WO2017016403A1 - 确定业务对象品牌指数信息的方法及装置 - Google Patents

确定业务对象品牌指数信息的方法及装置 Download PDF

Info

Publication number
WO2017016403A1
WO2017016403A1 PCT/CN2016/090285 CN2016090285W WO2017016403A1 WO 2017016403 A1 WO2017016403 A1 WO 2017016403A1 CN 2016090285 W CN2016090285 W CN 2016090285W WO 2017016403 A1 WO2017016403 A1 WO 2017016403A1
Authority
WO
WIPO (PCT)
Prior art keywords
brand
indicator
level
determining
index
Prior art date
Application number
PCT/CN2016/090285
Other languages
English (en)
French (fr)
Inventor
朱璐璐
李传福
Original Assignee
阿里巴巴集团控股有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 阿里巴巴集团控股有限公司 filed Critical 阿里巴巴集团控股有限公司
Publication of WO2017016403A1 publication Critical patent/WO2017016403A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce

Definitions

  • the present application relates to the field of business object information processing technology, and in particular, to a method and apparatus for determining business object brand index information.
  • the core of network brand refinement is to portray the concept and personality of the brand, and to study the relationship between the brand personality and image of the shopping website and the attitude of the consumer brand from the perspective of consumers, which is conducive to stimulating the purchase demand of consumers through brand recommendation.
  • the index system of the brand index in the prior art includes brand identification, information construction, channel construction, customer development, media performance, search power, market activities, and word of mouth.
  • the content of each item basically comes from the information that the brand officially announces and is known to the public.
  • Sources of information include corporate website information, corporate brochures and product brochures, advertisements placed by companies in magazines and websites, exhibiting materials and exhibits, events attended by companies or self-sponsored, descriptions of brands by corporate agents, and user pairs. Brand descriptions, industry magazines and industry websites, personal websites and forums that involve brand information, blogs, and more.
  • the existing brand index mainly measures the situation of the offline brand, and does not consider the different performance characteristics of the brand in the e-commerce sales platform, so it is impossible to deeply describe the e-commerce sales.
  • the personality image of the brand in the platform, and thus the existing brand index can not be used to meet the consumer demand for the brand in the e-commerce sales platform.
  • the embodiment of the present application provides a method and device for determining business object brand index information, and can more accurately describe a brand index of each brand from the perspective of an e-commerce sales platform.
  • a method for determining business object brand index information including:
  • the quantized value, the weight, and the N-1th index of each Nth level indicator are used to determine the quantized value of each index of each N-1 level, and the higher level indicator is sequentially followed according to the method of the step.
  • the layer-by-layer calculation is performed until the quantized value of each index of the first level is calculated, and the brand index of the brand is determined by using the quantized value and the weight on each level 1 index.
  • a device for determining brand index information of a business object comprising:
  • An indicator weight determining unit is configured to determine a weight of each indicator, where the weight is used to indicate a relative importance level of the indicator when characterizing a higher-level indicator to which the indicator belongs;
  • An indicator quantization value determining unit is configured to determine, according to a historical operation record of the user, a statistical value of each of the Nth level indicators of the same brand, and determine a quantized value of each Nth level indicator according to the statistical value;
  • the brand index determining unit is configured to determine the quantized value of each index of each N-1 level by using the quantized value, the weight, and the N-1 level indicator to which each Nth level indicator belongs, and according to the The method of the step is to perform layer-by-layer calculation in the upper level index until the quantized value of each index of the first level is calculated, and each level 1 is utilized.
  • the quantified value and weight of the indicator determine the brand index of the brand.
  • the present application discloses the following technical effects:
  • the relationship between the brand personality of the shopping website and the consumer brand attitude can be studied from the perspective of the e-commerce consumer, and the brand can be described.
  • it can be realized by establishing a hierarchical structural model, in which a plurality of hierarchical indicators are included, and the last-level indicators are statistic or computable values, and each Indicators can each have their own weights to indicate the relative importance of their role in characterizing the higher-level indicators to which they belong. In this way, it is possible to start from the last level of indicators and adopt a step-by-step upward calculation method to finally obtain the index quantitative value of the brand.
  • the historical operation information of the network consumer users in the sales platform is utilized for data mining, which reflects the characteristics of the e-commerce transaction platform.
  • the hierarchical structure due to the hierarchical structure, for each indicator, it is only necessary to determine the importance level of the previous level of indicators to which it belongs, and perform level-by-level calculations without the need for cross-level weighting. Determined, more efficient, and more accurate calculation results.
  • FIG. 2 is a schematic diagram of an index hierarchical structure model in the embodiment of the present application.
  • FIG. 3 is a schematic diagram of a brand index scatter in the embodiment of the present application.
  • FIG. 4 is a schematic diagram of an apparatus provided by an embodiment of the present application.
  • the brand of the shopping website can be studied from the perspective of the cognition of the e-commerce consumer.
  • it can be achieved by establishing a hierarchical structure model.
  • N 2
  • the brand index can be characterized by two levels of indicators.
  • the first-level indicators can be identified as brand recognition, function positioning, quality identification, brand purchase and brand loyalty.
  • these five aspects have a strong correspondence with the perception, emotion, and intention of brand attitude, and can deeply portray the brand image of consumers.
  • each level indicator can be characterized by one or more secondary indicators, and the second level indicator is generally a specific indicator value that can be obtained, such as the transaction amount.
  • Brand awareness is characterized by some or all of the following secondary indicators: number of searches, number of favorites, and number of views;
  • Functional positioning is characterized by some or all of the following secondary indicators: number of business objects, presence or absence of designated types of stores, number of stores;
  • Quality certification is characterized by some or all of the following secondary indicators: brand price, high score description rate and refund complaint rate;
  • Brand purchase is characterized by some or all of the following secondary indicators: number of transactions, turnover, conversion rate, sales, and unit price;
  • Brand loyalty is characterized by the following secondary indicators: repeat purchase rate.
  • the establishment of the above-mentioned second-level hierarchical model is based on the shopping characteristics and individual needs of online consumers, starting from the process of consumers' attitude towards the brand, along the consumer from the contact with the cognitive brand to the purchase to the second purchase.
  • the route comprehensively measures the consumer shopping choice experience, selects and builds a brand indicator system that can describe the personality needs of online consumers. This way, you can unearth the brands that you like and care about in the inner world of online consumers.
  • brand information to provide users with various types of business object information (for example, search results, results by category browsing, recommendation information, etc.), such brands that are generally liked or concerned can be preferentially provided.
  • the user's purchase demand can be stimulated, the time for the user to select the product can be saved, and the concept of “user-centered” can be more satisfied, and the user's stay time on the website can be increased, the consumption can be stimulated, and the user's stickiness to the website can be improved.
  • the quantized value of the last-level indicator (level N) can be directly calculated according to the statistical data, and the index of the N-th level belongs to the N-1 level. Therefore, the quantized value of the index of the N-th level can be specifically
  • the linear addition method obtains the quantized value of each index of the N-1th level. In this way, the higher-level index is calculated layer by layer, and the quantized value of each index of the first level can be calculated, and then each level 1 can be utilized.
  • the quantified value on the indicator is determined by linear summation to determine the brand's brand index quantified value.
  • each nth level indicator can be characterized by multiple n+1th level indicators
  • each n+1th level usually has different importance when characterizing the nth level indicator to which it belongs. Therefore, in the linear summation of the quantitative values of the indicators, it is also possible to consider the difference in importance, so as to more accurately depict the brand index. Specifically, in order to reflect this difference in importance, the weight of each indicator can also be determined. This weight identifies the relative importance of the indicator when characterizing the higher-level indicator to which it belongs.
