Brand value intelligent evaluation method and system based on multi-source data fusion

Abstract

The present application provides a brand value intelligent evaluation method and system based on multi-source data fusion, the method comprising: obtaining brand events, public response summary and operating result data, and converting them into a standard brand observation record set; taking brand event observation as the starting point, calculating the attribution support degree of public response and operating result observation and candidate events, determining the unique attribution relationship, and constructing a brand influence unit set; determining the value adjustment state according to the observation composition of each unit, calculating the unit adjustment proportion, and generating a value adjustment result set; updating the current settled brand value according to the set and the settlement registration record, and outputting the evaluation result. The present application effectively avoids the problems of cross-source dispersion and cross-cycle repeated valuation of the same matter, and improves the accuracy and consistency of the evaluation.

Description

Technical Field

[0001] This invention belongs to the field of computer data processing technology, and in particular relates to a method and system for intelligent brand value assessment based on multi-source data fusion. Background Technology

[0002] As brand management activities gradually shift towards online, platform-based, and periodic monitoring, the data used for brand valuation has expanded from traditional financial data, market share, brand investment, and survey results to include corporate announcements, regulatory information, authorized news, search trends, review summaries, complaint summaries, and operational data such as sales volume, returns, conversion rates, and repeat purchases. Existing technologies typically perform data cleaning, indicator transformation, and weighting according to data source or indicator category, then use weighted aggregation, statistical analysis, or predictive models to generate brand value scores or changes. While this approach broadens the data coverage for brand valuation, its basic evaluation object remains a single data point or indicator, making it difficult to adapt to the reality that the same brand issue may generate multiple observations at different business stages. For example, a product quality issue might first generate corporate announcements, regulatory notices, or news records, followed by an increase in related searches, negative reviews, or complaints, and after a certain period, this might manifest as an increase in returns, a decrease in sales volume, or a decrease in repeat purchases. The aforementioned data originates from public information systems, public response platforms, and enterprise management systems, differing in content format, statistical methods, target audiences, and generation times. Ordinary text deduplication can only identify records with similar content, fixed time windows can only process data from similar time periods, and conventional correlation analysis struggles to determine whether different records originate from the same brand event. If valued separately according to news, search, complaints, and operational metrics, the same brand event would be split into multiple independent impacts, repeatedly affecting brand value. If directly aggregated according to evaluation cycles, the public response and operational results of the same event might fall into different cycles, causing the event to undergo a complete adjustment in subsequent cycles. Furthermore, brand event records, public response records, and operational result records play different roles in the formation of brand influence: event records indicate that the event has occurred, public response records indicate that the event has generated a market response, and operational result records indicate that the event has produced actual changes in operations. Existing technology treats these data as equal indicators, easily leading to short-term dissemination popularity, concentrated reposting, or abnormal changes from a single source directly translating into brand value adjustments. It also fails to explain the specific event, target audience, and multi-source data basis corresponding to a single adjustment. Existing methods lack a unique attribution mechanism and settlement registration mechanism centered on the same brand. Items that have already been valued may still be identified as new impacts and re-involved in the calculation when new reports, complaint records, or subsequent operating data are obtained. Summary of the Invention

[0003] This invention discloses a brand value intelligent evaluation method and system based on multi-source data fusion to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the first aspect of the present invention provides a brand value intelligent assessment method based on multi-source data fusion, the method comprising: Acquire brand event data, public response summary data, and business results data, and convert the brand event data, public response summary data, and business results data into a standard brand observation record set; Starting with brand event observation, calculate the degree of attribution support between public response observation and business results observation and candidate brand events, determine the unique attribution relationship based on the degree of attribution support, and construct a set of brand influence units; The value adjustment status is determined based on the observation composition of each brand influence unit in the brand influence unit set, the unit adjustment ratio is calculated, and a value adjustment result set is generated. Based on the aforementioned set of value adjustment results and brand value settlement registration records, the currently settled brand value is updated, and the brand value assessment result is output.

[0005] Furthermore, the conversion of the brand event data, the public response summary data, and the business results data into a standard brand observation record set includes: Based on the brand directory and object directory, brand identifiers and object identifiers are mapped to the brand event data, the public response summary data, and the business results data; Based on the data interface category and the recorded content, determine the observation type and observation category; the observation type includes brand event observation, public response observation, and business results observation; Observation results are formed based on the meaning of the data itself, and the observation time and direction of influence are uniformly determined; The standard brand observation record set is formed by writing the brand identifier, the object identifier, the observation type, the observation category, the observation content, the observation result, the observation time, the data source, the source record identifier, and the direction of influence.

[0006] Furthermore, the calculation of the degree of attribution support between public response observations and business results observations and candidate brand matters includes: Brand consistency is used as the gating condition. When the record matches the brand identity of the candidate item, the gating result is the first preset value; otherwise, it is the second preset value. Calculate the degree of correspondence between objects, the degree of correspondence between event chains, and the degree of correspondence between time. The degree of attribution support is obtained by combining the gating result, the degree of object correspondence, the degree of event chain correspondence, and the degree of time correspondence.

[0007] Further, determining the unique attribution relationship based on the degree of attribution support includes: Candidate items whose affiliation support level reaches a preset threshold will be included in the affiliation comparison. When a candidate item reaches the preset threshold, it is directly classified into that candidate item. When two candidate items reach the preset threshold, the candidate item with the highest degree of support for attribution is selected, and the unique attribution is determined when the difference between the highest and second-highest values ​​reaches the preset difference threshold. When the difference is less than the preset difference threshold, the attribution is determined in the following order: priority of specific object consistency, priority of direct event chain correspondence, and priority of observation time closer to the center position of the main related time period. If the above steps still fail to distinguish between the two brands, a candidate brand influence unit is created and marked as pending verification.

