CN106126499A - User satisfaction and loyalty degree analysis method and device - Google Patents

User satisfaction and loyalty degree analysis method and device Download PDF

Info

Publication number
CN106126499A
CN106126499A CN201610462852.1A CN201610462852A CN106126499A CN 106126499 A CN106126499 A CN 106126499A CN 201610462852 A CN201610462852 A CN 201610462852A CN 106126499 A CN106126499 A CN 106126499A
Authority
CN
China
Prior art keywords
type
data
user
dimension
comments
Prior art date
Application number
CN201610462852.1A
Other languages
Chinese (zh)
Inventor
赵斐
朱建坤
克远
于芝涛
Original Assignee
青岛海信传媒网络技术有限公司
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 青岛海信传媒网络技术有限公司 filed Critical 青岛海信传媒网络技术有限公司
Priority to CN201610462852.1A priority Critical patent/CN106126499A/en
Publication of CN106126499A publication Critical patent/CN106126499A/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/20Handling natural language data
    • G06F17/27Automatic analysis, e.g. parsing
    • G06F17/2705Parsing
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/20Handling natural language data
    • G06F17/27Automatic analysis, e.g. parsing
    • G06F17/2765Recognition
    • G06F17/277Lexical analysis, e.g. tokenisation, collocates
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/20Handling natural language data
    • G06F17/27Automatic analysis, e.g. parsing
    • G06F17/2785Semantic analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce, e.g. shopping or e-commerce
    • G06Q30/02Marketing, e.g. market research and analysis, surveying, promotions, advertising, buyer profiling, customer management or rewards; Price estimation or determination
    • G06Q30/0201Market data gathering, market analysis or market modelling

Abstract

The invention provides a user satisfaction and loyalty degree analysis method and device; the method comprises the following steps: obtaining user comment data, wherein the user comment data is formed by users according to usage results; determining comment dimensions types and common result types corresponding to the user comment data according to the user comment data; determining the user satisfaction and loyalty degree result according to the comment dimensions types and common result types corresponding to the user comment data. The comment data actively started by the users in the method is the behavior data formed after user real usage, so the user satisfaction and loyalty degree result obtained by analyzing the comment data can express real opinions of users on certain product, thus ensuring the credibility of the user satisfaction and loyalty degree result.

Description

用户满意度和忠诚度分析方法及装置 Customer satisfaction and loyalty analysis method and apparatus

技术领域 FIELD

[0001] 本发明设及计算机技术,尤其设及一种用户满意度和忠诚度分析方法及装置。 [0001] The present invention is provided computer technology and, in particular, is provided, and one user satisfaction and loyalty analysis method and apparatus.

背景技术 Background technique

[0002] 用户满意度和忠诚度是衡量一个产品的顾客对企业产品或服务的满意程度及忠诚程度,通过获取用户满意度和忠诚度的数据,并对运些数据进行分析,可W帮助企业找出提高产品或服务水平的切入点,W及与竞争对手相比的优势及劣势。 [0002] user satisfaction and loyalty is a measure of a product's customer business product or service satisfaction and loyalty degree, by obtaining data user satisfaction and loyalty, and operational analysis of these data, can help businesses W improve product or service to find out the level of entry point, W and compared with competitors strengths and weaknesses.

[0003] 现有技术中,主要通过问卷调查的方式来获取用户满意度和忠诚度数据,并对运些数据进行分析,进而根据分析结果采取相应的措施。 [0003] prior art, mainly to obtain customer satisfaction and loyalty data through questionnaires, and some transport data for analysis, and then take appropriate action based on the analysis results. 其中,问卷调查的方式可W包括发放问卷调查表、电话询问、通过访谈对话询问等,在运些方式中,都是先由问卷发起者提出一个或多个问题,然后再由被调查用户回答问题。 Among them, the way the survey W may include issuing questionnaires, telephone inquiries, interviews and other inquiries dialogue, in operation in some way, they are the first by the initiator of the questionnaire asked one or more questions, then answered by the users surveyed problem.

[0004] 但是,通过现有技术所获取到的用户满意度和忠诚度数据,并不能完全反应用户对于产品的实际看法,导致通过运些数据所分析出的结果的可信度不高。 [0004] However, the prior art acquired through customer satisfaction and loyalty data, and not fully reflect the actual users think about the product, leading to the results of the analysis of these data by running out of credibility is not high.

发明内容 SUMMARY

[0005] 本发明提供一种用户满意度和忠诚度分析方法及装置,用于解决使用现有技术所导致的用户满意度和忠诚度分析结果可信度不高的问题。 [0005] The present invention provides a user satisfaction and loyalty analysis method and apparatus for solving the problem of low reliability of the results of user satisfaction and loyalty analyzed using prior art caused.

[0006] 本发明第一方面提供一种用户满意度和忠诚度分析方法,包括: [0006] a first aspect the present invention provides a user satisfaction and loyalty analysis method, comprising:

[0007] 获取用户评论数据,所述用户评论数据由用户根据使用结果生成; [0007] The data acquiring user reviews, user reviews the data generated by the user according to the results of use;

[000引根据所述用户评论数据,确定所述用户评论数据对应的评论维度类型W及评论结果类型; [000 reviews cited data according to the user, determines that the user reviews the data type corresponding to the dimension W and review comments result types;

[0009] 根据用户评论数据对应的评论维度类型W及评论结果类型,确定用户满意度和忠诚度结果。 [0009] The types of dimensions W and review comments result data corresponding to the type of user comments, determines user satisfaction and loyalty results.

[0010] 进一步地,所述获取用户评论数据,包括: [0010] Further, the user reviews data acquisition, comprising:

[0011] 从至少一个网络渠道获取所述用户评论数据,所述网络渠道至少包括:论坛、博客、电商平台。 [0011] User comments obtaining the data from at least one network channel, the network channels, including at least: forum, blog, electronic business platform.

[0012] 进一步地,所述根据所述用户评论数据,确定所述用户评论数据对应的评论维度类型W及评论结果类型,包括: [0012] Further, according to the user reviews the data, determining that the user reviews the data type corresponding to the dimension W and review comments result types, comprising:

[0013] 识别所述用户评论数据中的关键词; [0013] identifying the user reviews data Image;

[0014] 根据所述用户评论数据中的关键词,确定所述用户评论数据对应的评论维度类型; [0014] The user reviews the data keywords to determine the dimensions of the type of comment data corresponding to the user reviews;

[0015] 对所述用户评论数据进行语义分析,获取所述用户评论数据对应的评论结果类型。 [0015] Type Comment semantic analysis result of the user reviews data, obtaining data corresponding to the user reviews.

[0016] 进一步地,所述用户评论数据的评论维度类型包括多个评论父维度类型,每个评论父维度类型下包括多个评论子维度类型;所述根据所述用户评论数据中的关键词,确定所述用户评论数据对应的评论维度类型,包括: [0016] Further, the user reviews a review dimension type comprising a plurality of data types comments parent dimension, comprising a plurality of sub-types of dimensions at each comment Comments parent dimension type; according to the user reviews the data in the Image determining the user reviews data corresponding comments dimension type, comprising:

[0017] 判断所述用户评论数据中的第一关键词是否与第一评论父维度类型对应的关键词一致,若是,则: [0017] determines whether the user reviews the data in the first keyword and a first comment corresponding to the type of parent dimension same keyword, and if so, then:

[0018] 确定所述用户评论数据的评论父维度类型为所述第一评论父维度类型;W及, [0018] determining a type of the parent dimension comment user reviews the data as a first type of parent dimension reviews; W is and,

[0019] 判断所述用户评论数据中的第二关键词是否与第一评论子维度类型对应的关键词一致,若是,则确定所述用户评论数据的评论子维度类型为所述第一评论子维度类型,其中,所述第一评论子维度类型为所述第一评论父维度类型下所包括的评论子维度类型。 [0019] Analyzing the data in the second user reviews keyword coincides with a first dimension corresponding to the type of comment sub keyword, and if so, the type of comment sub-dimensions user reviews the data is determined to be a first sub reviews dimension type, wherein the first type of comment sub dimension in said first dimension of the type comprising a review of the parent comment sub-dimension type.

[0020] 进一步地,所述对所述用户评论数据进行语义分析,获取所述用户评论数据对应的评论结果类型,包括: [0020] Further, a semantic analysis of the user reviews data, obtaining data corresponding to the user reviews a review result types, comprising:

[0021] 对所述用户评论数据进行语义分析,判断语义分析结果是否与第一评论结果类型一致,若是,则确定所述用户评论数据对应的评论结果类型为所述第一评论结果类型。 [0021] The semantic analysis of the user reviews the data, it is determined whether the semantic analysis result is consistent with the result type of the first comments, and if so, determining that the user reviews the data corresponding to the first type of result of review comments result type.