  • the weight of each indicator can also be determined.
  • the linear weighting calculation can be performed based on the weights, so that the obtained results are more accurate.
  • an embodiment of the present application provides a method for determining business object brand index information, and the method may include the following steps:
  • S102 Determine a weight of each indicator, where the weight is used to indicate a relative importance level of the indicator when characterizing a higher-level indicator to which the indicator belongs;
  • the weights of the respective indicators may also be determined by: firstly, generating m*m judgment matrices based on m n+1th level indicators belonging to the same nth level indicator, and then generating a m ⁇ m judgment matrix, and then The maximum eigenvalue of the judgment matrix and the eigenvector corresponding to the maximum eigenvalue can be calculated, and finally the weights of the m n+1th metrics can be determined according to the eigenvector corresponding to the maximum eigenvalue.
  • Other indicators can also be weighted in this way.
  • the judgment matrix represents the relative weight of each element in each level relative to its upper element in the form of a matrix To the extent.
  • the value of each element in the judgment matrix is determined by: comparing the importance degree of the m n+1th level indicators when characterizing the nth level indicator, according to the ratio
  • the difference degree quantified value is determined for the degree of difference of the results, and the difference degree quantized value obtained by the pairwise alignment is determined as each element in the judgment matrix.
  • a quantized value of the degree of difference of 1 to 9 may be introduced, as shown in Table 1:
  • a ij definition 1 The i factor is as important as the j factor 3 The i factor is slightly more important than the j factor 5 The i factor is more important than the j factor 7 The i factor is very important to the j factor 9 The i factor is absolutely important than the j factor 2,4,6,8
  • the above-mentioned difference degree quantization value can also be defined by other means. After the above-mentioned difference degree quantization value is defined, each index belonging to the same superior indicator can be compared in pairs, and a judgment matrix is established according to the comparison result.
  • Z is a quantitative value of the brand index that needs to be calculated finally, wherein the first-level indicators are A1, A2, A3, A4, and A5, and the second-level indicators are B1, B2, ..., B15, respectively.
  • B1, B2, and B3 belong to A1, B4, B5, and B6 belong to A2, and so on. That is, A1 to A5 are five factors that affect the quantized value of Z, and B1, B2, and B3 are three factors that affect the quantized value of A1, and so on.
  • the quantized value of the difference degree of A1 with respect to A2 can be expressed as 3
  • the quantized value of the difference degree of A2 with respect to A1 can be expressed as 1/3, and so on. It is possible to establish the first judgment matrix:
  • each element in the matrix represents the difference between the five factors A1 to A5, and the quantized value of the degree of difference. It can be seen that the elements on the diagonal are all 1, because it is the same The result of the indicator when compared to itself.
  • the elements in the symmetrical position are reciprocal to each other, for example, the element value of the first row and the second column is 1/2, and the element value of the first column of the second row is 2, which is Representative:
  • the quantized value of the difference degree of A1 with respect to A2 is 1/2.
  • the quantified value of the difference degree of A2 with respect to A1 is its reciprocal, that is, 2, and so on. .
  • each of the first-level indicators is characterized by a plurality of secondary indicators, respectively.
  • a judgment matrix can also be obtained. For example, for the hierarchical structure shown in FIG. 2, based on the five first-level indicators A1 to A5, the following five judgment matrices can be separately established:
  • the judgment matrix of M2 represents the comparison of the two levels of the second-level indicators B1, B2, and B3 belonging to the first-level index A1, and the difference degree quantitative values obtained respectively.
  • the judgment matrix of M3 represents the comparison of the second-level indicators B4, B5, and B6 belonging to the first-level indicator A2, and the quantized value of the degree of difference obtained, and so on. If there are other levels of indicators, you can also establish other judgment matrices as described above.
  • each judgment matrix may be determined according to experience or the like.
  • the weight of each indicator can be determined based on the judgment matrix. Specifically, for a judgment matrix, a maximum eigenvalue of the judgment matrix and an eigenvector corresponding to the maximum eigenvalue may be calculated, and the eigenvector includes a quantity of component values that are the same as the order of the matrix, that is, The number of indicators of the same superior indicator is the same. Therefore, each component value of the feature vector can be used as the weight of each indicator.
  • the maximum eigenvalue is calculated, and the eigenvector corresponding to the maximum eigenvalue is, the weight of the index A1 is determined to be 0.263, the weight of A2 is 0.475, and the weight of A3 is 0.055, A4 The weight is 0.099, and the weight of A5 is 0.110.
  • the maximum eigenvalue is calculated, and if the eigenvector corresponding to the maximum eigenvalue is, the weight of B1 may be determined to be 0.595.
  • the weight of B2 is 0.277, and the weight of A3 is 0.129.
  • its respective weights can be determined in the above manner.
  • the consistency check may be performed on the judgment matrix, if If the verification passes, the component values included in the feature vector corresponding to the maximum eigenvalue are used as the weights of the corresponding indicators. Otherwise, if the verification fails, it may prove that the element values in the judgment matrix may have some irrationality. Therefore, the value of the element in the judgment matrix can also be adjusted. After the adjustment, the maximum eigenvalue is recalculated, and the consistency check is performed again. After the verification is passed, the eigenvector is used to determine the weight of each indicator.
  • the so-called consistency check the purpose is to test the coordination between the importance of each element in the judgment matrix, to avoid the contradiction between "A1 is more important than A2, A2 is more important than A3, and A3 is more important than A1".
  • S103 Determine, according to a historical operation record of the user, statistical values of the same brand on each level N indicator, and determine a quantized value of each level N indicator according to the statistical value;
  • the model After determining the hierarchical structure of each indicator and its respective weights, the model can be used to evaluate the index quantified values of each brand. Specifically, since the last level, that is, the level N indicator, can usually be directly obtained through statistics, etc., it is possible to first determine the statistical value of the same brand on each level N indicator, and determine each according to the statistical value. Quantitative value of the Nth level indicator.
  • each indicator and the weight information can be expressed by the following Table 2.
  • Table 2 above gives the calculation method of each secondary indicator, which can be calculated according to the historical operation records of a large number of users in the sales platform, including historical purchase records, collection records, browsing records, etc. of a given brand.
  • the statistical value of the level indicator The specific calculation method for each secondary indicator will not be described in detail here.
  • each indicator on the Nth level can also calculate the statistical value by the above method, and then determine the quantized value of the Nth level indicator according to the method.
  • the foregoing indicator may also be calculated by using the other statistical values.
  • the corresponding calculation method may also be adjusted.
  • Table 2 is only an example, and should not be regarded as Limitations on the scope of protection of this application.
  • the statistical values of the above-mentioned Nth-level indicators are directly compared, there may be cases where they are not of the same order of magnitude and therefore have no direct comparability with each other.
  • a secondary indicator is the number of times a brand is viewed, and another secondary indicator is the number of times the brand is viewed.
  • the number of views is much larger than the number of collections, and the two are not on the same order of magnitude. Therefore, in order to make the quantized values of the respective N-th level indicators comparable, and finally make the quantized values of the respective brand indices comparable, and also to make the statistical results smoother, the statistical values of the respective N-th level indicators can be performed.
  • the smoothing and normalization processing are performed, and then the obtained values are determined as the quantized values of the respective Nth-level indicators.
  • the specific processing manner may be various.
  • the ln function and the inverse tangent function may be used to process the statistical values of the respective Nth level indicators, so that the quantization of each Nth level indicator is performed. Values fall within the range [0,1].
  • S104 using the quantized value, the weight, and the N-1th index to which each Nth level indicator belongs, determining the quantized value of each index of each N-1 level by linear weighted summation.
  • the upper-level index is calculated layer by layer until the quantized value of each index of the first level is calculated, and the quantized value and the weight on each level 1 index are used to determine by linear weighted summation.