[0008] Furthermore, the construction of the brand influence unit set includes: Candidate items are merged based on source record identifiers, observation content, and time boundaries; Cross-source records that have the same brand identity, the same main target, the same brand event category, and correspond to the same original event will be merged into one brand influence unit. Generate a unique influence unit identifier for each brand influence unit; The influence direction of the brand influence unit is determined based on the brand event observation direction, public response observation direction, and business results observation direction.

[0009] Furthermore, determining the value adjustment state based on the observation composition of each brand influence unit in the brand influence unit set includes: When brand event observation, public response observation, and business results observation all exist, and the records and merging basis indicate that the three types of observations have a clear attribution relationship, it is determined to be a complete realization status; When brand event observation and business result observation have a consistent specific object, a direct event chain correspondence, and an effective time correspondence, they are determined to be in the direct realization state. When brand event observation and public response observation exist, but business results observation has not yet been formed, it is determined to be a market confirmation state; When only brand events are observed, the event is identified as being in the formation stage. When only public response observations or operational results observations are available, the status is determined to be pending verification.

[0010] Furthermore, the calculation unit adjusts the ratio, including: Obtain the basic adjustment ratio from the value adjustment rule table; Confirmation coefficients for matters obtained based on public response observations or direct fulfillment of relationships; The operating realization coefficient is obtained based on the object relationship, continuous status, and number of unidirectional operating indicators observed in the operating results; The impact chain closure coefficient is obtained based on the complete or direct realization status. The unit adjustment ratio is obtained by multiplying the basic adjustment ratio, the item confirmation coefficient, the business realization coefficient, and the impact chain closure coefficient.

[0011] Furthermore, updating the currently settled brand value based on the value adjustment result set and the brand value settlement registration record includes: The results that participated in the value adjustment of this period but have not yet formed a valid settlement registration in the value adjustment result set shall be regarded as the set of results to be settled; By using a composite rate of change relationship, the unit adjustment ratio and unit adjustment direction of each brand influence unit in the set of results to be settled are applied to the currently settled brand value to obtain the updated brand value; The settlement registration record, brand value adjustment details, and the updated brand value are written into the same database transaction; the database sets unique constraints on the influencing unit identifier and evaluation cycle.

[0012] Furthermore, the output brand value assessment results include: Save the brand logo, basic brand value, brand value before settlement, adjusted brand value, brand value change results, settled brand influence units, brand influence units not yet involved in the adjustment, and their corresponding value adjustment status; The brand value assessment results will be written into the brand value database; The evaluation terminal reads the brand value database and generates a brand value evaluation report; the brand influence units in the brand value evaluation report are arranged according to the earliest observation time.

[0013] In a second aspect, the present invention provides a brand value intelligent assessment system based on multi-source data fusion, the system comprising: The data access module is used to acquire brand event data, public response summary data, and business results data; The observation record standardization module is used to convert the brand event data, the public response summary data, and the business result data into a standard brand observation record set. The Brand Influence Unit Construction Module is used to calculate the degree of attribution support between public response observations and business results observations and candidate brand events, starting from brand event observations, and to determine a unique attribution relationship based on the degree of attribution support, thereby constructing a set of brand influence units. The value adjustment result generation module is used to determine the value adjustment status based on the observation composition of each brand influence unit in the brand influence unit set, calculate the unit adjustment ratio, and generate a value adjustment result set. The settlement registration module is used to update the currently settled brand value based on the value adjustment result set and the brand value settlement registration record; The evaluation results output module is used to output the brand value evaluation results.

[0014] The beneficial technical effects of the present invention are at least as follows: This invention addresses the problem of multiple heterogeneous records generated and repeatedly valued for the same brand event at different stages. By establishing a standard set of brand observation records and a unique attribution mechanism, heterogeneous data from different sources are grouped into a single brand influence unit, effectively avoiding the issues of the same event being scattered across sources, appearing across stages, and being repeatedly valued across periods. Simultaneously, the system establishes a unique brand value settlement registration record based on the influence unit identifier, ensuring that each change in brand value corresponds to a specific brand event, target, observation chain, and rule version. This eliminates the potential for duplicate attribution of cross-source records and inconsistencies in database adjustment details from a data processing perspective. The final evaluation result has a clear single settlement status, which can be queried, traced, and verified by the evaluation terminal, significantly improving the accuracy, consistency, and interpretability of brand value evaluation in a multi-source data fusion environment. Attached Figure Description

[0015] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0016] Figure 1 This is a flowchart of the intelligent brand value assessment method based on multi-source data fusion, as described in this invention.

[0017] Figure 2 This is a framework diagram of the brand value intelligent evaluation system based on multi-source data fusion, as described in this invention. Detailed Implementation

[0018] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0019] Figure 1 This is a flowchart of a brand value intelligent assessment method based on multi-source data fusion according to an embodiment of this application. Figure 1 As shown, the method includes the following steps: Step S101: Obtain brand event data, public response summary data, and business results data, and convert the brand event data, the public response summary data, and the business results data into a standard brand observation record set.