[0022] 进一步地,所述根据用户评论数据对应的评论维度类型W及评论结果类型,确定用户满意度和忠诚度结果,包括: [0022] Further, the comment data corresponding to the user type of the dimension W and review comments result type, determines user satisfaction and loyalty results, comprising:

[0023] 获取每种评论维度类型的用户评论数据中每种评论结果类型的比例; [0023] Gets the proportion of each type of user reviews comments dimension data of each type of comment results;

[0024] 根据每种评论维度类型的用户评论数据中每种评论结果类型的比例,确定每种评论维度类型对应的用户满意度和忠诚度结果。 [0024] The proportion of each type of user reviews comments dimension data of each type of comment result, determines user satisfaction and loyalty results for each dimension corresponding to the type of comment.

[0025] 进一步地,所述根据用户评论数据对应的评论维度类型W及评论结果类型,确定用户满意度和忠诚度结果,包括: [0025] Further, the comment data corresponding to the user type of the dimension W and review comments result type, determines user satisfaction and loyalty results, comprising:

[0026] 获取所有用户评论数据中每种评论结果类型的比例; [0026] All user reviews the data acquisition percentage of each type of comment results;

[0027] 根据所有用户评论数据中每种评论结果类型的比例,确定整体的用户满意度和忠诚度结果。 [0027] According to the proportion of all user reviews of each data type of the result of a review to determine overall customer satisfaction and loyalty results.

[0028] 进一步地,所述第一评论结果类型为推荐型评论、消极型评论、贬低型评论中的任意一种。 [0028] Further, the result of the first type is recommended type review comments, reviews passive type, any one of comments demeaning type.

[0029] 本发明第二方面提供一种用户满意度和忠诚度分析装置,包括: [0029] The second aspect of the present invention provides a user satisfaction and loyalty analysis apparatus, comprising:

[0030] 获取模块,用于获取用户评论数据,所述用户评论数据由用户根据使用结果生成; [0030] obtaining module, configured to obtain user reviews the data, the user reviews the data generated by the user according to the results of use;

[0031] 第一确定模块,用于根据所述用户评论数据,确定所述用户评论数据对应的评论维度类型W及评论结果类型; [0031] The first determining module configured to review the data according to the user, determines the user reviews and comments dimension W type comment data corresponding to the type of result;

[0032] 第二确定模块,用于根据用户评论数据对应的评论维度类型W及评论结果类型, 确定用户满意度和忠诚度结果。 [0032] The second determination module, according to type comments dimension W type user reviews the results and comment data corresponding to the determined customer satisfaction and loyalty results.

[0033] 进一步地,所述获取模块包括: [0033] Further, the obtaining module comprises:

[0034] 第一获取单元,用于从至少一个网络渠道获取所述用户评论数据,所述网络渠道至少包括:论坛、博客、电商平台。 [0034] The first obtaining unit configured to obtain the user reviews data from at least one network channel, the network channels, including at least: forum, blog, electronic business platform.

[0035] 进一步地,第一确定模块包括: [0035] Further, the first determination module comprises:

[0036] 识别单元,用于识别用户评论数据中的关键词。 [0036] identification unit for identifying a user reviews the data keywords.

[0037] 确定单元,用于根据用户评论数据中的关键词,确定用户评论数据对应的评论维度类型。 [0037] The determination means according to the keyword data, user reviews, user reviews the data corresponding to the determined dimension type comments.

[0038] 分析单元,用于对用户评论数据进行语义分析,获取用户评论数据对应的评论结果类型。 [0038] analysis unit for user reviews data semantic analysis, data corresponding to the acquired user reviews comments result type.

[0039] 进一步地,用户评论数据的评论维度类型包括多个评论父维度类型,每个评论父维度类型下包括多个评论子维度类型。 [0039] Further, the type of comment dimensional data comprises a plurality of user reviews comments parent dimension type, comprising a plurality of sub-types of dimensions at each review comments dimension type parent.

[0040] 相应地,确定单元具体用于: [0040] Accordingly, the determining unit is configured to:

[0041] 判断用户评论数据中的第一关键词是否与第一评论父维度类型对应的关键词一致,若是,则: [0041] Analyzing the data in the user reviews a first keyword and the keyword coincides first comment corresponding to the type of parent dimension, and if so, then:

[0042] 确定用户评论数据的评论父维度类型为第一评论父维度类型;W及,判断用户评论数据中的第二关键词是否与第一评论子维度类型对应的关键词一致,若是,则确定用户评论数据的评论子维度类型为第一评论子维度类型,其中,第一评论子维度类型为第一评论父维度类型下所包括的评论子维度类型。 [0042] determined that the user reviews the data type is a first dimension of the parent comment Comments type parent dimension; W and consistent, user reviews determines whether the keyword data in the second first dimension corresponding to the type of comment sub-keyword, if yes, determining a user comment sub-dimensions comment data type is a first type of comment sub-dimension, wherein the first type of comment sub-dimensions at a first type of sub-dimensions reviews comments included parent dimension types.

[0043] 分析单元具体用于: [0043] The analyzing unit is configured to:

[0044] 对用户评论数据进行语义分析,判断语义分析结果是否与第一评论结果类型一致,若是,则确定用户评论数据对应的评论结果类型为第一评论结果类型。 [0044] The user reviews the data semantic analysis to determine the semantic analysis result is consistent with the result type of the first comments, and if yes, determining that the user reviews the data corresponding to a first type of review comments result of the result type.

[0045] 进一步地,第二确定模块包括: [0045] Further, the second determination module comprises:

[0046] 第一获取单元,用于获取每种评论维度类型的用户评论数据中每种评论结果类型的比例。 [0046] a first obtaining unit, configured to obtain the proportion of each type of user reviews comments dimension data of each type of review results.

[0047] 第一确定单元,用于根据每种评论维度类型的用户评论数据中每种评论结果类型的比例,确定每种评论维度类型对应的用户满意度和忠诚度结果。 [0047] The first determining unit, according to the proportion of each type of user reviews comments dimension data of each type of comment result, determines user satisfaction and loyalty results for each dimension corresponding to the type of comment.

[004引进一步地,第二确定模块还包括: [004 cited Further, the second determination module further comprises:

[0049] 第二获取单元,用于获取所有用户评论数据中每种评论结果类型的比例。 [0049] The second obtaining unit, configured to obtain the proportions of all user reviews the data type of each review the results.

[0050] 第一确定单元,用于根据所有用户评论数据中每种评论结果类型的比例,确定整体的用户满意度和忠诚度结果。 [0050] The first determining unit, according to the proportions of all user reviews the data type of each review the results to determine customer satisfaction and loyalty overall results.

[0051] 本发明所提供的用户满意度和忠诚度分析方法及装置,通过获取用户主动发起的评论数据,并基于该评论数据确定其对应的评论维度类型W及评论结果类型,得到用户满意度和忠诚度结果。 [0051] User satisfaction with the present invention is provided and loyalty analysis method and apparatus, by obtaining the user initiates comment data, and based on the comment data determining corresponding comments dimension type W Reviews and result type, obtained user satisfaction and loyalty results. 由于用户主动发起的评论数据是用户经过实际使用后所生成的"行为数据",因此经过分析运些评论数据所得到的用户满意度和忠诚度结果能够反应用户对某产品的实际看法,从而保证用户满意度和忠诚度结果的可信度。 Since the user initiates review data after the user through the actual use of the generated "behavioral data" and therefore through the analysis of user satisfaction and loyalty results shipped some comment data obtained can react users actually think of a product, thus ensuring customer satisfaction and loyalty, confidence in the results. 进一步地,本实施例中通过对用户评论数据进行分析后确定出用户评论数据对应的评论维度类型W及评论结果类型, 基于运些评论维度类型W及评论结果类型,可W得到用户在不同维度上或者整体上对于某产品的评价结果,从而帮助企业识别出更有针对性的改进方向和改进点。 Further, the present embodiment by determining the user comment data corresponding to the comment types of dimensions W and reviews the results of the type the user reviews the data analysis based on operation of these reviews types of dimensions W and reviews the results of type, W obtained in Example users in different dimensions or on the overall results of the evaluation of a product to help companies identify more focused direction for improvement and points of improvement.

附图说明 BRIEF DESCRIPTION

[0052] 为了更清楚地说明本发明或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍,显而易见地,下面描述中的附图是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可W根据运些附图获得其他的附图。 [0052] In order to more clearly illustrate the present invention or the technical solution in the prior art, accompanying drawings for describing the embodiments are introduced briefly described in the embodiments or the prior art are required Apparently, the following description is attached FIG some embodiments of the present invention, those of ordinary skill in the art is concerned, without any creative effort, and W may be transported in accordance with these drawings other drawings.

