between the first and third brands is the cosine similarity or Euclidean distance between the first vector and the third vector whose elements are the number of occurrences of a plurality of words (co-occurring words) attached to the same content as the third brand word together with hashtags.
6. A computer that communicates with an SNS server device that provides SNS services via an internet connection, A means for requesting and receiving from the SNS server device the transmission of multiple words attached with hashtags for each of multiple posts made by a user using the aforementioned SNS service, A means for extracting a word representing the first brand (first brand word) from the aforementioned multiple words, A means for calculating the ratio of the number of posts in which the first brand word is attached with a hashtag to the total number of posts by the user (first brand posting rate), A program that functions as an affinity score calculation means for calculating an affinity score between a user and the second brand (user / second brand affinity score) by multiplying the first brand posting rate by a score representing the affinity between the first brand and the second brand not attached to the user's posts (first / second brand affinity score).