[0020] Specifically, the system acquires brand event data, public response summary data, and business results data directly related to changes in brand value. Brand event data is obtained through interfaces such as the enterprise announcement system, regulatory public information, and authorized news data. The fields retrieved include brand name, target audience, event content, event occurrence time, public disclosure time, record number, and data source, indicating specific events that have occurred for the brand, product, or service. Public response summary data is obtained through interfaces such as search trend summary, evaluation category summary, and complaint category summary. The fields retrieved include brand name, target audience, response category, current summary result, corresponding benchmark result, statistical period, and data source, indicating changes in market attention, evaluation trends, or complaint categories related to the relevant events. Business results data is obtained through the enterprise business management system's summary interface. The fields retrieved include brand name, product or service identifier, business indicator category, current summary result, corresponding benchmark result, statistical period, and data source, indicating changes in sales volume, returns, conversion rates, or repeat purchases for the relevant brand, product, or service. Both public response data and operational results data are aggregated and recorded according to brand, product, or service dimensions, with the collected fields limited to those required for brand value assessment. Before the assessment period begins, each data interface is configured with data source identifiers, allowed fields, data acquisition methods, data update times, and data object scope. The system periodically calls the interfaces according to fixed configurations or receives new records pushed by the business system, writing the acquired data to the temporary access area for the current assessment period. The system then checks core fields such as brand name, target audience, event content or indicator category, and observation time, supplementing with definitively identifiable content based on internal product codes, service codes, announcement numbers, or registered directories, forming a pending record with complete object and time information.

[0021] The system uses the brand directory and object directory provided by the Enterprise Master Data System to unify brands and their target objects. The brand directory stores the official brand name, common abbreviations, historical names, internal brand codes, and the correspondence between these names. The object directory stores product codes, product names, product series, service names, store names, and the hierarchical relationship between each object and the brand. When the input data contains internal brand codes, product codes, or service codes, the system prioritizes determining the brand and object identifiers based on these codes. When the input data only contains names, the system first processes spaces, character formats, common suffixes, and abbreviations according to the pre-registered naming rules in the directories, and then determines the corresponding identifiers based on the brand and object directories. Records where the brand can be identified but only the product series or the brand as a whole can be located are set as either product series objects or brand-wide objects, respectively, ensuring that the object identifiers remain consistent with the finest object range supported by the original data. For example, when a company announcement uses the official brand name, a search trend interface uses the brand abbreviation, and the business management system uses the internal brand code, the system unifies these three expressions into a single brand identifier based on the brand directory. When a news record describes a product series and a business results record describes a specific model within that series, the system retains both the product series object identifier and the specific model object identifier, along with the hierarchical relationships registered in the object directory, for use in determining the object association between the two records during subsequent brand influence unit construction. When the original record involves multiple products or services and the content of each object can be determined separately, the system generates separate pending records for each object; when the original record only determines the overall scope of the matter, the system creates a single pending record based on the product series or the brand as a whole.

[0022] The system determines the observation type and category based on the data interface category and record content. Company announcements, regulatory notices, product release records, and authorized news records are identified as brand event observations; summary records of search changes, evaluation category changes, and complaint category changes are identified as public response observations; summary records of sales changes, return changes, conversion changes, and repeat purchase changes are identified as business results observations. Each input record corresponds to one observation type. Brand event observation categories include product releases, quality issues, price adjustments, service interruptions, and public marketing activities; public response observation categories include search responses, evaluation responses, and complaint responses; and business results observation categories include sales changes, return changes, conversion changes, and repeat purchase changes. Structured interfaces determine observation categories based on the event type field or indicator type field; announcements and news texts determine observation categories based on pre-configured event expressions, object names, and their combination relationships within the same or adjacent sentences. For example, if the text simultaneously contains a known product object and event expressions such as recall, malfunction, or non-conformity, the record is identified as a quality issue; if the text simultaneously contains a product object and event expressions such as release, launch, or rollout, the record is identified as a product release. When the original text simultaneously records multiple clearly distinguishable objects and events, the system generates observation records separately according to the correspondence between objects and events; the causes, scope of impact, and handling measures in the same event are collectively organized into one observation entry. The observation content for brand event observations saves the event behavior, the target of action, and the scope of the event; the observation content for public response observations saves the response target and response category; and the observation content for business results observations saves the objects and indicator categories corresponding to the business indicators.

[0023] The system generates observation results based on the meaning of each data type. For brand event observations, the results retain the status of events that have occurred or been made public, as well as the scope and severity of events explicitly recorded in the original records. Public response observations compare the current aggregated results with benchmark results under the same data interface, object, observation category, and statistical caliber to determine whether the change is increasing, stable, or decreasing. The benchmark result prioritizes the corresponding aggregated result from the previous evaluation period. If records from the previous evaluation period are missing, the normal state range determined before the start of the evaluation period is used. The normal state range is determined by the historical aggregated results of the corresponding interface or verified evaluation period records and is saved along with the data source identifier, statistical caliber, and rule version. Business results observations adhere to the same object consistency and statistical caliber consistency requirements, comparing the current aggregated results with the corresponding benchmark results to determine the corresponding business change status. Search response, evaluation response, complaint response, sales changes, return changes, conversion changes, and repeat purchase changes each retain their own observation results, which are then used for different correlation judgments and value confirmation purposes in subsequent steps. When different interfaces use different value ranges for the same observation category, the system converts the original change results into a unified change state based on the interval correspondence configured for that interface before the start of the evaluation period. For example, after a product quality notification is made public, the complaint classification summary interface shows that the product quality complaints have entered an increasing range relative to their own historical benchmark, and the system generates a public response observation with the observation category of complaint response and the observation result of increase; the enterprise management system shows that the product return situation has entered an increasing range relative to the benchmark under the same statistical caliber, and the system generates an operating result observation with the observation category of return change and the observation result of increase. The two observation records retain their source, category, and change state respectively, and in step S102, they are combined with the quality event observation to determine whether they jointly reflect the same brand event.