[0053] 图1为本发明提供的用户满意度和忠诚度分析方法实施例一的流程示意图; Example of a flow [0053] FIG customer satisfaction and loyalty analysis method of the present invention provides a schematic embodiment;

[0054] 图2为本发明提供的用户满意度和忠诚度分析方法实施例二的流程示意图; Process Example II [0054] 2 user satisfaction and loyalty analysis method of the present invention provides a schematic embodiment;

[0055] 图3为本发明提供的用户满意度和忠诚度分析方法实施例=的流程示意图; = Example of Process [0055] 3 customer satisfaction and loyalty analysis method of the present invention provides a schematic view;

[0056] 图4为本发明提供的用户满意度和忠诚度分析方法实施例四的流程示意图; Process Example IV [0056] FIG customer satisfaction and loyalty analysis method of the present invention provides a schematic embodiment;

[0057] 图5为本发明提供的用户满意度和忠诚度分析装置实施例一的模块结构图; User satisfaction and loyalty [0057] Figure 5 provides the analyzing apparatus of the present invention, the module structure diagram of an embodiment;

[0058] 图6为本发明提供的用户满意度和忠诚度分析装置实施例二的模块结构图; [0058] FIG customer satisfaction and loyalty to the present invention provides an analytical device module configuration diagram according to a second embodiment;

[0059] 图7为本发明提供的用户满意度和忠诚度分析装置实施例=的模块结构图; Block configuration diagram of an embodiment [0059] FIG 7 customer satisfaction and loyalty analysis apparatus of the present invention provides a =;

[0060] 图8为本发明提供的用户满意度和忠诚度分析装置实施例四的模块结构图; [0060] FIG 8 customer satisfaction and loyalty analysis apparatus of the present invention to provide a block configuration diagram according to a fourth embodiment;

[0061] 图9为本发明提供的用户满意度和忠诚度分析装置实施例五的模块结构图。 [0061] FIG 9 customer satisfaction and loyalty to the present invention provides an analytical device module according to a fifth embodiment of the structure of FIG.

具体实施方式 Detailed ways

[0062] 为使本发明的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本发明一部分实施例,而不是全部的实施例。 [0062] To make the objectives, technical solutions, and advantages of the present invention will become more apparent below in conjunction with the present invention in the accompanying drawings, technical solutions of embodiments of the present invention are clearly and completely described, obviously, the described EXAMPLE some embodiments of the present invention rather than all embodiments. 基于本发明中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。 Based on the embodiments of the present invention, those of ordinary skill in the art to make all other embodiments without creative work obtained by, it falls within the scope of the present invention.

[0063] 现有通过问卷调查方式获取用户满意度和忠诚度数据的过程中,由问卷发起者提出问题,被调查用户回答问题。 [0063] acquiring existing customer satisfaction and loyalty survey data by way of the process, ask questions from the questionnaire initiator, answered questionnaires users. 首先,对于问卷中的问题,被调查用户可能并没有遇到过,因此,被调查用户所提供的答案,只是一种"态度数据",而并不是被调查者已经使用某产品之后的"行为数据",因此,其参考价值并不能完全保证。 First, the question in the questionnaire, surveyed users may not have encountered before, therefore, investigated the answers provided by the user, just an "attitudinal data", rather than "behavior after the respondents have been using a product data "and therefore, its reference value is not guaranteed. 其次,同样的调查问卷,在不同的时间发起、由不同的问卷发起者发起,其调查的结果可能并不相同。 Secondly, the same survey, initiated at different times, different questionnaire initiated by the initiator of the results of its investigation may not be the same. 例如,如果调查问卷是在被调查者的闲暇时间发起,那么调查结果一般比较好,而如果是在被调查者忙碌的时间发起, 则被调查者提供的答案可能和闲暇时完全不同。 For example, if the survey respondents in leisure time of launch, the survey results are generally better, and if it is initiated at a busy time of respondents, the respondents may provide answers and leisure completely different. 另外,调查问卷的问题设置也会影响最终的调查结果。 In addition, the survey question set will influence the final survey results. 总之,现有使用调查问卷的方式来获取用户满意度和忠诚度数据的方式,容易受到各种因素的影响,因此,并不能完全反应用户对于产品的实际看法,导致通过运些数据所分析出的结果的可信度不高。 In short, using existing questionnaire way to get customer satisfaction and loyalty the way data is easily affected by various factors, therefore, it can not completely reacted the user actually views on the product, resulting in the analysis of the data by running some the reliability of the results is not high.

[0064] 本发明基于上述问题,提出一种用户满意度和忠诚度测量方法,通过收集用户评论数据的方式来获取用户对于某产品所主动提供的满意度和忠诚度数据,W提升满意度和忠诚度结果的可信度。 [0064] The present invention is based on the above problems, we propose a method of customer satisfaction and loyalty measurements, to obtain data collected by way of user reviews user satisfaction and loyalty to the data of a certain product unsolicited, W and improve satisfaction loyalty confidence in the results.

[0065] 图1为本发明提供的用户满意度和忠诚度分析方法实施例一的流程示意图,如图1 所示,该方法包括: [0065] FIG customer satisfaction and loyalty analysis method of the present invention provides a schematic flow chart of Example 1, the method comprising:

[0066] S101、获取用户评论数据,该用户评论数据由用户根据使用结果生成。 [0066] S101, the data acquiring user reviews, user reviews the data generated by the user according to the results of use.

[0067] 具体地,可W通过多种方式来获取用户评论数据,例如网络、报刊、杂志、电视等媒体上收集用户对于某产品的评论数据。 [0067] Specifically, W can be obtained through the user reviews the data in various ways, such as collecting user data for a product review on the networks, newspapers, magazines, television and other media.

[0068] 用户对于某产品的评论数据并不是通过被动的回答而产生的,而是用户在购买、 使用过某产品后的真实体会,即,用户评论数据是一种"行为数据",而与现有技术中的"态度数据"完全不同,并且,运些评论数据都是由用户主动生成的。 [0068] For the user of a product review data not generated by passive answer, but users in the purchase, used after the real experience of a product, namely, user reviews the data is a "behavioral data", with the prior art "attitudinal data" is completely different, and transported some comment data is automatically generated by the user. 因此,获取运些评论数据并对其进行分析,所得到的结果能够反应用户对某产品的实际看法。 Therefore, access to transport some reviews and analyze the data, the results obtained can react the user actually views on a product.

[00例別02、根据用户评论数据,确定用户评论数据对应的评论维度类型W及评论结果类型。 [02 respectively 00 cases, the user reviews the data, determines the type of the dimension W and reviews the results of a review comment data corresponding to the user type.

[0070] 其中,用户评论数据对应的评论维度类型是指用户的某条评论是针对产品的哪个维度的,例如质量维度、服务维度等。 [0070] where, user reviews comments dimension corresponding to the data type refers to the user that a particular review is for a product which dimension, such as dimensions of quality, service and other dimensions. 评论维度类型可W根据产品的实际需要进行具体设置,并且,评论维度也可W不限于一级,可W根据实际需要设置为多级维度,即可W对每种评论维度进行细分,将其划分为多种评论子维度。 Reviews dimension W may be based on the actual type of product requires specific settings, and may also review the dimension W is not limited to one, according to actual needs W may be provided in multiple stages dimension W can be broken down for each dimension reviews the divided into several sub-dimensions comment.

[0071] 用户评论数据对应的评论结果类型是指用户的某条评论的分析结果属于哪种范畴,根据产品的实际需要,评论结果类型可W进行不同的划分。 [0071] The comment data corresponding to the user type refers to review analysis results of a review of which belongs to the category of users, according to the actual needs of the product, the results of a review of different types can be divided W. 例如,可W将评论结果类型划分为:非常满意、满意、较满意、一般、不太满意、不满意和很不满意。 For example, W will review the results of the type divided into: very satisfied, satisfied, more satisfied, in general, not very satisfied, dissatisfied and very dissatisfied. 或者,基于特定的满意度和忠诚度分析方法来设置评论结果类型,例如,如果要应用净推荐值(Net Promoter Score,简称NPS)方法来分析某产品的用户满意度和忠诚度,则评论结果类型为NPS中所规定的推荐型、消极型W及贬低型。 Or, to set the review results based on the type-specific satisfaction and loyalty analysis, for example, if you want to apply the Net Promoter Score (Net Promoter Score, referred to as NPS) methods to analyze customer satisfaction and loyalty of a product, the review results NPS types as specified in the recommended type, W-type negative and demeaning type.

[0072] S103、根据用户评论数据对应的评论维度类型W及评论结果类型,确定用户满意度和忠诚度结果。 [0072] S103, the comment data corresponding to the user reviews and comments dimension W type result type, determines user satisfaction and loyalty results.

[0073] 当获取到用户评论数据之后,通过对用户评论数据进行分析,可W确定其对应的评论维度类型W及评论结果类型。 [0073] When the user reviews the data obtained by data analysis of user comments, W may be determined corresponding dimension type W review Review result type. 基于所确定出的评论维度类型W及评论结果类型,就可W得到用户在不同维度上或者整体上对于某产品的评价结果,从而帮助企业识别出更有针对性的改进方向和改进点。 Based on the determined dimension type comments and reviews the results of the type W, W can get the user on a different dimension to the overall results of the evaluation of a product, or to help companies identify more focused direction for improvement and points of improvement.