  • the brand's brand index is used to determine by linear weighted summation.
  • the quantized values, the weights, and the N-1th level indicators to which the respective Nth level indicators are attached can be utilized by linear weighted summation.
  • the quantized values of each index of each N-1th level are determined.
  • N 2
  • the secondary indicators B1, B2, and B3 belong to the primary indicator A1. Therefore, the quantized value of A1 can be calculated as:
  • VA1 VB1 ⁇ w1+VB2 ⁇ w2+VB3 ⁇ w3
  • the secondary indicators B4, B5, and B6 belong to the first-level indicator A2. Therefore, the quantized value of A2 can be calculated as:
  • VA2 VB4 ⁇ w4+VB5 ⁇ w5+VB6 ⁇ w6
  • the other individual and the quantized values of the indicators can also be linearly weighted and summed in the above manner.
  • the level indicator can be calculated layer by layer according to the step, until the level 1 is calculated.
  • the brand index of the brand is determined by linear weighted summation using the quantized values and weights on the respective first level indicators.
  • VZ VA1 ⁇ W1+VA2 ⁇ W2+VA3 ⁇ W3+VA4 ⁇ W4+VA5 ⁇ W5
  • brand indices of other brands can also be calculated by the above steps.
  • the correlation between the brand personality of the shopping website and the consumer brand attitude can be studied from the perspective of the e-commerce consumer, and the brand can be described.
  • it can be realized by establishing a hierarchical structural model, in which a plurality of hierarchical indicators are included, and the last-level indicators are statistic or computable values, and each Indicators can each have their own weights to indicate the relative importance of their role in characterizing the higher-level indicators to which they belong. In this way, it is possible to start from the last level of indicators and adopt a step-by-step upward calculation method to finally obtain the index quantitative value of the brand.
  • the historical operation information of the network consumer users in the sales platform is utilized for data mining, which reflects the characteristics of the e-commerce transaction platform.
  • the hierarchical structure due to the hierarchical structure, for each indicator, it is only necessary to determine the importance level of the previous level of indicators to which it belongs, and perform level-by-level calculations without the need for cross-level weighting. Determined, more efficient, and more accurate calculation results.
  • the brand index pair can be utilized.
  • the business object list is processed. For example, business objects under brands with higher brand indices can be given priority, and so on.
  • a scatter plot of the brand index can be made according to the order of the brand index from small to large, and then the corresponding point can be found according to the point at which the slope of the trend line in the scatter plot changes.
  • the brand index which uses these brand indices as a threshold, achieves stratification of the brand.
  • this level information can be applied to a specific application scenario. For example, in mention When the business object information list is provided, an operation option for selecting according to the brand level can be provided, so that the consumer can select a brand suitable for the target level according to his actual needs, and then the server can select the selected target according to the operation selection. Level, and provide business object information of the brand within the target level, thus saving the time for the user to select the product.
  • the embodiment of the present application further provides a device for determining the business object brand index information.
  • the device may specifically include:
  • the index weight determining unit 402 is configured to determine a weight of each indicator, where the weight is used to indicate a relative importance level of the indicator when characterizing the upper level indicator to which the indicator belongs;
  • the index quantization value determining unit 403 is configured to determine, according to the historical operation record of the user, statistical values of the same brand on each of the Nth level indicators, and determine a quantized value of each Nth level indicator according to the statistical value;
  • the brand index determining unit 404 is configured to determine the quantized value of each index of each N-1 level by using the quantized value, the weight, and the N-1th index to which each Nth level indicator belongs.
  • the method of this step is to perform layer-by-layer calculation in the upper level index until the quantized value of each index of the first level is calculated, and the brand index of the brand is determined by using the quantized value and weight on each level 1 index.
  • the indicator weight determining unit 402 may include:
  • the judgment matrix generation subunit is configured to generate a m ⁇ m judgment matrix based on the m n+1th level indicators belonging to the same nth level indicator, and the values of the respective elements in the judgment matrix are determined by:
  • a calculation subunit configured to calculate a maximum eigenvalue of the judgment matrix and a feature vector corresponding to the maximum eigenvalue
  • a weight determining subunit configured to determine, according to the feature vector corresponding to the maximum feature value, a weight of each of the m n+1th level indicators.
  • the weight determining subunit includes:
  • a consistency check subunit configured to perform a consistency check on the judgment matrix
  • Determining a subunit configured to determine, as a check pass, each component value in the feature vector corresponding to the maximum feature value as a weight of each of the m n+1th level indicators.
  • the matrix adjustment unit is configured to adjust the judgment matrix if the consistency check fails.
  • the quantization value determining unit 403 is specifically configured to:
  • the statistical value is smoothed and normalized to obtain a quantized value of the Nth level indicator.
  • the brand cognition is characterized by some or all of the following secondary indicators: number of searches, number of favorites, and number of views;
  • the function positioning is characterized by some or all of the following secondary indicators: the number of business objects, the presence or absence of a specified type of store, and the number of stores;
  • the quality determination is characterized by some or all of the following secondary indicators: brand price, high-scoring description matching rate, and refund complaint rate;
  • the brand purchase is characterized by some or all of the following secondary indicators: the number of transactions, the turnover, the conversion rate, the sales degree, and the customer unit price;
  • the brand loyalty is characterized by the following secondary indicators: repeat purchase rate.
  • the device may further include:
  • the business object list processing unit is configured to process the provided business object list according to the brand index when the business object list information is provided to the client.
  • the device may further comprise:
  • Brand stratification unit for stratifying brands according to brand indices of various brands
  • the information providing unit is configured to provide the business object information to the client by using the hierarchical information.
  • the brand stratification unit comprises:
  • a target point determining sub-unit for determining a target point at which a slope of a trend line in the scattergram changes
  • a threshold value determining subunit is configured to determine a brand index value at the target point as a threshold value of the layer to layer the brand according to the threshold value.
  • the information providing unit includes:
  • An operation option provides a sub-unit for providing operational options for selecting a brand hierarchy
  • the information providing subunit is configured to provide the business object information of the brand within the target level after determining the selected target level according to the operation selection.
  • the relationship between the brand personality of the shopping website and the consumer brand attitude can be studied from the perspective of the e-commerce consumer, and the brand can be described.
  • it can be realized by establishing a hierarchical structural model, in which a plurality of hierarchical indicators are included, and the last-level indicators are statistic or computable values, and each Indicators can each have their own weights to indicate the relative importance of their role in characterizing the higher-level indicators to which they belong. In this way, it is possible to start from the last level of indicators and adopt a step-by-step upward calculation method to finally obtain the index quantitative value of the brand.
  • the historical operation information of the network consumer users in the sales platform is utilized for data mining, which reflects the characteristics of the e-commerce transaction platform.
  • the hierarchical structure due to the hierarchical structure, for each indicator, it is only necessary to determine the importance level of the previous level of indicators to which it belongs, and perform level-by-level calculations without the need for cross-level weighting. Determined, more efficient, and more accurate calculation results.
  • the present application can be implemented by means of software plus a necessary general hardware platform. Based on such understanding, the technical solution of the present application may be embodied in the form of a software product in essence or in the form of a software product, which may be stored in a storage medium such as a ROM/RAM or a disk. , an optical disk, etc., includes instructions for causing a computer device (which may be a personal computer, server, or network device, etc.) to perform the methods described in various embodiments of the present application or portions of the embodiments.
  • a computer device which may be a personal computer, server, or network device, etc.