[0024] The system uniformly determines the observation time and impact direction of each record. For brand event observations, the actual occurrence time of the event explicitly recorded in the original record is prioritized. If the original record only records the public disclosure time, the first public disclosure time of the company announcement or regulatory information is used. Public response observations and business result observations use the statistical cutoff time of the corresponding summary records and retain their respective evaluation periods. The system determines the impact direction based on the corresponding relationship between the event direction and the indicator direction configured before the start of the evaluation period. Quality issues and service interruptions correspond to a negative impact; product launches correspond to a positive impact or are undetermined based on the event content; price adjustments and public marketing activities determine the corresponding direction based on the rules for registered events. Increases in negative evaluations, quality complaints, and returns correspond to a negative impact; increases in positive evaluations, sales, and repeat purchases correspond to a positive impact. The impact direction, as the direction of change reflected in the observation record, is used in step S102 to determine the directional correspondence between different observations. The system then identifies duplicate records within the same data source. When two records have the same brand identifier, object identifier, observation type, observation category, observation time, and observation result, the record with complete original information and earlier reception time is retained. After completing the above processing, the system writes brand identifier, object identifier, observation type, observation category, observation content, observation result, observation time, data source, source record identifier, and influence direction for each valid record, forming a standard brand observation record set. The source record identifier is formed by combining the data source identifier and the original record number, used to distinguish different original records from the same source and to identify whether multiple sources cite the same public matter. The observation content stores the specific content corresponding to the event behavior, response object, or business indicator, used to distinguish multiple similar events occurring within a short period for the same brand and the same object. Step S102 uses this set as input to construct brand influence units using the above fields.

[0025] Step S102: Starting with brand event observation, calculate the degree of attribution support between public response observation and business results observation and candidate brand events, determine the unique attribution relationship based on the degree of attribution support, and construct a set of brand influence units.

[0026] Specifically, using the standard brand observation record set output in step S101 as input, the system reads the brand identifier, object identifier, observation type, observation category, observation content, observation result, observation time, data source, source record identifier, and direction of influence from each record. The system first selects brand event observations and establishes candidate items based on the brand identifier, object identifier, observation category, observation content, observation time, and source record identifier. When a company announcement or regulatory public information clearly records the time of occurrence, the target audience, and the scope of the event, the corresponding record is used as the starting point for the candidate item. When an authorized news record has the same brand, compatible target audience, and the same event category as a candidate item and cites the same announcement or regulatory record, it is used as the public dissemination record of that candidate item. The target audience of the candidate item is determined according to the target catalog. Specific product items cover the product itself, product series items cover the series of products registered in the catalog, and brand-wide items cover the target audience of the brand. Observation content is used to distinguish multiple events of the same brand, the same target audience, and the same category occurring at similar times. The source record identifier is used to identify whether multiple sources jointly point to the same original event.

[0027] The system establishes data correspondences between brand event categories, public response categories, and business result categories according to the brand event association rules determined before the start of the evaluation cycle. Quality events correspond to an increase in quality complaints, an increase in negative reviews, an increase in returns, or a decrease in sales; product launches correspond to an increase in new product searches, an increase in new product reviews, an increase in new product sales, or an increase in conversion rates; service interruptions correspond to an increase in service complaints, an increase in negative reviews, or a decrease in repeat purchases; price adjustments correspond to changes in price searches, changes in reviews, changes in sales, or changes in conversion rates. The main association period and extended association period for each event category are statistically formed based on the interval between the event occurrence time and the public response or business result occurrence time in historically verified brand event records. The main association period adopts the more concentrated coverage of the historical intervals for the corresponding event category, while the extended association period adopts the boundary of a broader historical interval. The specific boundaries are written into the rule version for the current evaluation cycle along with the brand event association rules.

[0028] For each public response observation and business outcome observation, the system calculates the degree of attribution support between it and each candidate item. This calculation restricts the overall attribution relationship based on the weakest of the necessary attribution conditions, and uses brand consistency as a gating condition, specifically: ; in, This indicates the degree of support for classifying the current standard brand observation record into the current candidate item; This indicates the brand-specific gating result; if the current record matches the brand logo of the candidate item, it is selected. Otherwise take ; Indicates the degree of correspondence between objects; when the specific objects are identical, take... When a specific product has a direct subordinate relationship with the product series covered by the candidate item, it is considered... When the overall brand record corresponds to the overall brand event, take In other cases, take ; This indicates the degree of correspondence in the event chain. It is used when the current observation category and observation results form a direct data correspondence with the brand event association rules. When forming indirect data correspondence, take When there is no data correspondence, take ; Indicating the degree of time correspondence, when the current observation time falls within the main related time period, it is taken as... When located within the extended associated time period, take If the event occurs earlier than the time of occurrence or exceeds the extended related period, then take... The above items all represent the degree of attribution, which are dimensionless values ​​with a range of values ​​of [value range missing]. to The degree of object correspondence, the degree of event chain correspondence, the degree of time correspondence, and the subsequent attribution threshold are all saved as part of the current evaluation cycle rule version, and the calculation results are used to determine the unique brand event to which the record belongs.

[0029] The level of belonging support reached Candidate items are then compared for their affiliation. If only one candidate item in the current record meets this condition, it is directly assigned to that candidate item; if multiple candidate items exist, the candidate item with the highest affiliation support is selected, with the difference between the highest and second-highest values ​​reaching a certain threshold. A unique attribution is determined at that time. Insufficient difference. When the specific object is consistent, the corresponding direct event chain is prioritized, and the observation time is closer to the center of the main related time period. If the above order still cannot distinguish them, a candidate brand influence unit is established based on the brand identifier, object identifier, observation category, observation content and observation time of the record, and it is marked as pending verification to avoid a single observation record being assigned to multiple brand influence units in the same evaluation period.

[0030] For example, a brand experiences a product recall and a brand-wide promotion within the same evaluation period, subsequently generating a public response observation of increased quality complaints for that model. This observation, along with the brand-specific gating results of the recall, is... The degree of object correspondence is The level of correspondence in the event chain is The degree of time correspondence is The degree of support is The degree to which this observation corresponds to the target audience of the brand's overall promotional activities is: The level of correspondence in the event chain is The degree of support is Therefore, this complaint is classified as a recall. Returns for the corresponding model are also classified as recalls in the same way, while increases in overall brand search activity are classified as promotions based on the brand's overall target audience and marketing activity chain.

[0031] After completing the observation attribution, the system merges candidate items based on source record identifiers, observation content, and time boundaries. Cross-source records with the same brand identifier, main target, and brand event category, and corresponding to the same announcement, regulatory record, or original event, are merged into one brand influence unit. When the same brand, target, and event category correspond to different source record identifiers or different event content, forming independent event time ranges, separate brand influence units are established. Each brand influence unit generates a unique influence unit identifier and saves the brand identifier, main target identifier, brand event observation, public response observation, business result observation, earliest observation time, last observation time, direction of influence, and record merging basis.

[0032] The direction of influence for brand influence units is determined according to the following rules: When the direction of brand event observation is clear and the directly corresponding business result observation is consistent with its direction, that common direction is adopted; when the direction of brand event observation is not determined, but public response observation and business result observation form a common direction, that common direction is adopted; when the direction of brand event observation conflicts with the direction of direct business result observation, or when existing observations cannot form a common direction, the direction of influence is recorded as undetermined. Records are structured and stored based on the degree of correspondence between objects, the degree of correspondence between event chains, the degree of correspondence between time, the degree of attribution support, the data source relationship, and the source record identifier relationship corresponding to each observation. Finally, a set of brand influence units is output, which is used in step S103 to determine the value adjustment status and value adjustment ratio of each brand influence unit.

[0033] Step S103: Determine the value adjustment status based on the observation composition of each brand influence unit in the brand influence unit set, calculate the unit adjustment ratio, and generate a value adjustment result set.

[0034] Specifically, using the brand influence unit set output in step S102 as input, the system reads the influence unit identifier, brand identifier, main object identifier, brand event observation, public response observation, business result observation, earliest observation time, last observation time, influence direction, and record merging basis for each brand influence unit. The system determines the value adjustment status based on the observation composition: when brand event observation, public response observation, and business result observation all exist, and the record merging basis indicates a clear attribution relationship among the three types of observations, it is determined to be in a complete realization status; when brand event observation and business result observation have a specific object consistency, a direct event chain correspondence, and an effective time correspondence, it is determined to be in a direct realization status; when brand event observation and public response observation exist but business result observation has not yet been formed, it is determined to be in a market confirmation status; when only brand event observation exists, it is determined to be in an event formation status; candidate brand influence units with only public response observation or business result observation are determined to be in a pending verification status. Complete realization status and direct realization status are included in the value adjustment ratio calculation, while other statuses result in results that are not currently included in the current evaluation cycle adjustment.

[0035] The system uses a pre-determined value adjustment rule table for the current assessment period to obtain calculation parameters. The value adjustment rule table sets basic adjustment ratios according to brand item category, main target scope, and item severity, and saves fixed values ​​corresponding to item confirmation status, operational realization status, and impact chain type. The basic adjustment ratio is formed based on historically confirmed and settled brand item records. Historical records are grouped by item category, target scope, and item severity, and the median of the effective historical adjustment ratios for each group is used as the corresponding basic adjustment ratio. For item categories lacking historical records, the corresponding level in the company's existing brand valuation rules is used. All basic adjustment ratios in the value adjustment rule table are non-negative and less than [a certain value]. And it is saved together with the rule version of the current evaluation cycle.

[0036] In the case of full fulfillment, the system selects the record with the highest degree of attribution support, the most direct correspondence with the main target, and the observation time falling within the main related period from the public response observations as the main public response observation. When two different data sources form the same direction of influence, the event confirmation coefficient... Pick When a data source forms a direct response, Pick When public response is stable or conflicts with the brand's direction, Pick In the direct realization scenario, brand event observations are derived from corporate announcements or publicly available regulatory information, and the observations of operating results have a direct object relationship, a direct event chain relationship, and a degree of attribution support. At that time, the confirmation coefficient of the item Pick .

[0037] The system selects records from the operational results observations that have the highest degree of attribution support, the most direct correspondence with the main target, and a direct data correspondence between operational indicators and brand matters as the main operational results observations. When two direct operational indicators form the same direction of influence, or when the same direct operational indicator maintains the same direction for two consecutive statistical periods, the operational realization coefficient is... Pick When a direct operating indicator changes within a statistical period, Pick When operating results are stable or conflict with the direction of brand matters, Pick Influences the chain closure factor Determined solely based on the type of observed chain; complete fulfillment status is taken as... Directly redeemable status Other states take The object relationships, event chain relationships, and time relationships have been determined through the degree of attribution support in step S102 and the selection of the main observation in this step. The influence chain closure coefficient is used to distinguish between the three types of complete observation chains and the direct chain from event to business results. , and All are dimensionless coefficients, with a range of values ​​of 1000. to And it is saved along with the rule version of the value adjustment rule table.

[0038] The unit adjustment ratio for the brand influence unit uses the weaker of the necessary conditions to represent the combined effectiveness of event confirmation and operational realization, and uses the influence chain closure coefficient to represent the chain type, specifically: ; in, The unit adjustment ratio of the current brand influence unit is always expressed as a non-negative ratio; This indicates the base adjustment percentage obtained from the value adjustment rule table based on the observation category, observation results, and main object identifier of the brand event observation; The confirmation coefficient for an item is obtained based on public response observations or direct fulfillment relationships. This represents the operational realization coefficient, which is obtained based on the relationships between the objects observed in the operational results, their continuous status, and the number of similar operational indicators. This indicates the factor affecting chain closure, obtained based on either the complete realization status or the direct realization status. Because... Non-negative and less than , , and All located in to ,therefore It is a dimensionless, non-negative proportion and less than Unit adjustment direction Adjustment ratio of unit Save them separately, when the direction of influence is positive. Pick When the direction of influence is negative Pick If the direction of the impact is not yet determined, the brand's impact unit will not be included in the current assessment cycle adjustment.

[0039] For example, if a product line experiences a quality recall, the value adjustment rule table determines the base adjustment percentage based on the quality issue, the product line scope, and the extent of the recall. for The increase in quality complaints and negative evaluations comes from two different data sources, the item confirmation coefficient. Pick An increase in returns within a statistical period directly impacts business operations, and the business realization coefficient is calculated accordingly. Pick The three types of observations form a complete chain, affecting the chain closure coefficient. Pick Therefore, the unit adjustment ratio for The influence direction formed in step S102 is negative, and the element adjustment direction... Pick If returns increase for two consecutive statistical periods or if repeat purchases decrease simultaneously, the operating realization coefficient will be affected. Pick Unit adjustment ratio for .

[0040] The system generates a brand influence unit value adjustment result for each brand influence unit based on the influence unit identifier, saving the influence unit identifier, brand identifier, main object identifier, whether it participates in the current cycle's value adjustment, unit adjustment direction, unit adjustment ratio, value adjustment status, value adjustment basis, and rule version identifier. The value adjustment basis saves the brand event observations, main public response observations, main business result observations, event confirmation coefficients, business realization coefficients, influence chain closure coefficients, earliest observation time, last observation time, and the record merging basis formed in step S102. Finally, it outputs a set of brand influence unit value adjustment results for step S104 to perform a single settlement.

[0041] Step S104: Based on the value adjustment result set and brand value settlement registration record, update the currently settled brand value and output the brand value assessment result.

[0042] Specifically, the system takes the brand influence unit value adjustment result set output in step S103 as input, reads the influence unit identifier, brand identifier, main object identifier, whether it participates in the current period's value adjustment, unit adjustment direction, unit adjustment ratio, value adjustment status, value adjustment basis, and rule version identifier, and aggregates the corresponding results according to the brand identifier. Based on the brand identifier, the system reads the brand base value and the currently settled brand value from the brand value database for the current evaluation period. The brand base value is the fixed benchmark result at the beginning of the evaluation period, and the currently settled brand value is the latest valid result formed after the brand base value has been adjusted by the brand influence units that have completed settlement within this evaluation period; if settlement has not yet been performed in the current evaluation period, the currently settled brand value is equal to the brand base value.

[0043] The system establishes brand value settlement registration records in the brand value database, using the influence unit identifier and evaluation period as the sole search criteria. Step S103 identifies results that participate in the value adjustment for this period, whose rule version is consistent with the current evaluation period, and which have not yet formed a valid settlement registration; these results are added to the set of results awaiting settlement. Results with existing valid settlement registrations maintain their original settlement status. Results that do not participate in the value adjustment for this period retain their value adjustment status and the basis for value adjustment. The settlement registration record stores the influence unit identifier, brand identifier, main object identifier, unit adjustment direction, unit adjustment ratio, rule version identifier, settlement status, and settlement time.

[0044] The brand influence units in this pending settlement result set are applied to the currently settled brand value using a composite rate of change relationship. The composite rate of change represents the cumulative result when multiple independent proportional changes collectively affect the same current value, specifically: ; in, This represents the brand value obtained after completing the processing of the current set of results pending settlement; This indicates the currently settled brand value retrieved from the brand value database, which already includes the adjustment results generated by previously settled brand influence units. This indicates the number of brand influence units belonging to the current brand in the current set of results to be settled; Indicates the first The adjustment direction of each brand influence unit follows the meaning of step S103, taking a positive approach. negative take ; Indicates the first The unit adjustment ratio for each brand influence unit follows the non-negative ratio meaning of step S103 and is less than... . and Using the same unit of value, For directional indication, Since it is a dimensionless proportion, each adjustment factor is dimensionless and greater than... The calculated brand value maintains the same numerical type as the currently settled brand value; when there are no valid brand influence units belonging to the current brand in the current set of results to be settled, for The system remains It only saves records of unsettled status for the current period.

[0045] For example, the basic brand value of a certain brand is The product series quality recall has been settled, bringing the current settled brand value to [value missing]. Subsequently, a new product launch brand influence unit that has not yet been settled was acquired, with the unit adjustment direction being positive and the unit adjustment ratio being [missing information]. Therefore, the brand value after this settlement is Multiply ,get The affected units corresponding to the quality recall already have valid settlement registrations and will not be included in the current pending settlement result set. The adjustments already made will be retained in the currently settled brand value.

[0046] The system writes the settlement registration record, brand value adjustment details, and updated brand value for the current brand impact unit to be settled within the same database transaction. The database sets unique constraints on the impact unit identifier and evaluation period. When the same impact unit is submitted repeatedly due to concurrent processing, only the record successfully written for the first time is included in the actual adjustment. Each brand value adjustment detail saves the impact unit identifier, brand identifier, main object identifier, brand value before settlement, unit adjustment direction, unit adjustment ratio, brand value after settlement, value adjustment basis, rule version identifier, and settlement time. The brand impact units in the evaluation report are arranged according to the earliest observation time; this arrangement is used to show the sequence of brand events.

[0047] After settlement, the system generates the brand value assessment results for the current evaluation period, saving the brand logo, basic brand value, brand value before settlement, adjusted brand value, changes in brand value, settled brand influence units, brand influence units not yet adjusted and their value adjustment status, the main object identifiers corresponding to each brand influence unit, unit adjustment direction, unit adjustment ratio, value adjustment basis, rule version identifier, and evaluation period. The system writes the brand value assessment results into the brand value database, and the evaluation terminal generates a brand value assessment report; the next evaluation period reads the brand value finally confirmed in this period as the new basic brand value.

[0048] Figure 2This is a structural block diagram of a brand value intelligent evaluation system based on multi-source data fusion, according to an embodiment of this application. Figure 2 As shown, the system includes: The data access module 201 is used to acquire brand event data, public response summary data, and business results data.

[0049] The observation record standardization module 202 is used to convert the brand event data, the public response summary data, and the business results data into a standard brand observation record set.

[0050] The brand influence unit construction module 203 is used to calculate the degree of attribution support between public response observation and business result observation and candidate brand events, starting from brand event observation, and to determine the unique attribution relationship based on the degree of attribution support, thereby constructing a set of brand influence units.

[0051] The value adjustment result generation module 204 is used to determine the value adjustment status based on the observation composition of each brand influence unit in the brand influence unit set, calculate the unit adjustment ratio, and generate a value adjustment result set.

[0052] The settlement registration module 205 is used to update the currently settled brand value based on the value adjustment result set and the brand value settlement registration record.

[0053] The evaluation result output module 206 is used to output the brand value evaluation results.

[0054] It is worth noting that the specific workflow of the brand value intelligent evaluation system based on multi-source data fusion provided in this embodiment of the invention is the same as that of the brand value intelligent evaluation method based on multi-source data fusion described in the above embodiments, and will not be repeated here.

[0055] This invention also provides a brand value intelligent evaluation device based on multi-source data fusion, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps as described in the above embodiments of the brand value intelligent evaluation method based on multi-source data fusion; or, when the processor executes the computer program, it implements the functions of each module in the above system embodiments.

[0056] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the brand value intelligent evaluation device based on multi-source data fusion.

[0057] The brand value intelligent assessment device based on multi-source data fusion can be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. This device may include, but is not limited to, processors and memory. Those skilled in the art will understand that the device may also include input / output devices, network access devices, buses, etc.

[0058] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the brand value intelligent assessment device based on multi-source data fusion, connecting all parts of the device via various interfaces and lines.

[0059] The memory can be used to store the computer programs and / or modules. The processor, by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory, realizes various functions of the brand value intelligent evaluation device based on multi-source data fusion. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function, etc.; the data storage area may store data created according to the operation of the controller, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital card (SD card), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0060] The modules integrated into the brand value intelligent evaluation device based on multi-source data fusion, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0061] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0062] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A brand value intelligent evaluation method based on multi-source data fusion, characterized in that, include: Acquire brand event data, public response summary data, and business results data, and convert the brand event data, public response summary data, and business results data into a standard brand observation record set; Starting with brand event observation, calculate the degree of attribution support between public response observation and business results observation and candidate brand events, determine the unique attribution relationship based on the degree of attribution support, and construct a set of brand influence units; The value adjustment status is determined based on the observation composition of each brand influence unit in the brand influence unit set, the unit adjustment ratio is calculated, and a value adjustment result set is generated. Based on the aforementioned set of value adjustment results and brand value settlement registration records, the currently settled brand value is updated, and the brand value assessment result is output.

2. The brand value intelligent evaluation method based on multi-source data fusion according to claim 1, characterized in that, The process of converting the brand event data, the public response summary data, and the business results data into a standard brand observation record set includes: Based on the brand directory and object directory, brand identifiers and object identifiers are mapped to the brand event data, the public response summary data, and the business results data; Based on the data interface category and the recorded content, determine the observation type and observation category; the observation type includes brand event observation, public response observation, and business results observation; Observation results are formed based on the meaning of the data itself, and the observation time and direction of influence are uniformly determined; The standard brand observation record set is formed by writing the brand identifier, the object identifier, the observation type, the observation category, the observation content, the observation result, the observation time, the data source, the source record identifier, and the direction of influence.