[0074] 需要说明的是,本实施例中在获取用户评论数据时,可W同时获取针对多个产品的用户评论数据。 [0074] Incidentally, in the present embodiment, when the user reviews the data acquired, W can simultaneously acquire data for a plurality of user reviews of products. 当获取到多个产品的用户评论数据后,可W通过分类筛选,将各个产品的用户评论数据筛选出来,并针对每个产品的用户评论数据,采用步骤S102-S103的方法来确定用户满意度和忠诚度结果。 When a plurality of the acquired product data user comments, W can be classified by screening the product of the respective filter out data the user reviews and comments for each product the user data, the method steps S102-S103 to determine user satisfaction and loyalty results. 从而实现通过一次采集用户评论数据而获取多个产品的用户满意度和忠诚度结果的效果。 And in order to achieve customer satisfaction and obtain the loyalty of the results of the effect of multiple products through one data collection user reviews.

[0075] 本实施例中,通过获取用户主动发起的评论数据,并基于该评论数据确定其对应的评论维度类型W及评论结果类型,得到用户满意度和忠诚度结果。 [0075] In this embodiment, the comment is acquired by the user initiates the data, the comment data is determined based on its corresponding dimension types W and review comments result type, to obtain customer satisfaction and loyalty results. 由于用户主动发起的评论数据是用户经过实际使用后所生成的"行为数据",因此经过分析运些评论数据所得到的用户满意度和忠诚度结果能够反应用户对某产品的实际看法,从而保证用户满意度和忠诚度结果的可信度。 Since the user initiates review data after the user through the actual use of the generated "behavioral data" and therefore through the analysis of user satisfaction and loyalty results shipped some comment data obtained can react users actually think of a product, thus ensuring customer satisfaction and loyalty, confidence in the results. 进一步地,本实施例中通过对用户评论数据进行分析后确定出用户评论数据对应的评论维度类型W及评论结果类型,基于运些评论维度类型W及评论结果类型,可W得到用户在不同维度上或者整体上对于某产品的评价结果,从而帮助企业识别出更有针对性的改进方向和改进点。 Further, the present embodiment by determining the user comment data corresponding to the comment types of dimensions W and reviews the results of the type the user reviews the data analysis based on operation of these reviews types of dimensions W and reviews the results of type, W obtained in Example users in different dimensions or on the overall results of the evaluation of a product to help companies identify more focused direction for improvement and points of improvement.

[0076] 在上述实施例的基础上,本实施例设及获取用户评论数据的具体方法,即,上述步骤SlOl的一种具体实施方法为: [0076] On the basis of the above-described embodiment, the specific method of the present embodiment is provided and access to data user reviews embodiment, i.e., a particular embodiment of the above method steps SlOl:

[0077] 从至少一个网络渠道获取用户评论数据,其中,网络渠道至少包括:论坛、博客、电商平台。 [0077] obtain a user review data from at least one network channel, wherein the network channels, including at least: forum, blog, electronic business platform.

[0078] 随着网络技术的不断发展,越来越多的用户更倾向于通过网络渠道来发表评论, 通过网络发表评论的方式相比于其他方式要更加便捷和高效,并且能够获取到比其他方式更多的样本数据。 [0078] With the continuous development of network technology, more and more users prefer to comment through online channels, network comments by the way compared to other ways to be more convenient and efficient, and able to get to than others way more sample data. 例如,如果用户使用电商平台购买了一台电视,在购买、接收W及使用该电视的过程中,用户可能会有很多使用感受,用户可W将运些感受通过电商平台提供的评价页面来提交自己对于该电视的评价,运些评价反应的是用户的真实使用感受。 For example, if the user uses electronic business platform purchased a TV in the purchase and receipt of W and use the TV in the process, users may have to use a lot of experience, users can evaluate the page W will be shipped some feel provided by the electronic business platform to commit themselves to evaluate the TV, shipped some evaluation reaction is the user's real experience. 如果该电商平台的交易量很大,则通过电商平台所获取到用户评论数据就会很多,即能够保证获取到足够多的样本数据。 If the volume is large electronic business platform, acquired by the electronic business platform will be a lot of data to user comments, which can ensure to get enough sample data. 或者,用户也可W通过论坛、博客、微博、微信等各种网络渠道来发表自己对于某产品的评论数据。 Alternatively, users can express their comments to W data for a product through a variety of network channels forum, blog, microblogging, letters and so on.

[0079] 可选地,在通过网络渠道获取用户评论数据时,可W通过直接访问特定网络渠道的评价数据库来来获取用户评论数据。 [0079] Alternatively, the user comments in obtaining data through the network channels, can be used to obtain user comments W data by directly accessing a specific channel network evaluation database. 例如,对于某个特定产品,可W选定多个销售该产品的电商平台,通过定期访问运些电商平台的评价数据库,来获取该特定产品的用户评论数据。 For example, for a particular product, you can select multiple sales of the product W electronic business platform, transport evaluation database some electronic business platform through regular visits, to obtain the specific product data user comments.

[0080] 或者,也可W使用某个或某类产品对应的关键字,在各种网络渠道上捜索该产品对应的评论数据,并将捜索的评论数据集中保存起来,后续直接对所保存的评论数据进行分析即可。 [0080] Alternatively, W may be used or a type of product corresponding to the keyword, in a variety of cable network channels Dissatisfied comment data corresponding to the product, and the cable Dissatisfied comment dataset saved, stored directly subsequent comments can analyze the data.

[0081 ]本实施例中,通过从网络渠道获取用户评论数据,不仅能够保证运些评论数据能够反应用户的真实感受,同时,由于网络用户数量非常大,因此,还能够保证获取到足够数量的样本数据,运些足够数量的样本数据能够进一步保证用户满意度和忠诚度结果的可信度。 [0081] In this embodiment, by obtaining user reviews data from network sources, not only to guarantee the operation of these reviews data can reflect the real feel of users, but, since the number of network users is very large, therefore, it is possible to ensure obtaining a sufficient number of sample data, some transport a sufficient number of sample data to further ensure customer satisfaction and loyalty confidence in the results.

[0082] 在上述实施例的基础上,本实施例设及确定用户评论数据对应的评论维度类型W 及评论结果类型的具体方法,即,图2为本发明提供的用户满意度和忠诚度分析方法实施例二的流程示意图,如图2所示,上述步骤S102具体包括: [0082] Based on the foregoing embodiment, the present embodiment is provided user reviews and comments corresponding to the data type of the dimension W and the specific method of determining the type of comment results, i.e., FIG. 2 is a user satisfaction and loyalty to provide analysis of the present invention schematic flow chart according to a second embodiment of the method, as shown in FIG. 2, step S102 comprises:

[0083] S201、识别用户评论数据中的关键词。 [0083] S201, the user identification data review keywords.

[0084] 可选地,可W使用分词技术来识别出用户评论数据中的关键词,其中,分词技术的具体实现方法可W参考现有技术。 [0084] Alternatively, W can be used to identify the word segmentation data user reviews keyword, wherein the specific implementation technology word W can refer to the prior art method.

[0085] 需要说明的是,使用分词技术识别用户评论数据中的关键词时,可W识别出不同类型的关键词,主要包括评论主体关键词和情感类关键词。 [0085] Incidentally, when using segmentation techniques to identify keywords user reviews data, W can identify different types of keywords, and keyword including the keyword body emotional type comments. 例如,用户所发表的一条评论数据为"购买XX电视时销售人员的服务态度很好",则使用分词技术识别出的两类关键词为: 评论主体关键词为"销售"W及"服务态度",情感类关键词为"很好"。 For example, a user comment published data is "sales person of good service attitude when buying XX TV", the use of segmentation techniques identified two types of keywords are: Comments main keyword is "sales" W and "attitude "emotion class keyword is" very good. "

[0086] S202、根据用户评论数据中的关键词,确定用户评论数据的评论维度类型。 [0086] S202, the user reviews data keywords to determine the type of user reviews comments dimension data.

[0087] 如前所述,用户评论数据对应的评论维度类型是指用户的某条评论是针对产品的哪个维度的,例如质量维度、服务维度等。 [0087] As described above, the data corresponding to the user reviews comments dimension refers to the type of user for which a comment is the dimension of the product, e.g. dimensions of quality, service and other dimensions. 本步骤中,当获取到用户评论数据中的关键词之后,就可W根据关键词来确定评论维度类型。 In this step, when the user reviews the data acquired keyword, W can be determined based on keywords reviews dimension type.

[0088] 举例来说,假设评论维度类型包括一级,分别为:质量维度、服务维度、销售维度。 [0088] For example, if comments dimension types include one, namely: quality dimensions, the dimensions service, sales dimension. 用户所发表的一条评论数据为"购买XX电视时销售人员的服务态度很好",通过前述步骤所识别出的评论主体关键词为"销售"W及"服务态度",评论主体关键词即对应于用户评论数据的评论维度类型。 User comments published data is a "good time to buy XX TV sales service attitude", through the aforementioned steps identified Reviews main keyword is "sales" W and "attitude", a comment that is subject keyword correspondence user reviews dimension to the type of data. 其中,关键词"销售"与销售维度对应,因此,可W确定本条评论数据的评论维度类型为销售维度。 Among them, the keyword "sales" sales correspond to the dimensions, therefore, can determine Reviews dimension W type article reviews data for marketing dimension.