Landscapes

  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Finance (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本申请实施例公开了确定业务对象品牌指数信息的方法及装置,所述方法包括:预先确定用于刻画品牌指数的指标,所述指标包括N个层级,每个层级上包括至少一个指标,第n级指标通过至少一个第n+1级指标进行刻画;确定各个指标的权重;根据用户的历史操作记录,确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值;利用各个第N级指标上的量化值、权重依次向上级指标进行逐层计算,直到计算出第1级各个指标的量化值时,利用各个第1级指标上的量化值、权重,确定该品牌的品牌指数。通过本申请实施例,能够从电商销售平台的角度更准确地刻画各个品牌的品牌指数。

Description

确定业务对象品牌指数信息的方法及装置
本申请要求2015年07月27日递交的申请号为201510447140.8、发明名称为“确定业务对象品牌指数信息的方法及装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及业务对象信息处理技术领域,特别是涉及确定业务对象品牌指数信息的方法及装置。
背景技术
电子商务的重要组成部分——购物网站的蓬勃发展,带动了网络营销的迅速发展,开始进入了网络品牌细化的时代,消费者的购物行为中开始出现自主性和个性化的需求。消费者对单向的、被动式的“填鸭式”信息获取方式感到厌倦和不信任,希望在购买产品时通过各种途径能够主动获取相关信息。网络品牌细化适应了消费者的这一特殊需求。网络为消费者个性化的选择提供了便利,个性化消费成为消费的主流,而这种个性化恰好是品牌战略的重点。网络品牌细化的核心即为刻画品牌的理念和个性,从消费者的角度研究购物网站的品牌个性和形象以及消费者品牌态度之间的关系,有利于通过品牌推荐激发消费者的购买需求,精准定位品牌活动营销,提高购物网站的品牌影响力和品质感。
现有技术中的品牌指数的指标体系包括品牌识别、信息化建设、渠道建设、客户拓展、媒体表现、搜索力、市场活动、口碑八项。每项的内容基本上来自品牌正式对外公布、且被公众所了解的信息。信息来源包含企业网站信息、企业宣传画册和产品说明书、企业在杂志和网站上投放的广告、企业参展资料和参展情况、企业参加或自行主办的活动情况、企业代理商对品牌的描述、用户对品牌的描述、行业杂志和行业网站、涉及品牌信息的个人网站和论坛、博客,等等。
上述现有技术虽然能够从一定程度上刻画品牌指数,但是这种现有的品牌指数主要衡量线下品牌的情况,没有考虑电商销售平台中品牌的不同表现特点,因此无法深刻刻画电商销售平台中品牌的个性形象,进而也就无法用该现有的品牌指数来满足消费者对电商销售平台中品牌的需求。
因此,如何从电商销售平台的角度更准确地刻画各个品牌的品牌指数,成为需要本领域技术人员解决的技术问题。
发明内容
本申请实施例提供了确定业务对象品牌指数信息的方法及装置,能够从电商销售平台的角度更准确地刻画各个品牌的品牌指数。
本申请实施例提供了如下方案:
一种确定业务对象品牌指数信息的方法,包括:
预先确定用于刻画品牌指数的指标,所述指标包括N个层级,每个层级上包括至少一个指标,第n级指标通过至少一个第n+1级指标进行刻画;其中,n=1,2……N,N为大于1的正整数;
确定各个指标的权重,所述权重用于表示所述指标在刻画其所隶属的上一级指标时体现的相对重要程度;
根据用户的历史操作记录,确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值;
利用各个第N级指标上的量化值、权重以及各个第N级指标所隶属的第N-1级指标,确定各个第N-1级各个指标量化值,并按照该步骤的方式依次向上级指标进行逐层计算,直到计算出第1级各个指标的量化值时,利用各个第1级指标上的量化值、权重,确定该品牌的品牌指数。
一种确定业务对象品牌指数信息的装置,包括:
指标确定单元,用于预先确定用于刻画品牌指数的指标,所述指标包括N个层级,每个层级上包括至少一个指标,第n级指标通过至少一个第n+1级指标进行刻画;其中,n=1,2……N,N为大于1的正整数;
指标权重确定单元,用于确定各个指标的权重,所述权重用于表示所述指标在刻画其所隶属的上一级指标时体现的相对重要程度;
指标量化值确定单元,用于根据用户的历史操作记录,确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值;
品牌指数确定单元,用于利用各个第N级指标上的量化值、权重以及各个第N级指标所隶属的第N-1级指标,确定各个第N-1级各个指标量化值,并按照该步骤的方式依次向上级指标进行逐层计算,直到计算出第1级各个指标的量化值时,利用各个第1级 指标上的量化值、权重,确定该品牌的品牌指数。
根据本申请提供的具体实施例,本申请公开了以下技术效果:
通过本申请实施例,可以从电子商务消费者的认知角度出发,研究购物网站的品牌个性与消费者品牌态度之间的相关关系,并以此来刻画品牌。并且,为了体现这种相关关系,可以通过建立层次化结构模型的方式来实现,在该模型中,包括多个层级的指标,最后一级指标为可统计或者可计算的值,并且,每个指标都可以具有各自的权重,用于表明其在刻画其所隶属的上一级指标时所起作用的相对重要程度。这样,就可以从最后一级指标开始,采用逐级向上计算的方式,最终得到品牌的指数量化值。在整个计算过程中,利用了网络消费者用户在销售平台中的历史操作信息进行数据挖掘,体现了电子商务交易平台的特性。并且,由于采用了层级化的结构,因此,对于各个指标而言,只需要确定出对于其所隶属的上一级别指标的重要程度,并进行逐级的计算,而不需要进行跨级别的权重确定,效率更高,计算结果也更准确。
当然,实施本申请的任一产品并不一定需要同时达到以上所述的所有优点。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案,下面将对实施例中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本申请实施例提供的方法的流程图;
图2是本申请实施例中指标层次化结构模型示意图;
图3是本申请实施例中品牌指数散点示意图;
图4是本申请实施例提供的装置的示意图。
具体实施方式
下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员所获得的所有其他实施例,都属于本申请保护的范围。
在本申请实施例中,可以从电子商务消费者的认知角度出发,研究购物网站的品牌 个性与消费者品牌态度之间的相关关系,并以此来刻画品牌。为了体现这种相关关系,可以通过建立层次化结构模型的方式来实现。具体的,可以预先确定用于刻画品牌指数的指标,这种指标可以包括N个层级,每个层级上包括至少一个指标,其中,第n级指标通过至少一个第n+1级指标进行刻画;其中,n=1,2……N,N为大于1的正整数。
例如,在一种具体的实现方式下,N=2,也即,可以通过两个层级的指标,对品牌指数进行刻画。其中,考虑到消费者对品牌的态度形成包括“认知——情感——意动”的过程,因此,可以将一级指标确定为品牌认知、功能定位、品质认定、品牌购买和品牌忠诚度五个方面中的部分或全部,这五个方面与品牌态度的认知、情感、意动有很强的对应关系,能够很深刻的刻画消费者心中的品牌形象。其中,每个一级指标又可以通过一个或者多个二级指标来进行刻画,二级指标一般是具体可以获取的指标值,例如成交额等。具体的,对应上述各个一级指标,二级指标一共可以有15个,并且,一级指标与二级指标的关系可以如下:
品牌认知通过以下部分或全部二级指标进行刻画:搜索次数、收藏次数以及浏览次数;
功能定位通过以下部分或全部二级指标进行刻画:业务对象数量、有无指定类型店铺、店铺数量;
品质认定通过以下部分或全部二级指标进行刻画:品牌价位、高评分的描述相符率以及退款投诉率;
品牌购买通过以下部分或全部二级指标进行刻画:成交笔数、成交额、转化率、热销度、客单价;
品牌忠诚度通过以下二级指标进行刻画:重复购买率。
上述二级层级模型的建立,是根据网络消费者的购物特点和个性需求,从消费者对品牌的态度形成的过程出发,沿着消费者从接触认知品牌到购买再到二次购买的购物路线,全面衡量消费者购物选择体验,选取并构建能够刻画网络消费者个性需求的品牌指标体系。这种,可以挖掘出网络消费者内心世界里的喜欢和关注的品牌。利用这种品牌信息在向用户提供各种类型的业务对象信息(例如,搜索结果,按类目浏览的结果,推荐信息等等)时,可以将这种普遍被喜欢或者关注的品牌优先提供,这样,可以激发用户的购买需求,节省用户选商品的时间,更加符合“以用户为中心”的理念,并且能增加用户在网站上的停留时间,拉动消费,提升用户对网站的粘性。
在建立了上述递阶层级结构之后,上下层之间的元素隶属关系就确定了,接下来就 可以通过对各级指标进行量化,最终得到品牌的量化表示结果,以便在不同的品牌之间进行比较。
其中,最后一级指标(第N级)的量化值可以直接根据统计数据中计算获得,而第N级指标隶属于第N-1级,因此,具体可以通过将第N级指标的量化值进行线性相加的方式,获得第N-1级各指标的量化值,以此方式向上级指标逐层计算,就可以计算出第1级各个指标的量化值,进而,就可以利用各个第1级指标上的量化值,通过线性求和的方式,确定该品牌的品牌指数量化值。
具体实现时,虽然每个第n级指标可以通过多个第n+1级指标来刻画,但是,每个第n+1级在刻画其隶属的第n级指标时,通常具有不同的重要性,因此,在进行指标量化值的线性求和时,还可以考虑体现出这种重要性方面的差别,从而更准确的刻画出品牌指数。具体的,为了体现这种重要性方面的差别,还可以确定出各个指标的权重,这种权重就标识该指标在刻画其所隶属的上一级指标时体现的相对重要程度。