3. The brand value intelligent evaluation method based on multi-source data fusion according to claim 1, characterized in that, The calculation of the degree of attribution support between public response observations and business results observations and candidate brand matters includes: Brand consistency is used as a gating condition. When the record matches the brand identity of the candidate item, the gating result is the first preset value; otherwise, it is the second preset value. Calculate the degree of correspondence between objects, the degree of correspondence between event chains, and the degree of correspondence between time. The degree of attribution support is obtained by combining the gating result, the degree of object correspondence, the degree of event chain correspondence, and the degree of time correspondence.

4. The brand value intelligent evaluation method based on multi-source data fusion according to claim 1, characterized in that, Determining the unique attribution relationship based on the degree of attribution support includes: Candidate items whose affiliation support level reaches a preset threshold will be included in the affiliation comparison. When a candidate item reaches the preset threshold, it is directly classified into that candidate item. When two candidate items reach the preset threshold, the candidate item with the highest degree of support for attribution is selected, and the unique attribution is determined when the difference between the highest and second-highest values ​​reaches the preset difference threshold. When the difference is less than the preset difference threshold, the attribution is determined in the following order: priority of specific object consistency, priority of direct event chain correspondence, and priority of observation time closer to the center position of the main related time period. If the above steps still fail to distinguish between the two brands, a candidate brand influence unit is created and marked as pending verification.

5. The brand value intelligent evaluation method based on multi-source data fusion according to claim 1, characterized in that, The set of brand influence units includes: Candidate items are merged based on source record identifiers, observation content, and time boundaries; Cross-source records that have the same brand identity, the same main target, the same brand event category, and correspond to the same original event will be merged into one brand influence unit. Generate a unique influence unit identifier for each brand influence unit; The influence direction of the brand influence unit is determined based on the brand event observation direction, public response observation direction, and business results observation direction.

6. The brand value intelligent evaluation method based on multi-source data fusion according to claim 1, characterized in that, The step of determining the value adjustment state based on the observation composition of each brand influence unit in the brand influence unit set includes: When brand event observation, public response observation, and business results observation all exist, and the records and merging basis indicate that the three types of observations have a clear attribution relationship, it is determined to be a complete realization status; When brand event observation and business result observation have a consistent specific object, a direct event chain correspondence, and an effective time correspondence, they are determined to be in the direct realization state. When brand event observation and public response observation exist, but business results observation has not yet been formed, it is determined to be a market confirmation state; When only brand events are observed, the event is identified as being in the formation stage. When only public response observations or operational results observations are available, the status is determined to be pending verification.

7. The brand value intelligent evaluation method based on multi-source data fusion according to claim 6, characterized in that, The adjustment ratio of the computing unit includes: Obtain the basic adjustment ratio from the value adjustment rule table; Confirmation coefficients for matters obtained based on public response observations or direct fulfillment of relationships; The operating realization coefficient is obtained based on the object relationship, continuous status, and number of unidirectional operating indicators observed in the operating results; The impact chain closure coefficient is obtained based on the complete or direct realization status. The unit adjustment ratio is obtained by multiplying the basic adjustment ratio, the item confirmation coefficient, the business realization coefficient, and the impact chain closure coefficient.

8. The brand value intelligent evaluation method based on multi-source data fusion according to claim 1, characterized in that, The step of updating the currently settled brand value based on the value adjustment result set and the brand value settlement registration record includes: The results that participated in the value adjustment of this period but have not yet formed a valid settlement registration in the value adjustment result set shall be regarded as the set of results to be settled; By using a composite rate of change relationship, the unit adjustment ratio and unit adjustment direction of each brand influence unit in the set of results to be settled are applied to the currently settled brand value to obtain the updated brand value; The settlement registration record, brand value adjustment details, and the updated brand value are written into the same database transaction; the database sets unique constraints on the influencing unit identifier and evaluation cycle.

9. The brand value intelligent evaluation method based on multi-source data fusion according to claim 1, characterized in that, The output brand value assessment results include: Save the brand logo, basic brand value, brand value before settlement, adjusted brand value, brand value change results, settled brand influence units, brand influence units not yet involved in the adjustment, and their corresponding value adjustment status; The brand value assessment results will be written into the brand value database; The evaluation terminal reads the brand value database and generates a brand value evaluation report; the brand influence units in the brand value evaluation report are arranged according to the earliest observation time.

10. A brand value intelligent evaluation system based on multi-source data fusion, characterized in that: include: The data access module is used to acquire brand event data, public response summary data, and business results data; The observation record standardization module is used to convert the brand event data, the public response summary data, and the business result data into a standard brand observation record set. The Brand Influence Unit Construction Module is used to calculate the degree of attribution support between public response observations and business results observations and candidate brand events, starting from brand event observations, and to determine a unique attribution relationship based on the degree of attribution support, thereby constructing a set of brand influence units. The value adjustment result generation module is used to determine the value adjustment status based on the observation composition of each brand influence unit in the brand influence unit set, calculate the unit adjustment ratio, and generate a value adjustment result set. The settlement registration module is used to update the currently settled brand value based on the value adjustment result set and the brand value settlement registration record; The evaluation results output module is used to output the brand value evaluation results.

Images