[0089] S203、对用户评论数据进行语义分析,获取用户评论数据对应的评论结果类型。 [0089] S203, the user reviews data semantic analysis, data corresponding to the acquired user reviews comments result type.

[0090] 具体地,在上述步骤S201中使用分词技术识别关键词时,可W识别出情感类关键词,本步骤中,即可W对已识别出的情感类关键词进行语义分析,确定用户在运条评论中的情感倾向。 When [0090] Specifically, in the above-described step S201 keyword word recognition technology, W may be identified emotion class keyword, in this step, W have been identified to emotion class keyword semantic analysis to determine user emotional tendencies in the operation reviews. 例如,前述的用户所发表的一条评论数据"购买XX电视时销售人员的服务态度很好"中的情感类关键词为"很好",则可W分析出该情感类关键词"很好"对应的语音应该是用户满意并可能会推荐的,进而可W确定出该条用户评论数据对应的评论结果类型。 For example, the aforementioned user published a review of data "XX TV sales staff when buying good service attitude" in the emotional type keyword is "very good", it can analyze the emotional type W keyword "very good" corresponding voice should be user satisfaction and may recommend, and thus can determine the article W user comments corresponding to the data type of review results.

[0091] 本实施例中,通过对用户评论数据进行识别关键词和语义分析处理,可W得到用户评论数据对应的评论主体和情感态度,进而,可W通过所确定出的评论主题和情感态度确定出用户评论数据对应的评论维度类型W及评论结果类型。 [0091] In this embodiment, the user reviews the data to identify keywords and semantic analysis process, may be obtained W data corresponding to user reviews and comments body attitudes, thereby, W can be determined by the attitudes and comments relating to comment data corresponding to the user is determined to reviews and comments dimension W type result type. 由于通过关键词识别和语义分析可W准确得到评论主体和情感态度,因此,可W保证基于其所确定出的评论维度类型W及评论结果类型的准确性。 Since keyword recognition and semantic analysis can accurately get Comments W body and emotional attitude, therefore, it can be determined based on W assurance reviews dimension type W and reviews the results of the type of accuracy.

[0092] 如前所述,评论维度类型可W根据产品的实际需要进行具体设置,并且,评论维度也可W不限于一级,可W根据实际需要设置为多级维度,即可W对每种评论维度进行细分, 将其划分为多种评论子维度。 [0092] As described above, review the dimension W type can be specifically set according to the actual needs of the product, and may also review the dimension W is not limited to one, according to actual needs W may be provided in multiple stages dimension to each W comments kinds of dimensions segments will be divided into several sub-dimensions comment. 在一种优选的实施方式中,可W将评论维度划分为两级,即, 评论维度类型包括多个评论父维度类型,每个评论父维度类型下包括多个评论子维度类型。 In a preferred embodiment, the comment dimension W may be divided into two, i.e., includes a plurality of review comments dimension type parent dimension type, comprising a plurality of sub-types of dimensions at each review comments dimension type parent. W下W智能电视环节分类为例来说明两级评论维度下的各级评论维度类型。 W W Smart TV link under the classification example to illustrate the type of comment at all levels in the two dimensions comments dimension.

[0093] 智能电视相关的环节可W总结为:研发设计、销售、质量W及服务,在每个环节下又包含多项子环节。 [0093] Smart TV related links can be summarized as W: R & D, sales, and service quality of W, at every link also contains a number of sub-sectors. 智能电视的环节和子环节可W分别对应评论父维度类型W及评论子维度类型。 Smart TV links and sub-links may correspond to review parent dimension W W types and sub-types of dimensions comment. 表1为智能电视的评论维度类型的划分。 Table 1 Smart TV comments dimension type of division.

[0094] 表1 [0094] TABLE 1

Figure CN106126499AD00101

[0095] [0095]

[0096] [0096]

[0097] 表1仅是智能电视的评论维度类型的一种示例性划分,根据实际需要,可W将智能电视或其他产品的评论维度类型进行更细致的划分,形成更加细化的评论维度类型层次结构。 [0097] Table 1 is only one type of smart TV dimension comment exemplary division, according to actual needs, W can be a smart television or other type of product reviews dimension of more detailed classification, more refined form dimension type Comments Hierarchy. 基于运些层次的评论维度类型,可W得出更加详细的用户满意度和忠诚度结果,从而为企业提供更加有针对性的提升建议。 Based on some hierarchy shipped comments dimension type, W can draw a more detailed user satisfaction and loyalty results, thus providing more targeted improvement suggestions.

[0098] 当评论维度类型为上述的两层结构,即包括评论父维度类型W及评论子维度类型时,上述步骤S202的一种优选实施方式为: [0098] When the above-described review of dimension type two-layer structure, i.e., including comments type parent dimension W sub-dimension type Reviews, step S202 described above as a preferred embodiment:

[0099] 判断用户评论数据中的第一关键词是否与第一评论父维度类型对应的关键词一致,若是,则: [0099] Analyzing the data in the user reviews a first keyword and the keyword coincides first comment corresponding to the type of parent dimension, and if so, then:

[0100] 确定用户评论数据的评论父维度类型为该第一评论父维度类型;W及,判断用户评论数据中的第二关键词是否与第一评论子维度类型对应的关键词一致,若是,则确定用户评论数据的评论子维度类型为该第一评论子维度类型,其中,第一评论子维度类型为第一评论父维度类型下所包括的评论子维度类型。 [0100] determined that the user reviews the data type for the first dimension of the parent comment Comments parent dimension type; and W is, determines that the user reviews the data in the second keyword coincides with a first dimension corresponding to the type of comment sub-keyword, if, user comments comment data is determined for the first type of sub-dimensions comment sub-dimension type, wherein the first type of comment sub-dimensions at a first type of sub-dimensions reviews comments included parent dimension types.

[0101] 其中,第一关键词和第二关键词既可W是同一关键词,也可W是不同的关键词。 [0101] wherein the first keyword and the second keyword W is either the same keywords, W may be different keywords.

[0102] 当第一关键词和第二关键词是不同的关键词时,首先将字符数量较少的关键词作为第一关键词,判断该第一关键词与第一评论父维度类型是否一致,具体可W直接通过比较字符的匹配程度来确定是否一致。 [0102] When the first keyword and the second keyword are different keywords, the first small number of characters as the first keyword of the keyword, determining whether the first keyword and the dimension of the first type are the same parent Comments specific W directly to determine compliance by comparing the degree of matching characters. 例如,第一关键词为"销售人员",而第一评论父维度类型为"销售",即第一关键词中包括了第一评论父维度类型,则可W确定第一关键词与第一评论父维度类型匹配。 For example, the first keyword is "sales", and the first comment parent dimension of type "sales", that is the first keyword is included in the first comment parent dimension type, you can determine the first keyword and first W comments parent dimension types match. 进而,判断第二关键词与第一评论子维度类型是否一致,也可W通过比较字符的匹配程度来确定是否一致。 Furthermore, keywords and determining a second dimension of the first sub-types are consistent review, W may be determined by the matching degree comparison is consistent character. 例如,第二关键词为"服务的态度",而第一评论子维度类型为"态度",则可W确定第二关键词与第一评论子维度类型匹配。 For example, the second keyword is "service attitude", and the first sub-dimension type comments as "attitude", W may be determined that the second keyword and the first sub-dimensions match the type of comment.

[0103] 当第一关键词和第二关键词是相同的关键词时,例如,一条用户评论数据为"XX电视屏幕质量不错",识别出的评论主体关键词只有一个,为"屏幕质量",则在确定用户评论数据的评论父维度类型时W及评论子维度类型时,都使用该关键词进行匹配,由于"屏幕质量"可W和"质量"运个评论父维度类型匹配上,因此,可W确定本条用户评论数据的评论父维度类型为"质量",同时,"屏幕质量"也可W和"屏幕质量"运个评论子维度类型匹配上,因此,可W确定本条用户评论数据的评论子维度类型为"屏幕质量"。 [0103] When the first keyword and the second keyword are the same keywords, for example, a user reviews the data of "XX TV screen good quality", identified only a comment subject keyword for "display quality" when, in determining when a parent commentary and reviews dimension type W sub-dimensions data type user comments, use the keyword matching, due to the "screen quality" to W and "quality" of the parent dimension shipped comments on the type of match, so can W determine comment parent dimension type article user comment data is "quality", while "screen quality" also W and "screen quality" Win the comments sub-dimension type matching, therefore, may be W is determined under this section user reviews data comments sub-dimensions of type "screen quality."

[0104] 其中,上述第一评论父维度类型是指任意一个评论父维度类型,第一评论子维度类型是一个特定评论父维度类型下的任意一个评论子维度类型。 [0104] wherein the first dimension of the parent comment refers to any type of a review type parent dimension, the first dimension comment sub-type is arbitrary comment at a specific type parent dimension a comment sub-dimension type. 即,在匹配时,可W按照评论维度类型的顺序,逐个将关键词与评论维度类型进行匹配,W确定关键词所对应的评论维度类型。 That is, when matching, the dimensions may be W type sequence of comments, comment one by one with the keyword matching dimension type, W type determining keyword comment corresponding dimension.