也就是说,在本申请实施例中,在建立了指标之间的递阶层级结构之后,还可以确定出每个指标的权重。在依次向上级指标进行逐层计算时,就可以基于这种权重,进行线性加权计算,从而使得得出的结果更准确。
下面对具体的实现方式进行详细介绍。
参见图1,本申请实施例提供了一种确定业务对象品牌指数信息的方法,该方法可以包括以下步骤:
S101:预先确定用于刻画品牌指数的指标,所述指标包括N个层级,每个层级上包括至少一个指标,第n级指标通过至少一个第n+1级指标进行刻画;其中,n=1,2……N,N为大于1的正整数;
S102:确定各个指标的权重,所述权重用于表示所述指标在刻画其所隶属的上一级指标时体现的相对重要程度;
具体确定各个指标的权重的方式可以有多种,例如,可以直接根据经验等方式,判断各个指标的重要程度,并直接赋予权重值。或者,在本申请实施例中,还可以通过以下方式来确定各个指标的权重:首先,基于隶属于同一第n级指标的m个第n+1级指标,生成m×m的判断矩阵,然后,可以计算判断矩阵的最大特征值以及该最大特征值对应的特征向量,最终就可以根据该最大特征值对应的特征向量,来确定该m个第n+1级指标各自的权重。其他各指标也均可以按照此方式确定权重。
其中,判断矩阵是以矩阵的形式来表述每一层次中各要素相对其上层要素的相对重 要程度。具体实现时,该判断矩阵中各个元素的取值通过以下方式确定:将所述m个第n+1级指标的在刻画该第n级指标时体现的重要程度进行两两比对,根据比对结果的差异程度确定差异程度量化值,将所述两两比对得到的所述差异程度量化值,确定为所述判断矩阵中的各个元素。
在一种具体实现方式下,为了使各因素之间进行两两比较得到量化的判断矩阵,可以引入1~9的差异程度量化值,如表1所示:
表1
差异程度量化值aij 定义
1 i因素与j因素同等重要
3 i因素比j因素略重要
5 i因素比j因素较重要
7 i因素比j因素非常重要
9 i因素比j因素绝对重要
2,4,6,8 为以上判断之间的中间状态对应的标度值
倒数 若i因素与j因素比较,得到判断值为,aji=1/aij,aii=1
当然,在实际应用中,还可以通过其他方式定义上述差异程度量化值。在定义了上述差异程度量化值之后,就可以将隶属于同一上级指标的各个指标进行两两比较,并根据比较的结果建立起判断矩阵。
例如,如图2所示,假设Z为最终需要计算的品牌指数的量化值,其中,一级指标为A1、A2、A3、A4、A5,二级指标分别为B1、B2……B15,其中,B1、B2、B3隶属于A1,B4、B5、B6隶属于A2,以此类推。也就是说,A1至A5是影响Z的量化值的五个因素,B1、B2、B3是影响A1量化值的三个因素,等等。接下来,就可以在隶属于同一上级指标的各个指标之间,进行重要程度的两两比对。例如,对于A1至A5这五个因素,分别在A1与A2、A1与A3、A1与A4、A1与A5、A2与A3……等之间进行两两比对。其中,假设,A1比A2略重要,则A1相对于A2的差异程度量化值就可以表示为3,相应的,A2相对于A1的差异程度量化值就可以表示为1/3,以此类推,就可以建立起第一个判断矩阵:
Figure PCTCN2016090285-appb-000001
其中,矩阵中的各个元素就代表A1至A5这五个因素两两对比时,得到的差异程度量化值,可以看出,对角线上的元素均为1,这是因为,其为同一个指标与自身比较时的结果。另外,以该对角线为轴,对称位置上的元素之间互为倒数,例如,第一行第二列的元素值为1/2,第二行第一列的元素值为2,这代表:元素A1与A2进行两两比较时,A1相对于A2的差异程度量化值为1/2,相应的,A2相对于A1的差异程度量化值就为其倒数,也就是2,以此类推。
另外,由于各个一级指标分别由多个二级指标来刻画,因此,也分别可以得到一个判断矩阵。例如对于图2中所示的层级结构,基于A1至A5这五个一级指标,还可以分别建立以下五个判断矩阵:
Figure PCTCN2016090285-appb-000002
Figure PCTCN2016090285-appb-000003
其中,M2这一判断矩阵代表隶属于一级指标A1的各个二级指标B1、B2、B3两两比较,分别得到的差异程度量化值。M3这一判断矩阵代表隶属于一级指标A2的各个二级指标B4、B5、B6两两比较,分别得到的差异程度量化值,以此类推。如果还有其他层级的指标,则还可以按照上述方式建立起其他的判断矩阵。
需要说明的是,各个判断矩阵中各个元素上的具体取值可以是根据经验等确定。在建立起各个判断矩阵之后,就可以基于这种判断矩阵,来确定各个指标的权重。具体的,对于一个判断矩阵,可以计算出该判断矩阵的最大特征值以及该最大特征值对应的特征向量,该特征向量包含的分量值数量,与该矩阵的阶数相同,也就是与隶属于同一上级指标的指标数目相同,因此,就可以将该特征向量的各个分量值分别作为各个指标的权重。
例如,对于判断矩阵M1,计算得到其最大特征值为,该最大特征值对应的特征向量为,则可以确定出指标A1的权重为0.263,A2的权重为0.475,A3的权重为0.055,A4的权重为0.099,A5的权重为0.110。
对于其他各判断矩阵,也可以分别作类似处理,例如,对于判断矩阵M2,计算得到其最大特征值为,该最大特征值对应的特征向量为,则可以确定出B1的权重为0.595, B2的权重为0.277,A3的权重为0.129。总之,对于各个指标,均可以通过上述方式确定出其各自的权重。
需要说明的是,关于计算判断矩阵的最大特征值以及特征向量的具体实现,可以根据数学领域的计算公式来实现,这里不再详述。
另外,在具体实现时中,为了进一步提高权重确定过程中使用数据的合理性,在计算出各个判断矩阵的最大特征值以及对应的特征向量之后,还可以对判断矩阵进行一致性校验,如果校验通过,则将该最大特征值对应的特征向量中包含的各个分量值作为对应指标的权重,否则,如果校验不通过,则证明判断矩阵中的元素值可能存在一定的不合理性,因此,还可以对判断矩阵中的元素值进行调整,调整之后再重新计算最大特征值,重新进行一致性校验,直到校验通过后,再用特征向量确定各个指标的权重。
其中,所谓一致性校验,其目的是为了检验判断矩阵中各元素重要度之间的协调性,避免出现“A1比A2重要,A2比A3重要,而A3又比A1重要”这样的矛盾情况出现。具体在进行一致性校验时可以有多种方式,例如,在其中一种方式下,可以CR=CI/RI进行检验,其中CR是判断矩阵的随机一致性比率,一致性指标
Figure PCTCN2016090285-appb-000004
是平均随机一致性指标,RI是多次(例如500次以上)重复随机判断矩阵特征值的计算之后取算术平均得到的。一般认为,当一致性比例CR<0.1时,判断矩阵的一致性是可以接受的。否则需要对判断矩阵进行调整,直到达到满意的一致性为止。
例如,对于前述判断矩阵M1,λmax=5.073,n=5,也即矩阵M1的阶数,则可以计算出CI=(5.073-5)/(5-1)=0.018,通过查表可得RI=1.12,因此,可以计算出CR=0.018/1.12=0.016<0.1,因此,该判断矩阵的一致性满足条件,进而,直接将该最大特征值对应的特征向量中的各个分量,作为A1至A5各个指标的权重即可。其他各判断矩阵也都可以通过该方式进行一致性校验。通过计算,前述例子中的各个判断矩阵均能通过一致性校验,因此,可以分别基于前述矩阵确定出各个指标的权重。
S103:根据用户的历史操作记录,确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值;
在确定了各个指标的层级化结构以及各自的权重之后,就可以利用该模型,对各个品牌的指数量化值进行评估。具体的,由于最后一级也即第N级指标通常是可以通过统计等方式直接得到的,因此,可以首先确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值。
例如,在前述二级层级结构中,可以通过以下表2表达各个指标的关系以及权重信息。
表2
Figure PCTCN2016090285-appb-000005
上述表2中给出了各个二级指标的计算方式,可以根据销售平台中海量用户的历史操作记录,包括对某指定品牌的历史购买记录、收藏记录、浏览记录等等,分别计算出各个二级指标的统计值。关于各个二级指标的具体计算方式这里不再进行详述。当然,在具体实现时,如果N>2,则第N级上各个指标也可以通过上述方式计算统计值,再据此确定出第N级指标的量化值。并且,在具体实现时,上述指标也可以用该其他的统计值来计算,另外,根据具体指标的不同,对应的计算方法也可以进行调整,上述表2仅作为示例,而不应看作是对本申请保护范围的限制。
需要说明的是,在具体实现时,上述各个第N级指标的统计值,如果直接进行比较,则可能存在由于不是同一数量级,因此,相互之间不具有直接的可比性等情况。例如,某二级指标为对某品牌的浏览次数,另一二级指标为对该品牌的收藏次数,显然,浏览次数远大于收藏次数,两者并不在同一数量级上。因此,为了使得各个第N级指标的量化值之间具有可比性,并最终使得各个品牌指数的量化值具有可比性,也为了使得统计结果更平滑,可以对各个第N级指标的统计值进行平滑以及归一化处理,之后,再将得到的值确定出各个第N级指标的量化值。其中,具体的处理方式可以有多种,例如,其中一种具体的实现方式下,可以采用ln函数以及反正切函数对各个第N级指标的统计值进行处理,使得各个第N级指标的量化值均落在[0,1]范围内。
S104:利用各个第N级指标上的量化值、权重以及各个第N级指标所隶属的第N-1级指标,通过线性加权求和的方式,确定各个第N-1级各个指标量化值,并按照该步骤的方式依次向上级指标进行逐层计算,直到计算出第1级各个指标的量化值时,利用各个第1级指标上的量化值、权重,通过线性加权求和的方式,确定该品牌的品牌指数。
在确定出各个第N级指标上的量化值之后,就可以利用各个第N级指标上的量化值、权重以及各个第N级指标所隶属的第N-1级指标,通过线性加权求和的方式,确定各个第N-1级各个指标量化值。例如,在前述表2中,N=2,二级指标B1、B2、B3隶属于一级指标A1,因此,就可以计算出A1的量化值为:
VA1=VB1×w1+VB2×w2+VB3×w3
二级指标B4、B5、B6隶属于一级指标A2,因此,就可以计算出A2的量化值为:
VA2=VB4×w4+VB5×w5+VB6×w6
其他各个以及指标的量化值也可以通过上述方式进行线性加权求和。
当然,在具体实现时,如果N>2,则在计算出各个第N-1级各个指标量化值之后,还可以按照该步骤的方式依次向上级指标进行逐层计算,直到计算出第1级各个指标的 量化值时,利用各个第1级指标上的量化值、权重,通过线性加权求和的方式,确定该品牌的品牌指数。
例如,前述例子中,由于N=2,因此,计算出的A1、A2……A5的量化值已经是第1级指标,因此,就可以直接通过线性加权求和的方式,计算出对应品牌指数的量化值:
VZ=VA1×W1+VA2×W2+VA3×W3+VA4×W4+VA5×W5
类似的,其他品牌的品牌指数也可以通过上述各个步骤的方式计算出来。
总之,在本申请实施例中,可以从电子商务消费者的认知角度出发,研究购物网站的品牌个性与消费者品牌态度之间的相关关系,并以此来刻画品牌。并且,为了体现这种相关关系,可以通过建立层次化结构模型的方式来实现,在该模型中,包括多个层级的指标,最后一级指标为可统计或者可计算的值,并且,每个指标都可以具有各自的权重,用于表明其在刻画其所隶属的上一级指标时所起作用的相对重要程度。这样,就可以从最后一级指标开始,采用逐级向上计算的方式,最终得到品牌的指数量化值。在整个计算过程中,利用了网络消费者用户在销售平台中的历史操作信息进行数据挖掘,体现了电子商务交易平台的特性。并且,由于采用了层级化的结构,因此,对于各个指标而言,只需要确定出对于其所隶属的上一级别指标的重要程度,并进行逐级的计算,而不需要进行跨级别的权重确定,效率更高,计算结果也更准确。
在具体实现时,确定出各个品牌的品牌指数之后,就可以应用中具体的应用场景中,例如,在提供搜索结果、推荐信息等业务对象列表信息时,可以利用这种将品牌指数对提供的业务对象列表进行处理。例如,可以将品牌指数较高的品牌下的业务对象优先提供,等等。
或者,还可以首先对各个品牌进行分层,然后再将这种层次信息应用到具体的应用场景中。例如,在一种具体实现方式下,可以根据品牌指数的从小到大排序,做出品牌指数的散点图,然后,可以根据散点图中趋势线的斜率发生变化的点,找出对应的品牌指数,将这些品牌指数作为临界值,实现对品牌的分层。
例如,假设散点图如图3所示,在该图3中,假设有30个品牌,通过散点图可见,趋势线的斜率发生变化的点有A点和B点,因此,就可以将这两点对应的品牌指数(例如,分为为0.6和0.9),作为各个品牌层次的临界值,也即,将指数为0.9至1之间的品牌作为第一层,0.6至0.9之间的品牌作为第二层,0至0.6之间的品牌作为第三层。当然,在实际应用中,品牌数量众多,该图3仅作为示例性说明。
区分出品牌层次之后,可以将这种层次信息应用到具体的应用场景中。例如,在提 供业务对象信息列表时,可以提供按照品牌层次进行选择的操作选项,这样,消费者可以根据其实际需求,选择适合自己的目标层次的品牌,进而,服务器可以根据该操作选择确定被选中的目标层次,并提供该目标层次内的品牌的业务对象信息,这样,可以节省用户选商品的时间。
与本申请实施例提供的确定业务对象品牌指数信息的方法相对应,本申请实施例还提供了一种确定业务对象品牌指数信息的装置,参见图4,该装置具体可以包括:
指标确定单元401,用于预先确定用于刻画品牌指数的指标,所述指标包括N个层级,每个层级上包括至少一个指标,第n级指标通过至少一个第n+1级指标进行刻画;其中,n=1,2……N,N为大于1的正整数;
指标权重确定单元402,用于确定各个指标的权重,所述权重用于表示所述指标在刻画其所隶属的上一级指标时体现的相对重要程度;
指标量化值确定单元403,用于根据用户的历史操作记录,确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值;
品牌指数确定单元404,用于利用各个第N级指标上的量化值、权重以及各个第N级指标所隶属的第N-1级指标,确定各个第N-1级各个指标量化值,并按照该步骤的方式依次向上级指标进行逐层计算,直到计算出第1级各个指标的量化值时,利用各个第1级指标上的量化值、权重,确定该品牌的品牌指数。
具体实现时,所述指标权重确定单元402可以包括:
判断矩阵生成子单元,用于基于隶属于同一第n级指标的m个第n+1级指标,生成m×m的判断矩阵,所述判断矩阵中各个元素的取值通过以下方式确定:
将所述m个第n+1级指标的在刻画该第n级指标时体现的重要程度进行两两比对,根据比对结果的差异程度确定差异程度量化值,将所述两两比对得到的所述差异程度量化值,确定为所述判断矩阵中的各个元素;
计算子单元,用于计算所述判断矩阵的最大特征值以及该最大特征值对应的特征向量;
权重确定子单元,用于根据所述最大特征值对应的特征向量,确定所述m个第n+1级指标各自的权重。
其中,所述权重确定子单元包括:
一致性校验子单元,用于对所述判断矩阵进行一致性检验;
确定子单元,用于如果校验通过,则将所述最大特征值对应的特征向量中的各个分量值,确定为所述m个第n+1级指标各自的权重。
如果一致性校验未通过,还可以包括:
矩阵调整单元,用于如果一致性校验未通过,则对所述判断矩阵进行调整。
具体实现时,所述量化值确定单元403具体用于:
将所述统计值进行平滑以及归一化处理,得到所述第N级指标的量化值。
在一种具体的实现方式中,所述N=2,其中,一级指标包括品牌认知、功能定位、品质认定、品牌购买以及品牌忠诚度中的部分或全部;
其中,
所述品牌认知通过以下部分或全部二级指标进行刻画:搜索次数、收藏次数以及浏览次数;
所述功能定位通过以下部分或全部二级指标进行刻画:业务对象数量、有无指定类型店铺、店铺数量;
所述品质认定通过以下部分或全部二级指标进行刻画:品牌价位、高评分的描述相符率以及退款投诉率;
所述品牌购买通过以下部分或全部二级指标进行刻画:成交笔数、成交额、转化率、热销度、客单价;
所述品牌忠诚度通过以下二级指标进行刻画:重复购买率。
在实际应用中,该装置还可以包括:
业务对象列表处理单元,用于在向客户端提供业务对象列表信息时,根据所述品牌指数对提供的业务对象列表进行处理。
或者,该装置还可以包括:
品牌分层单元,用于根据各个品牌的品牌指数对品牌进行分层;
信息提供单元,用于利用分层信息向客户端提供业务对象信息。
其中,所述品牌分层单元包括:
排序子单元,用于将各个品牌的品牌指数进行排序;
散点图生成子单元,用于根据排序结果生成散点图;
目标点确定子单元,用于确定散点图中趋势线的斜率发生变化的目标点;
临界值确定子单元,用于将所述目标点处的品牌指数值确定为分层的临界值,以便根据所述临界值对品牌进行分层。
所述信息提供单元包括:
操作选项提供子单元,用于提供用于选择品牌层次的操作选项;
信息提供子单元,用于根据所述操作选择确定被选中的目标层次后,提供该目标层次内的品牌的业务对象信息。
通过本申请实施例,可以从电子商务消费者的认知角度出发,研究购物网站的品牌个性与消费者品牌态度之间的相关关系,并以此来刻画品牌。并且,为了体现这种相关关系,可以通过建立层次化结构模型的方式来实现,在该模型中,包括多个层级的指标,最后一级指标为可统计或者可计算的值,并且,每个指标都可以具有各自的权重,用于表明其在刻画其所隶属的上一级指标时所起作用的相对重要程度。这样,就可以从最后一级指标开始,采用逐级向上计算的方式,最终得到品牌的指数量化值。在整个计算过程中,利用了网络消费者用户在销售平台中的历史操作信息进行数据挖掘,体现了电子商务交易平台的特性。并且,由于采用了层级化的结构,因此,对于各个指标而言,只需要确定出对于其所隶属的上一级别指标的重要程度,并进行逐级的计算,而不需要进行跨级别的权重确定,效率更高,计算结果也更准确。
通过以上的实施方式的描述可知,本领域的技术人员可以清楚地了解到本申请可借助软件加必需的通用硬件平台的方式来实现。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品可以存储在存储介质中,如ROM/RAM、磁碟、光盘等,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行本申请各个实施例或者实施例的某些部分所述的方法。
本说明书中的各个实施例均采用递进的方式描述,各个实施例之间相同相似的部分互相参见即可,每个实施例重点说明的都是与其他实施例的不同之处。尤其,对于系统或系统实施例而言,由于其基本相似于方法实施例,所以描述得比较简单,相关之处参见方法实施例的部分说明即可。以上所描述的系统及系统实施例仅仅是示意性的,其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性劳动的情况下,即可以理解并实施。
以上对本申请所提供的确定业务对象品牌指数信息的方法及装置,进行了详细介绍,本文中应用了具体个例对本申请的原理及实施方式进行了阐述,以上实施例的说明只是 用于帮助理解本申请的方法及其核心思想;同时,对于本领域的一般技术人员,依据本申请的思想,在具体实施方式及应用范围上均会有改变之处。综上所述,本说明书内容不应理解为对本申请的限制。

Claims (20)

  1. 一种确定业务对象品牌指数信息的方法,其特征在于,包括:
    预先确定用于刻画品牌指数的指标,所述指标包括N个层级,每个层级上包括至少一个指标,第n级指标通过至少一个第n+1级指标进行刻画;其中,n=1,2……N,N为大于1的正整数;
    确定各个指标的权重,所述权重用于表示所述指标在刻画其所隶属的上一级指标时体现的相对重要程度;
    根据用户的历史操作记录,确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值;
    利用各个第N级指标上的量化值、权重以及各个第N级指标所隶属的第N-1级指标,确定各个第N-1级各个指标量化值,并按照该步骤的方式依次向上级指标进行逐层计算,直到计算出第1级各个指标的量化值时,利用各个第1级指标上的量化值、权重,确定该品牌的品牌指数。
  2. 根据权利要求1所述的方法,其特征在于,所述确定各个指标的权重,包括:
    基于隶属于同一第n级指标的m个第n+1级指标,生成m×m的判断矩阵,所述判断矩阵中各个元素的取值通过以下方式确定:
    将所述m个第n+1级指标的在刻画该第n级指标时体现的重要程度进行两两比对,根据比对结果的差异程度确定差异程度量化值,将所述两两比对得到的所述差异程度量化值,确定为所述判断矩阵中的各个元素;
    计算所述判断矩阵的最大特征值以及该最大特征值对应的特征向量;
    根据所述最大特征值对应的特征向量,确定所述m个第n+1级指标各自的权重。
  3. 根据权利要求2所述的方法,其特征在于,所述根据所述最大特征值对应的特征向量,确定所述m个第n+1级指标各自的权重,包括:
    利用一致性指标、随机一致性指标和一致性比率,对所述判断矩阵进行一致性检验;
    如果校验通过,则将所述最大特征值对应的特征向量中的各个分量值,确定为所述m个第n+1级指标各自的权重。
  4. 根据权利要求3所述的方法,其特征在于,还包括:
    如果一致性校验未通过,则对所述判断矩阵进行调整。
  5. 根据权利要求1至4任一项所述的方法,其特征在于,所述根据该统计值确定各个第N级指标的量化值,包括:
    将所述统计值进行平滑以及归一化处理,得到所述第N级指标的量化值。
  6. 根据权利要求1至4任一项所述的方法,其特征在于,所述N=2,其中,一级指标包括品牌认知、功能定位、品质认定、品牌购买以及品牌忠诚度中的部分或全部;
    其中,
    所述品牌认知通过以下部分或全部二级指标进行刻画:搜索次数、收藏次数以及浏览次数;
    所述功能定位通过以下部分或全部二级指标进行刻画:业务对象数量、有无指定类型店铺、店铺数量;
    所述品质认定通过以下部分或全部二级指标进行刻画:品牌价位、高评分的描述相符率以及退款投诉率;
    所述品牌购买通过以下部分或全部二级指标进行刻画:成交笔数、成交额、转化率、热销度、客单价;
    所述品牌忠诚度通过以下二级指标进行刻画:重复购买率。
  7. 根据权利要求1至4任一项所述的方法,其特征在于,还包括:
    在向客户端提供业务对象列表信息时,根据所述品牌指数对提供的业务对象列表进行处理。
  8. 根据权利要求1至4任一项所述的方法,其特征在于,还包括:
    根据各个品牌的品牌指数对品牌进行分层;
    利用分层信息向客户端提供业务对象信息。
  9. 根据权利要求8所述的方法,其特征在于,所述根据各个品牌的品牌指数对品牌进行分层,包括:
    将各个品牌的品牌指数进行排序;
    根据排序结果生成散点图;
    确定散点图中趋势线的斜率发生变化的目标点;
    将所述目标点处的品牌指数值确定为分层的临界值,以便根据所述临界值对品牌进行分层。
  10. 根据权利要求8所述的方法,其特征在于,所述利用分层信息向客户端提供业务对象信息,包括:
    提供用于选择品牌层次的操作选项;
    根据所述操作选择确定被选中的目标层次后,提供该目标层次内的品牌的业务对象 信息。
  11. 一种确定业务对象品牌指数信息的装置,其特征在于,包括:
    指标确定单元,用于预先确定用于刻画品牌指数的指标,所述指标包括N个层级,每个层级上包括至少一个指标,第n级指标通过至少一个第n+1级指标进行刻画;其中,n=1,2……N,N为大于1的正整数;
    指标权重确定单元,用于确定各个指标的权重,所述权重用于表示所述指标在刻画其所隶属的上一级指标时体现的相对重要程度;
    指标量化值确定单元,用于根据用户的历史操作记录,确定同一品牌分别在各个第N级指标上的统计值,并根据该统计值确定各个第N级指标的量化值;
    品牌指数确定单元,用于利用各个第N级指标上的量化值、权重以及各个第N级指标所隶属的第N-1级指标,确定各个第N-1级各个指标量化值,并按照该步骤的方式依次向上级指标进行逐层计算,直到计算出第1级各个指标的量化值时,利用各个第1级指标上的量化值、权重,确定该品牌的品牌指数。
  12. 根据权利要求11所述的装置,其特征在于,所述指标权重确定单元包括:
    判断矩阵生成子单元,用于基于隶属于同一第n级指标的m个第n+1级指标,生成m×m的判断矩阵,所述判断矩阵中各个元素的取值通过以下方式确定:
    将所述m个第n+1级指标的在刻画该第n级指标时体现的重要程度进行两两比对,根据比对结果的差异程度确定差异程度量化值,将所述两两比对得到的所述差异程度量化值,确定为所述判断矩阵中的各个元素;
    计算子单元,用于计算所述判断矩阵的最大特征值以及该最大特征值对应的特征向量;
    权重确定子单元,用于根据所述最大特征值对应的特征向量,确定所述m个第n+1级指标各自的权重。
  13. 根据权利要求12所述的装置,其特征在于,所述权重确定子单元包括:
    一致性校验子单元,用于对所述判断矩阵进行一致性检验;
    确定子单元,用于如果校验通过,则将所述最大特征值对应的特征向量中的各个分量值,确定为所述m个第n+1级指标各自的权重。
  14. 根据权利要求13所述的装置,其特征在于,还包括:
    矩阵调整单元,用于如果一致性校验未通过,则对所述判断矩阵进行调整。
  15. 根据权利要求11至14任一项所述的装置,其特征在于,所述量化值确定单元 具体用于:
    将所述统计值进行平滑以及归一化处理,得到所述第N级指标的量化值。
  16. 根据权利要求11至14任一项所述的装置,其特征在于,所述N=2,其中,一级指标包括品牌认知、功能定位、品质认定、品牌购买以及品牌忠诚度中的部分或全部;
    其中,
    所述品牌认知通过以下部分或全部二级指标进行刻画:搜索次数、收藏次数以及浏览次数;
    所述功能定位通过以下部分或全部二级指标进行刻画:业务对象数量、有无指定类型店铺、店铺数量;
    所述品质认定通过以下部分或全部二级指标进行刻画:品牌价位、高评分的描述相符率以及退款投诉率;
    所述品牌购买通过以下部分或全部二级指标进行刻画:成交笔数、成交额、转化率、热销度、客单价;
    所述品牌忠诚度通过以下二级指标进行刻画:重复购买率。
  17. 根据权利要求11至14任一项所述的装置,其特征在于,还包括:
    业务对象列表处理单元,用于在向客户端提供业务对象列表信息时,根据所述品牌指数对提供的业务对象列表进行处理。
  18. 根据权利要求11至14任一项所述的装置,其特征在于,还包括:
    品牌分层单元,用于根据各个品牌的品牌指数对品牌进行分层;
    信息提供单元,用于利用分层信息向客户端提供业务对象信息。
  19. 根据权利要求18所述的装置,其特征在于,所述品牌分层单元包括:
    排序子单元,用于将各个品牌的品牌指数进行排序;
    散点图生成子单元,用于根据排序结果生成散点图;
    目标点确定子单元,用于确定散点图中趋势线的斜率发生变化的目标点;
    临界值确定子单元,用于将所述目标点处的品牌指数值确定为分层的临界值,以便根据所述临界值对品牌进行分层。
  20. 根据权利要求18所述的装置,其特征在于,所述信息提供单元包括:
    操作选项提供子单元,用于提供用于选择品牌层次的操作选项;
    信息提供子单元,用于根据所述操作选择确定被选中的目标层次后,提供该目标层次内的品牌的业务对象信息。
PCT/CN2016/090285 2015-07-27 2016-07-18 确定业务对象品牌指数信息的方法及装置 WO2017016403A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510447140.8 2015-07-27
CN201510447140.8A CN106408306A (zh) 2015-07-27 2015-07-27 确定业务对象品牌指数信息的方法及装置

Publications (1)

Publication Number Publication Date
WO2017016403A1 true WO2017016403A1 (zh) 2017-02-02

Family

ID=57884118

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/090285 WO2017016403A1 (zh) 2015-07-27 2016-07-18 确定业务对象品牌指数信息的方法及装置

Country Status (2)

Country Link
CN (1) CN106408306A (zh)
WO (1) WO2017016403A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111369300A (zh) * 2020-03-13 2020-07-03 浙江师范大学 满意度评价方法、装置、计算机设备和可读存储介质

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107808300A (zh) * 2017-09-15 2018-03-16 安徽中瑞通信科技股份有限公司 一种基于无线领域营销平台的广告推送系统及推送方法
CN108564314A (zh) * 2018-06-15 2018-09-21 杭州后博科技有限公司 一种出行时不同品牌共享单车的选择方法及系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20080004964A (ko) * 2006-07-07 2008-01-10 주식회사 동서리서치 브랜드의 자산가치평가 시스템 및 그 방법
CN104123656A (zh) * 2014-07-14 2014-10-29 广东省标准化研究院 一种基于层次分析法的信用评价方法
CN104424572A (zh) * 2013-08-28 2015-03-18 国家广播电影电视总局广播科学研究院 一种数字家庭购物商户信用评价方法
CN104732429A (zh) * 2015-04-21 2015-06-24 上海商学院 综合消费者品牌偏好的修正哈弗模型

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20080004964A (ko) * 2006-07-07 2008-01-10 주식회사 동서리서치 브랜드의 자산가치평가 시스템 및 그 방법
CN104424572A (zh) * 2013-08-28 2015-03-18 国家广播电影电视总局广播科学研究院 一种数字家庭购物商户信用评价方法
CN104123656A (zh) * 2014-07-14 2014-10-29 广东省标准化研究院 一种基于层次分析法的信用评价方法
CN104732429A (zh) * 2015-04-21 2015-06-24 上海商学院 综合消费者品牌偏好的修正哈弗模型

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111369300A (zh) * 2020-03-13 2020-07-03 浙江师范大学 满意度评价方法、装置、计算机设备和可读存储介质

Also Published As

Publication number Publication date
CN106408306A (zh) 2017-02-15

Similar Documents

Publication Publication Date Title
US11107118B2 (en) Management of the display of online ad content consistent with one or more performance objectives for a webpage and/or website
US11676087B2 (en) Systems and methods for vulnerability assessment and remedy identification
US11568334B2 (en) Adaptive workflow definition of crowd sourced tasks and quality control mechanisms for multiple business applications
Chen et al. Evaluating the enhancement of corporate social responsibility websites quality based on a new hybrid MADM model
CN109213936B (zh) 商品搜索方法和装置
US20210256445A1 (en) System and method for matching resource capacity with client resource needs
US20160132811A1 (en) Influential Peers
Zhao et al. Group preference aggregation: A nash equilibrium approach
Yildiz et al. A two-phased multi-criteria decision-making approach for selecting the best smartphone
US20240005368A1 (en) Systems and methods for an intelligent sourcing engine for study participants
WO2017016403A1 (zh) 确定业务对象品牌指数信息的方法及装置
CN113781139A (zh) 物品推荐方法、物品推荐装置、设备和介质
Yuan et al. Evaluation of service quality continuous improvement in coffee shops
US20240177184A1 (en) Self-learning valuation and access to digital content
Novotová Exploring customer loyalty to fashion brands on facebook fan pages
CN109829593B (zh) 目标对象的信用度确定方法、装置、存储介质及电子装置
Yu Evaluation of E-commerce service quality using the analytic hierarchy process
JP4962950B2 (ja) ネットワーク上のユーザに対するレコメンデーションの方法、レコメンデーションサーバ及びプログラム
CN112581281A (zh) 产品推荐方法及装置、存储介质及电子设备
Chen Optimizing MPBSM resource allocation based on revenue management: a China mobile Sichuan case
Lou et al. Combination of evaluation methods for assessing the quality of service for express delivery industry
Wang et al. Evaluating of factor affecting the cloud-based collaboration system based on FAHP
Aggarwal et al. Success Factors of M-Commerce: A Customer Perspective
Devi et al. Hybrid Filtering Algorithm in Event Manager Partner Recommendation System
Sun et al. A Study of Factors Influencing Customers' Online Shopping Behavior Based on ANP

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16829769

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16829769

Country of ref document: EP

Kind code of ref document: A1