[0105] 如果用户评论数据中所识别出的关键词与所有的评论父维度类型和评论子维度类型都匹配不上,则将该条用户评论数据确定为综合评论维度类型,被确定为综合评论维度类型的用户评论数据,可W作为产品整体评价时的指标,但是不会作为某个维度评价时的指标。 [0105] If the user reviews the data in the identified keyword and all types of dimensions and review comments parent sub-types of dimensions is not a match, then the user reviews the data piece determined to reviews dimension integrated type, is determined as a comprehensive reviews user comments dimension type data, when the index W can evaluate the product as a whole, but not as an indicator of when a dimension evaluated.

[0106] 本实施例中,通过将用户评论数据中的关键词与评论维度类型进行匹配,可W准确确定出用户评论数据所属的评论维度类型,W保证最终所确定出的用户满意度和忠诚度的结果的准确性。 [0106] This ultimately determined user satisfaction and loyalty embodiments, by matching the user reviews and comments keyword data types dimension, W may be accurately determined dimensions comments user reviews the data type belongs to, to ensure that W the degree of accuracy of the results.

[0107] 当评论维度类型为上述的两层结构,即包括评论父维度类型W及评论子维度类型时,上述步骤S203的一种优选实施方式为: [0107] When the above-described review of dimension type two-layer structure, i.e., including comments type parent dimension W sub Review dimension type, Step S203 is a preferred embodiment as:

[0108] 对用户评论数据进行语义分析,判断语义分析结果是否与第一评论结果类型一致,若是,则确定用户评论数据对应的评论结果类型为第一评论结果类型。 [0108] The user reviews the data semantic analysis to determine the semantic analysis result is consistent with the result type of the first comments, and if yes, determining that the user reviews the data corresponding to a first type of review comments result of the result type.

[0109] 具体地,进行语义分析的过程取决于评论结果类型的划分,如前所述,评论结果类型可W进行不同的划分。 Process [0109] Specifically, the semantic analysis result depends on the type of comments divided, as described above, review the results of different types can be divided W. 例如,可W将评论结果类型划分为:非常满意、满意、较满意、一般、 不太满意、不满意和很不满意。 For example, W will review the results of the type divided into: very satisfied, satisfied, more satisfied, in general, not very satisfied, dissatisfied and very dissatisfied. 或者划分为NPS所规定的推荐型、消极型W及贬低型。 Or division specified for the NPS recommended type, W-type negative and demeaning type. 即,不同的评论结果类型划分方案会有不同数量的等级,每种等级对应不同的评价程度。 That is, the different types of comments the results of the program will be divided into a number of different levels, each level corresponds to a different level of evaluation. 当进行语义分析时,根据用户评论数据中所识别出的情感关键词,根据该情感关键词的含义,将其对应到确定的等级。 When the semantic analysis, user reviews the data in accordance with the identified keywords emotions, feelings according to the meaning of the keyword, which corresponds to the determined level. 例如,一条用户评论数据为"购买XX电视时销售人员的服务态度很好", 情感关键词为"很好",使用NPS所规定的立种级别作为评论结果类型,贝幡义分析时,就会将"很好"与"推荐型"对应,即得到该条用户评论数据对应的评论结果类型为"推荐型"。 For example, when a user reviews the data is "XX TV when buying a good sales staff attitude" emotional keywords as "very good", the use of stand species level under the NPS as a result of the type of comment, shellfish streamers meaning analysis, we will be "very good" corresponds to the "recommend type", that is to get the article data corresponding user reviews comments result type is "recommended type."

[0110] 本实施例中,通过将用户评论数据中的情感关键词进行语义分析,并与评论结果类型进行匹配,可W准确确定出用户评论数据所属的评论结果类型,W保证最终所确定出的用户满意度和忠诚度的结果的准确性。 [0110] finally determined by the embodiment of the present embodiment, semantic analysis data by the user reviews the emotional keywords, results and comments matching type, W may be accurately determined user reviews the data belongs review result type, to ensure W the results of user satisfaction and loyalty of accuracy.

[0111] 图3为本发明提供的用户满意度和忠诚度分析方法实施例=的流程示意图,如图3 所示,上述步骤S103的一种具体实施方式为: = Flow diagram illustrating an embodiment of [0111] 3 customer satisfaction and loyalty analysis method of the present invention provides, as shown in FIG. 3, Step S103 as a specific embodiment:

[0112] S301、获取每种评论维度类型的用户评论数据中每种评论结果类型的比例。 [0112] S301, acquires the proportion of each type of user reviews comments dimension data of each type of review results.

[0113] S302、根据每种评论维度类型的用户评论数据中每种评论结果类型的比例,确定每种评论维度类型对应的用户满意度和忠诚度结果。 [0113] S302, according to the proportion of each type of user reviews comments dimension data of each type of comment result, determines user satisfaction and loyalty results for each dimension corresponding to the type of comment.

[0114] 本实施例中,对于每种评论维度类型分别进行用户满意度和忠诚度计算,从而得出每种评论维度类型下的用户满意度和忠诚度结果。 [0114] In this embodiment, dimensions for each type of comments were customer satisfaction and loyalty calculation to arrive at the user satisfaction and loyalty results for each dimension type comments. 从而使得企业可W清楚得了解哪些方面做得比较好,哪些方面还需要改进,W明确改进的方向。 W so that enterprises can get a clear understanding of what has done well and what areas need improvement, W clear directions for improvement.

[0115] 图4为本发明提供的用户满意度和忠诚度分析方法实施例四的流程示意图,如图4 所示,上述步骤S103的另一种具体实施方式为: [0115] FIG customer satisfaction and loyalty analysis method of the present invention provides a schematic flow diagram according to a fourth embodiment, shown in Figure 4, another specific embodiment of the above-described step S103:

[0116] S401、获取所有用户评论数据中每种评论结果类型的比例。 [0116] S401, the data acquisition percentage of all user reviews of each type of review results.

[0117] S402、根据所有用户评论数据中每种评论结果类型的比例,确定整体的用户满意度和忠诚度结果。 [0117] S402, according to the proportion of all user reviews of each data type of the result of a review to determine user satisfaction and loyalty results overall.

[0118] 本实施例中,对于每种评论维度类型,分别进行用户满意度和忠诚度计算,从而得出每种评论维度类型下的用户满意度和忠诚度结果。 [0118] In this embodiment, dimensions for each type of comments, respectively, the user satisfaction and loyalty calculation to arrive at the user satisfaction and loyalty results for each dimension type comments. 从而使得企业可W清楚得了解哪些方面做得比较好,哪些方面还需要改进,明确改进的方向。 W so that enterprises can get a clear understanding of what has done well and what areas need improvement, clear directions for improvement.

[0119] 在上述两个实施例的基础上,还可W进一步分析用户满意度和忠诚度的趋势发展,具体地,可W按照预设周期进行上述两种满意度和忠诚度的分析计算,经过多次的分析计算之后,就可W得到一段时间内某产品在某个维度上的用户满意度和忠诚度的变化情况,从而进一步帮助企业衡量和分析在各维度上的待改进点。 [0119] Based on the above two embodiments, the further analysis may be W tendency of the user satisfaction and loyalty development, specifically, analysis and calculation of the two W satisfaction and loyalty according to a preset period, after analysis and calculation many times, you can get W changes in customer satisfaction and loyalty on one dimension of a product over a period of time, thereby further enables companies to measure and analyze the points to be improved in each dimension.

[0120] 图5为本发明提供的用户满意度和忠诚度分析装置实施例一的模块结构图,如图5 所示,该装置包括: User satisfaction and loyalty analysis of the present invention provides [0120] FIG. 5 is a block configuration diagram of apparatus embodiment, shown in Figure 5, the apparatus comprising:

[0121] 获取模块501,用于获取用户评论数据,该用户评论数据由用户根据使用结果生成。 [0121] obtaining module 501, configured to obtain user reviews the data, the user reviews the data generated by the user according to the results of use.

[0122] 第一确定模块502,用于根据用户评论数据,确定用户评论数据对应的评论维度类型W及评论结果类型。 [0122] a first determining module 502, according to a user comment data, comment data corresponding to the determined user types comments dimension W Review result type.

[0123] 第二确定模块503,用于根据用户评论数据对应的评论维度类型W及评论结果类型,确定用户满意度和忠诚度结果。 [0123] The second determination module 503, according to a user comment data corresponding to the type of dimension W and review comments result type, determines user satisfaction and loyalty results.

[0124] 该装置用于实现前述的方法实施例,其实现原理和技术效果类似,此处不再寶述。 [0124] The apparatus for the accomplishment of the foregoing method embodiments, principles and techniques which achieve a similar effect, not Po described herein.

[0125] 图6为本发明提供的用户满意度和忠诚度分析装置实施例二的模块结构图,如图6 所示,获取模块501包括: [0125] FIG customer satisfaction and loyalty to the present invention to provide analysis module configuration diagram of apparatus according to a second embodiment, shown in Figure 6, the acquisition module 501 comprises:

[0126] 第一获取单元5011,用于从至少一个网络渠道获取用户评论数据,该网络渠道至少包括:论坛、博客、电商平台。 [0126] The first obtaining unit 5011, configured to obtain user reviews data from at least one network channel, the network channels, including at least: forum, blog, electronic business platform.

[0127] 图7为本发明提供的用户满意度和忠诚度分析装置实施例=的模块结构图,如图7 所示,第一确定模块502包括: = Module configuration diagram of the embodiment of customer satisfaction and loyalty analysis of the present invention provides [0127] FIG. 7 apparatus embodiment, shown in Figure 7, the first determination module 502 comprises:

[01%]识别单元5021,用于识别用户评论数据中的关键词。 [01%] recognizing unit 5021, data for identifying a user reviews the keywords.

[0129] 确定单元5022,用于根据用户评论数据中的关键词,确定用户评论数据对应的评论维度类型。 [0129] determination unit 5022, according to user reviews keyword data, data corresponding to the determined user reviews comments dimension type.

[0130] 分析单元5023,用于对用户评论数据进行语义分析,获取用户评论数据对应的评论结果类型。 [0130] analysis unit 5023, a user reviews the data semantic analysis, data corresponding to the acquired user reviews comments result type.

[0131] 另一实施例中,用户评论数据的评论维度类型包括多个评论父维度类型,每个评论父维度类型下包括多个评论子维度类型。 [0131] In another embodiment, user reviews comments dimension type comprising a plurality of data types comments parent dimension, comprising a plurality of sub-types of dimensions at each review comments dimension type parent.

[0132] 相应地,确定单元5022具体用于: [0132] Accordingly, the determining unit 5022 is configured to:

[0133] 判断用户评论数据中的第一关键词是否与第一评论父维度类型对应的关键词一致,若是,则: [0133] Analyzing the data in the user reviews a first keyword and the keyword coincides first comment corresponding to the type of parent dimension, and if so, then:

[0134] 确定用户评论数据的评论父维度类型为第一评论父维度类型;W及,判断用户评论数据中的第二关键词是否与第一评论子维度类型对应的关键词一致,若是,则确定用户评论数据的评论子维度类型为第一评论子维度类型,其中,第一评论子维度类型为第一评论父维度类型下所包括的评论子维度类型。 [0134] determined that the user reviews the data type is a first dimension of the parent comment Comments type parent dimension; W and consistent, user reviews determines whether the keyword data in the second first dimension corresponding to the type of comment sub-keyword, if yes, determining a user comment sub-dimensions comment data type is a first type of comment sub-dimension, wherein the first type of comment sub-dimensions at a first type of sub-dimensions reviews comments included parent dimension types.

[01巧]分析单元5023具体用于: [Qiao 01] analysis unit 5023 is configured to:

[0136] 对用户评论数据进行语义分析,判断语义分析结果是否与第一评论结果类型一致,若是,则确定用户评论数据对应的评论结果类型为第一评论结果类型。 [0136] The user reviews the data semantic analysis to determine the semantic analysis result is consistent with the result type of the first comments, and if yes, determining that the user reviews the data corresponding to a first type of review comments result of the result type.

[0137] 图8为本发明提供的用户满意度和忠诚度分析装置实施例四的模块结构图,如图8 所示,第二确定模块503包括: Block configuration diagram according to a fourth embodiment of the apparatus of customer satisfaction and loyalty analysis provided [0137] FIG. 8 of the present invention, shown in Figure 8, the second determination module 503 comprises:

[0138] 第一获取单元5031,用于获取每种评论维度类型的用户评论数据中每种评论结果类型的比例。 [0138] a first obtaining unit 5031, configured to obtain the proportion of each type of user reviews comments dimension data of each type of review results.

[0139] 第一确定单元5032,用于根据每种评论维度类型的用户评论数据中每种评论结果类型的比例,确定每种评论维度类型对应的用户满意度和忠诚度结果。 [0139] The first determining unit 5032, according to the proportion of each type of user reviews comments dimension data of each type of comment result, determines user satisfaction and loyalty results for each dimension corresponding to the type of comment.

[0140] 图9为本发明提供的用户满意度和忠诚度分析装置实施例五的模块结构图,如图9 所示,第二确定模块503还包括: [0140] FIG 9 customer satisfaction and loyalty analysis module of the present invention provides a configuration diagram of the apparatus according to the fifth embodiment, shown in Figure 9, the second determination module 503 further comprises:

[0141] 第二获取单元5033,用于获取所有用户评论数据中每种评论结果类型的比例。 [0141] The second obtaining unit 5033, configured to obtain the proportions of all user reviews the data type of each review the results.

[0142] 第二确定单元5034,用于根据所有用户评论数据中每种评论结果类型的比例,确定整体的用户满意度和忠诚度结果。 [0142] The second determining unit 5034, the proportions of all according to the data of each user reviews comments result type, determines the overall user satisfaction and loyalty results.

[0143] 另一实施例中,上述第一评论结果类型为推荐型评论、消极型评论、贬低型评论中的任意一种。 [0143] embodiment, the first type is a result of review comments recommended type, negative type comments demeaning any one type to another embodiment review.

[0144] 本领域普通技术人员可W理解:实现上述各方法实施例的全部或部分步骤可W通过程序指令相关的硬件来完成。 [0144] Those of ordinary skill in the art may be appreciated W: the foregoing method embodiments all or part of the steps may be W by a program instructing relevant hardware to complete. 前述的程序可W存储于一计算机可读取存储介质中。 W aforementioned program may be stored in a computer readable storage medium. 该程序在执行时,执行包括上述各方法实施例的步骤;而前述的存储介质包括:R〇M、RAM、磁碟或者光盘等各种可W存储程序代码的介质。 When the program is executed, comprising the step of performing the above-described method of the embodiment; and the storage medium comprising: a variety of medium storing program code R〇M W, RAM, magnetic disk, or optical disk.

[0145] 最后应说明的是:W上各实施例仅用W说明本发明的技术方案,而非对其限制;尽管参照前述各实施例对本发明进行了详细的说明,本领域的普通技术人员应当理解:其依然可W对前述各实施例所记载的技术方案进行修改,或者对其中部分或者全部技术特征进行等同替换;而运些修改或者替换,并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。 [0145] Finally, it should be noted that: each of the embodiments the W Example W only illustrate the technical solutions of the present invention, rather than limiting;. Although the embodiments of the present invention has been described in detail, those skilled in the art it should be understood that: W which can still make modifications to the technical solutions described in the embodiments, or to some or all of the technical features make equivalent replacements; the transport of these modifications or replacements do not cause the essence of corresponding technical solutions to depart from the present invention scope of the technical solutions of the embodiments.

Claims (10)

1. 一种用户满意度和忠诚度分析方法,其特征在于,包括: 获取用户评论数据,所述用户评论数据由用户根据使用结果生成; 根据所述用户评论数据,确定所述用户评论数据对应的评论维度类型以及评论结果类型; 根据所述用户评论数据对应的评论维度类型以及评论结果类型,确定用户满意度和忠诚度结果。 A customer satisfaction and loyalty analysis method comprising: acquiring data user reviews, user reviews the data generated by the user using the result; comment data according to the user, determines that the data corresponding to the user reviews comments comments result type and dimension type; according to the comment data corresponding to the user reviews and comments dimension type result type, determines user satisfaction and loyalty results.
2. 根据权利要求1所述的方法,其特征在于,所述获取用户评论数据,包括: 从至少一个网络渠道获取所述用户评论数据,所述网络渠道至少包括:论坛、博客、电商平台。 2. The method according to claim 1, wherein the obtaining user reviews data, comprising: obtaining user reviews the data, at least from the network channel comprises at least one network channel: forum, blog, business platform .
3. 根据权利要求1所述的方法,其特征在于,所述根据所述用户评论数据,确定所述用户评论数据对应的评论维度类型以及评论结果类型,包括: 识别所述用户评论数据中的关键词; 根据所述用户评论数据中的关键词,确定所述用户评论数据对应的评论维度类型; 对所述用户评论数据进行语义分析,获取所述用户评论数据对应的评论结果类型。 3. The method according to claim 1, wherein said data based on the user comments, comment data corresponding to the user is determined to reviews and comments dimension type result types, comprising: identifying the user reviews data Key words; comment data in accordance with said user keyword, determining whether the data corresponding to the user reviews comments dimension type; semantic analysis of the user reviews data, obtaining data corresponding to the user reviews comments result type.
4. 根据权利要求3所述的方法,其特征在于,所述用户评论数据的评论维度类型包括多个评论父维度类型,每个评论父维度类型下包括多个评论子维度类型;所述根据所述用户评论数据中的关键词,确定所述用户评论数据对应的评论维度类型,包括: 判断所述用户评论数据中的第一关键词是否与第一评论父维度类型对应的关键词一致,若是,则: 确定所述用户评论数据的评论父维度类型为所述第一评论父维度类型;以及, 判断所述用户评论数据中的第二关键词是否与第一评论子维度类型对应的关键词一致,若是,则确定所述用户评论数据的评论子维度类型为所述第一评论子维度类型,其中, 所述第一评论子维度类型为所述第一评论父维度类型下所包括的评论子维度类型。 4. The method according to claim 3, wherein said user reviews comments dimension type comprising a plurality of data types comments parent dimension, comprising a plurality of sub-types of dimensions at each comment Comments parent dimension type; according to the the user reviews the data in the keyword determining the data corresponding to the user reviews comments dimension type, comprising: determining whether the user reviews data in the first keywords are keywords coincide with the first dimension corresponding to the type of the parent comment, if so, then: determining a type of the dimension of the parent comment data for the user reviews a first dimension of the parent comment type; and, determining whether the user reviews the data in the first keyword and the second dimension corresponding to the type of comment sub-critical same words, if the user reviews the data type is determined to reviews of the first sub-dimensions comment sub-dimension type, wherein the first type of comment sub dimension in said first dimension of the type comprising a review of the parent comments sub-dimension type.
5. 根据权利要求3所述的方法,其特征在于,所述对所述用户评论数据进行语义分析, 获取所述用户评论数据对应的评论结果类型,包括: 对所述用户评论数据进行语义分析,判断语义分析结果是否与第一评论结果类型一致,若是,则确定所述用户评论数据对应的评论结果类型为所述第一评论结果类型。 5. The method according to claim 3, wherein said semantic analysis of the user reviews data, obtaining data corresponding to the user reviews a review result types, comprising: a semantic analysis of the data user reviews , it is determined whether the semantic analysis result is consistent with the result type of the first comment, if the user reviews the data corresponding to the result of determining the type of comment to review the results of the first type.
6. 根据权利要求1所述的方法,其特征在于,所述根据用户评论数据对应的评论维度类型以及评论结果类型,确定用户满意度和忠诚度结果,包括: 获取每种评论维度类型的用户评论数据中每种评论结果类型的比例; 根据每种评论维度类型的用户评论数据中每种评论结果类型的比例,确定每种评论维度类型对应的用户满意度和忠诚度结果。 6. The method according to claim 1, wherein the comment data corresponding to the user reviews and comments dimension type result type, determines user satisfaction and loyalty results, comprising: obtaining for each type of user comments dimension the proportion of each comment data type of the result comments; comments according to the proportion of each type of dimension data of each user reviews comments result type, determines user satisfaction and loyalty results for each dimension corresponding to the type of comment.
7. 根据权利要求1所述的方法,其特征在于,所述根据用户评论数据对应的评论维度类型以及评论结果类型,确定用户满意度和忠诚度结果,包括: 获取所有用户评论数据中每种评论结果类型的比例; 根据所有用户评论数据中每种评论结果类型的比例,确定整体的用户满意度和忠诚度结果。 7. The method according to claim 1, wherein the comment data corresponding to the user reviews and comments dimension type result type, determines user satisfaction and loyalty results, comprising: acquiring all the data of each user reviews comments proportion of the type of results; according to the proportion of all user reviews of each data type of the result of a review to determine user satisfaction and loyalty results overall.
8. 根据权利要求5所述的方法,其特征在于,所述第一评论结果类型为推荐型评论、消极型评论、贬低型评论中的任意一种。 8. The method as claimed in claim 5, wherein any one of the first type is a result of review comments recommended type, negative type comments, degrading type of comment.
9. 一种用户满意度和忠诚度分析装置,其特征在于,包括: 获取模块,用于获取用户评论数据,所述用户评论数据由用户根据使用结果生成; 第一确定模块,用于根据所述用户评论数据,确定所述用户评论数据对应的评论维度类型以及评论结果类型; 第二确定模块,用于根据所述用户评论数据对应的评论维度类型以及评论结果类型, 确定用户满意度和忠诚度结果。 A customer satisfaction and loyalty analysis apparatus, characterized by comprising: an obtaining module, configured to obtain user reviews the data, the user reviews the data generated by the user using the result; a first determining module, according to the said user reviews the data, determines the type of user reviews the comment data corresponding to the dimension and the comments result type; a second determining module configured according to the user data corresponding to the comment and review comments dimension type result type, determines user satisfaction and loyalty of the results.
10. 根据权利要求9所述的装置,其特征在于,所述获取模块包括: 第一获取单元,用于从至少一个网络渠道获取所述用户评论数据,所述网络渠道至少包括:论坛、博客、电商平台。 10. The apparatus according to claim 9, wherein the obtaining module comprises: a first acquisition unit configured to acquire from the at least one network user reviews the data channel, the channel network comprises at least: forum, blog ,Electronic business platform.
CN201610462852.1A 2016-06-22 2016-06-22 User satisfaction and loyalty degree analysis method and device CN106126499A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610462852.1A CN106126499A (en) 2016-06-22 2016-06-22 User satisfaction and loyalty degree analysis method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610462852.1A CN106126499A (en) 2016-06-22 2016-06-22 User satisfaction and loyalty degree analysis method and device

Publications (1)

Publication Number Publication Date
CN106126499A true CN106126499A (en) 2016-11-16

Family

ID=57269039

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610462852.1A CN106126499A (en) 2016-06-22 2016-06-22 User satisfaction and loyalty degree analysis method and device

Country Status (1)

Country Link
CN (1) CN106126499A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070294127A1 (en) * 2004-08-05 2007-12-20 Viewscore Ltd System and method for ranking and recommending products or services by parsing natural-language text and converting it into numerical scores
CN102214201A (en) * 2010-04-08 2011-10-12 微软公司 Deriving statement from product or service reviews
CN103078956A (en) * 2013-02-01 2013-05-01 杭州蓝谷科技有限公司 Information interaction system for realizing accurate data mining
CN103488635A (en) * 2012-06-11 2014-01-01 腾讯科技(深圳)有限公司 Method and device for acquiring product information
CN103631861A (en) * 2013-10-28 2014-03-12 百度在线网络技术(北京)有限公司 Method and device used for processing and providing evaluation information
CN105095411A (en) * 2015-07-09 2015-11-25 中山大学 Method and system for predicting APP ranking based on App quality

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070294127A1 (en) * 2004-08-05 2007-12-20 Viewscore Ltd System and method for ranking and recommending products or services by parsing natural-language text and converting it into numerical scores
CN102214201A (en) * 2010-04-08 2011-10-12 微软公司 Deriving statement from product or service reviews
CN103488635A (en) * 2012-06-11 2014-01-01 腾讯科技(深圳)有限公司 Method and device for acquiring product information
CN103078956A (en) * 2013-02-01 2013-05-01 杭州蓝谷科技有限公司 Information interaction system for realizing accurate data mining
CN103631861A (en) * 2013-10-28 2014-03-12 百度在线网络技术(北京)有限公司 Method and device used for processing and providing evaluation information
CN105095411A (en) * 2015-07-09 2015-11-25 中山大学 Method and system for predicting APP ranking based on App quality

Similar Documents

Publication Publication Date Title
Mercer The analysis of classroom talk: Methods and methodologies
Ludwig et al. More than words: The influence of affective content and linguistic style matches in online reviews on conversion rates
US9536269B2 (en) Method and apparatus for analyzing and applying data related to customer interactions with social media
Mohammad et al. Stance and sentiment in tweets
He et al. A novel social media competitive analytics framework with sentiment benchmarks
US20160162464A1 (en) Techniques for combining human and machine learning in natural language processing
Lynch et al. User participation in decision support systems development: influencing system outcomes
KR100919912B1 (en) Systems and methods for semantic knowledge assessment, instruction, and acquisition
MacPhail et al. Process guidelines for establishing Intercoder Reliability in qualitative studies
US10346542B2 (en) Human-to-human conversation analysis
EP2524348A2 (en) User communication analysis systems and methods
US9268766B2 (en) Phrase-based data classification system
US20120324353A1 (en) System and Method for Building and Managing User Experience for Computer Software Interfaces
Krishnamoorthy Linguistic features for review helpfulness prediction
US9646079B2 (en) Method and apparatus for identifiying similar questions in a consultation system
Angus et al. Visualising conversation structure across time: Insights into effective doctor-patient consultations
Krook Electoral gender quotas: A conceptual analysis
Feeney Organizational red tape: A measurement experiment
US20150179082A1 (en) Dynamic Identification and Validation of Test Questions from a Corpus
Tourani et al. Monitoring sentiment in open source mailing lists: exploratory study on the apache ecosystem
Figl Comprehension of procedural visual business process models
US20130290207A1 (en) Method, apparatus and computer program product to generate psychological, emotional, and personality information for electronic job recruiting
Carvalho et al. Liars and saviors in a sentiment annotated corpus of comments to political debates
KR20090106556A (en) Method and system for assessing and refining the quality of web services definitions
US9348900B2 (en) Generating an answer from multiple pipelines using clustering

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination