CN117649255A

CN117649255A - Product pain point identification method and device, electronic equipment and storage medium

Info

Publication number: CN117649255A
Application number: CN202410109913.0A
Authority: CN
Inventors: 王小红; 林瑶英; 胡晓君; 崔嶷
Original assignee: Meiyun Zhishu Technology Co ltd
Current assignee: Meiyun Zhishu Technology Co ltd
Priority date: 2024-01-26
Filing date: 2024-01-26
Publication date: 2024-03-05

Abstract

The application relates to the technical field of artificial intelligence and provides a method and a device for identifying pain points of products, electronic equipment and a storage medium. The method comprises the following steps: based on the pain point label library, performing pain point matching on the comment text of the target product with negative emotion in a preset time period to obtain a pain point matching result of the target product; classifying the pain points according to the pain point matching result based on the universality index of each pain point and the urgency index of each pain point to obtain a target product pain point identification result; the universality index of each pain point is determined based on the ratio of the number of each pain point to the number of all pain points in the pain point matching result; the urgency index of each pain point is determined based on the rate of increase of the number of each pain point in the pain point matching result compared to the corresponding number of pain points in the previous preset time period. The method, the device, the electronic equipment and the storage medium provided by the embodiment of the application can improve the efficiency and the accuracy of product pain point identification.

Description

Product pain point identification method and device, electronic equipment and storage medium

Technical Field

The application relates to the technical field of artificial intelligence, in particular to a method and a device for identifying pain points of a product, electronic equipment and a storage medium.

Background

Pain points are places of difficulty, problem or dissatisfaction that consumers or users encounter in using the product. These pain points may include aspects of product quality, price, functionality, user experience, after-market services, and the like. Knowing the pain points of the consumer can help the enterprise to better meet the demands of the consumer and improve the competitiveness of the product.

The prior art is limited to qualitative and quantitative investigation of small samples, and consumer feedback can be obtained more timely than the past from various electronic commerce platforms along with the continuous development of internet big data technology. The feedback corpus of the consumers has the characteristics of long text, context dependence and informal word usage. At present, the effect of identifying the pain points of the products in the feedback corpus of the consumers is poor, and the competitive power of the enterprise products is not improved.

Disclosure of Invention

The application aims to solve the technical problem that the effect of the pain point identification of the current product is poor. Therefore, the application provides a method for identifying the pain spots of the product, which can improve the accuracy of identifying the pain spots of the product.

The application also provides a product pain point identification device.

The application also proposes an electronic device and a non-transitory computer readable storage medium.

According to an embodiment of the first aspect of the present application, a method for identifying pain points of a product includes:

based on the pain point label library, performing pain point matching on the comment text of the target product with negative emotion in a preset time period to obtain a pain point matching result of the target product; the pain point tag library comprises an industry general tag library, an industry characteristic tag library and an inapplicable tag library; the industry universal tag library is a tag library determined based on universal pain point words of the industry to which the target product belongs; the industry characteristic tag library is a tag library determined based on characteristic pain point words of the industry to which the target product belongs; the inapplicable tag library is a tag library determined based on inapplicable pain spot words of target products in general pain spot words of the industry general tag library;

classifying the pain points according to the pain point matching result based on the universality index of each pain point and the urgency index of each pain point to obtain a target product pain point identification result; the universality index of each pain point is determined based on the ratio of the number of each pain point to the number of all pain points in the pain point matching result; the urgency index of each pain point is determined based on the rate of increase of the number of each pain point in the pain point matching result compared to the corresponding number of pain points in the previous preset time period.

According to the product pain point identification method provided by the embodiment of the application, the pain point is matched with the comment text of the target product with negative emotion through the pain point label library, and then pain point classification is carried out based on the universality index and the urgency index, so that the target product pain point identification result is obtained, the accuracy of product pain point identification can be improved, the readability of the pain point is improved, and the high cost problem of manual interpretation and correction is avoided. In addition, the pain points are classified by introducing universality and urgency, the value of the pain points is quantized, new people, new scenes and the like can be found, the opportunity points of new products are expanded, the decision-making efficiency of pain point processing is improved, the pain points of enterprises are obviously guided to optimize resource allocation, and the product competitiveness of the enterprises is improved.

According to an embodiment of the present application, the classifying the pain point matching result based on the universality index of each pain point and the urgency index of each pain point to obtain a target product pain point identification result includes:

determining the average value of the universality index of all pain points in the pain point matching result based on the universality index of each pain point; determining the average value of the urgency indexes of all pain points in the pain point matching result based on the urgency indexes of all pain points;

Determining a secondary pain point grading map based on the average value of the universality index of all the pain points and the average value of the urgency index of all the pain points;

classifying the pain points according to the pain point matching result based on the universality index of each pain point and the position of the urgency index of each pain point in the secondary pain point classification map to obtain a target product pain point identification result; the pain point identification result comprises core pain points, public pain points, emergency pain points and edge pain points; the core pain points represent the pain points of which the universality index of the pain points is larger than or equal to the average value of the universality indexes of all the pain points and the urgency index of the pain points is larger than the average value of the urgency indexes of all the pain points; the general pain points represent the pain points of which the universality index of the pain points is larger than or equal to the average value of the universality indexes of all the pain points and the urgency index of the pain points is smaller than or equal to the average value of the urgency indexes of all the pain points; the emergency pain points represent pain points with a universality index of the pain points smaller than the average value of the universality indexes of all the pain points and with an urgency index of the pain points larger than the average value of the urgency indexes of all the pain points; the marginal pain points represent pain points with a prevalence index of the pain points less than the average of the prevalence indexes of all the pain points, and with an urgency index of the pain points less than or equal to the average of the urgency indexes of all the pain points.

According to the product pain point identification method provided by the embodiment of the application, the secondary pain point classification map is determined through the average value of the universality index and the average value of the urgency index, so that pain points are classified, core pain points, public pain points, emergency pain points and edge pain points are obtained, new people, new scenes and the like can be found by quantifying the pain point value, new product opportunity points are expanded, the pain point processing decision efficiency is further improved, the enterprise pain point optimizing resource allocation is obviously guided, and the enterprise product competitiveness is further improved.

According to an embodiment of the application, the pain point matching is performed on a comment text of a target product with negative emotion in a preset time period based on a pain point tag library, so as to obtain a pain point matching result of the target product, and the method comprises the following steps:

carrying out pain point matching on the target product comment text with negative emotion in a preset time period based on an industry universal tag library and an industry characteristic tag library in the pain point tag library to obtain an initial target product pain point matching result;

and filtering the inapplicable pain point words in the initial target product pain point matching result based on the inapplicable label library in the pain point label library to obtain the target product pain point matching result.

According to an embodiment of the application, the pain point matching of the target product comment text with negative emotion in a preset time period based on the industry universal tag library and the industry characteristic tag library in the pain point tag library comprises the following steps:

generating a first vector based on product pain point words corresponding to an industry universal label library and an industry characteristic label library in the pain point label library; generating a second vector based on target keywords of target product comment texts with negative emotions in a preset time period; the target keywords represent feedback to a target product;

and carrying out regular matching on the first vector and the second vector.

According to one embodiment of the present application, the method further comprises:

and determining that the similarity between the target keyword and any product pain point word is greater than a similarity threshold value, and mapping the target keyword into the corresponding product pain point word.

determining that the ratio of the total number of pain points in the pain point matching result of the target product to the number of target product comment texts with negative emotions in a preset time period is smaller than a preset ratio, and updating the product pain point words corresponding to the industry universal tag library and the industry characteristic tag library in the pain point tag library based on the target keywords in the target product comment texts with negative emotions in the preset time period.

and carrying out negative emotion word matching on the comment text of the target product in a preset time period based on the negative emotion words of the industry to which the target product belongs, and determining the comment text of the target product with the negative emotion.

According to a second aspect of the present application, a product pain point identification device includes:

the matching module is used for carrying out pain point matching on the comment text of the target product with negative emotion in a preset time period based on the pain point tag library to obtain a pain point matching result of the target product; the pain point tag library comprises an industry general tag library, an industry characteristic tag library and an inapplicable tag library; the industry universal tag library is a tag library determined based on universal pain point words of the industry to which the target product belongs; the industry characteristic tag library is a tag library determined based on characteristic pain point words of the industry to which the target product belongs; the inapplicable tag library is a tag library determined based on inapplicable pain spot words of target products in general pain spot words of the industry general tag library;

the grading module is used for grading the pain points according to the pain point matching result based on the universality index of each pain point and the urgency index of each pain point to obtain a target product pain point identification result; the universality index of each pain point is determined based on the ratio of the number of each pain point to the number of all pain points in the pain point matching result; the urgency index of each pain point is determined based on the rate of increase of the number of each pain point in the pain point matching result compared to the corresponding number of pain points in the previous preset time period.

According to the product pain point identification device provided by the embodiment of the application, the pain point is matched with the comment text of the target product with negative emotion through the pain point label library, and then the pain points are classified based on the universality index and the urgency index, so that the target product pain point identification result is obtained, the accuracy of product pain point identification can be improved, the readability of the pain points is improved, and the high cost problem of manual interpretation and correction is avoided. In addition, the pain points are classified by introducing universality and urgency, the value of the pain points is quantized, new people, new scenes and the like can be found, the opportunity points of new products are expanded, the decision-making efficiency of pain point processing is improved, the pain points of enterprises are obviously guided to optimize resource allocation, and the product competitiveness of the enterprises is improved.

An electronic device according to an embodiment of the third aspect of the present application comprises a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing a product pain point identification method as in the embodiment of the first aspect when executing the computer program.

A non-transitory computer readable storage medium according to an embodiment of the fourth aspect of the present application has stored thereon a computer program which, when executed by a processor, implements a product pain point identification method as in the embodiment of the first aspect.

The above technical solutions in the embodiments of the present application have at least one of the following technical effects:

Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for a person having ordinary skill in the art.

Fig. 1 is a schematic flow chart of a method for identifying pain points of a product according to an embodiment of the present application;

fig. 2 is a schematic diagram of different product requirements under different scenes of different crowds according to an embodiment of the present application;

FIG. 3 is a schematic diagram of a hierarchical map of secondary pain points according to an embodiment of the present application;

FIG. 4 is a schematic diagram of a pain point identification and value quantification model provided in an embodiment of the present application;

fig. 5 is a schematic flow chart of the product pain point identification method provided in the embodiment of the present application in practical application;

fig. 6 is a schematic structural diagram of a product pain point recognition device according to an embodiment of the present application;

fig. 7 is a schematic entity structure of an electronic device according to an embodiment of the present application.

Detailed Description

The feedback of the consumer to the product can be obtained more timely than in the past from various e-commerce platforms, for example: on a certain e-commerce platform, the feedback of the user on the used toothpaste is as follows: the toothpaste is good, the hairs on the brushed teeth are clean, the toothpaste without stinging feeling is good to use, the toothpaste is very popular to a mother, the toothpaste is also good to use once, the teeth are felt to be whitened, and the teeth are clean. "the user feedback corpus has contextual dependency on toothpaste, and represents a generalization. On another e-commerce platform, the feedback of the user on the used toothpaste is as follows: "toothpaste 1: the oral cavity is sensitive, and can be used for refreshing breath and protecting gingiva for comfort; toothpaste 2: baking soda toothpaste, mainly beating yellow, removing stains and deeply cleaning tartar; toothpaste 3: jasmine flower fragrance, freshness and fragrance retention, and tooth stain and tartar improvement; toothpaste 4: the tea is suitable for people who drink tea and coffee frequently to improve dental stains. "the user's feedback corpus for toothpaste is long text and there is a contextual dependency.

At present, when pain points in the feedback corpus of a product are extracted, a natural language processing technology is mainly utilized, tag keyword matching is mainly utilized, and emotion tendencies are used for assisting in judging and screening the pain points. However, the overall recognition effect is poor, and the following defects exist in detail:

(1) The pain points extracted are too general and even wrong, and still consume a great deal of manpower to read and correct;

(2) The number of pain points is large, but the value cannot be quantified, and the optimization decision of the pain points cannot be directly guided;

(3) The pain point optimization decision is highly dependent on manual judgment, and experience cannot be converted into data precipitation.

For example: at a certain e-commerce platform, the feedback corpus of a user on a certain charger baby is as follows: the things are still in touch at present, the charging communication is slow, the logistics is very slow, the goods are received after more than one week, the goods are not useful at all after being required to be paid, and the meaning of the existence of the goods is not understood. "the user is dissatisfied with the charging speed in the corpus feedback, but describes the generalization; the pain point in the aspect of logistics is clearly described, and the pain point is particularly slow in logistics and is shown to be received for more than one week. The technology at present can only analyze two pain points of slow charge and slow logistics, and the detailed description of the pain points needs manual tracing and interpretation for the pain. Based on the method, the method can more accurately and specifically identify the pain points in the feedback of real consumers on social and electronic commerce platforms, and quantify and measure the value of the pain points, and the method can ensure that the identified pain points are clearer, so that the cost of manual interpretation and correction is reduced; the pain point grading optimization can be guided in an auxiliary manner, so that the decision making efficiency is improved.

Embodiments of the present application are described in further detail below with reference to the accompanying drawings and examples. The following examples are illustrative of the present application but are not intended to limit the scope of the present application.

In the description of the embodiments of the present application, it should be noted that, directions or positional relationships indicated by terms such as "center", "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., are based on those shown in the drawings, are merely for convenience in describing the embodiments of the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the embodiments of the present application. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.

In the description of the embodiments of the present application, it should be noted that, unless explicitly specified and limited otherwise, the terms "connected," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium. The specific meaning of the terms in the embodiments of the present application will be understood by those of ordinary skill in the art in a specific context.

In the examples herein, a first feature "on" or "under" a second feature may be either the first and second features in direct contact, or the first and second features in indirect contact via an intermediary, unless expressly stated and defined otherwise. Moreover, a first feature being "above," "over" and "on" a second feature may be a first feature being directly above or obliquely above the second feature, or simply indicating that the first feature is level higher than the second feature. The first feature being "under", "below" and "beneath" the second feature may be the first feature being directly under or obliquely below the second feature, or simply indicating that the first feature is less level than the second feature.

In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the embodiments of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

Fig. 1 is a schematic flow chart of a method for identifying pain points of a product according to an embodiment of the present application. Referring to fig. 1, a method for identifying pain points of a product according to an embodiment of the present application may include:

step 110, carrying out pain point matching on a target product comment text with negative emotion in a preset time period based on a pain point tag library to obtain a target product pain point matching result; the pain point label library comprises an industry general label library, an industry characteristic label library and an inapplicable label library; the industry universal tag library is a tag library determined based on universal pain point words of the industry to which the target product belongs; the industry characteristic tag library is a tag library determined based on characteristic pain point words of the industry to which the target product belongs; the inapplicable tag library is a tag library determined based on inapplicable pain point words of target products in general pain point words of the industry general tag library;

step 120, classifying the pain points according to the pain point matching result based on the universality index of each pain point and the urgency index of each pain point to obtain a target product pain point identification result; the universality index of each pain point is determined based on the ratio of the number of each pain point to the number of all pain points in the pain point matching result; the urgency index of each pain point is determined based on the rate of increase of the number of each pain point in the pain point matching result compared to the corresponding number of pain points in the previous preset time period.

It should be noted that, in the product pain point identification method provided in the embodiment of the present application, the execution body may be a computer device, for example, a mobile phone, a tablet computer, a notebook computer, a palm computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (ultra-mobile personal computer, UMPC), an internet book, a personal digital assistant (personal digital assistant, PDA), or the like. The pain point recognition system for executing the product pain point recognition method is taken as an execution subject to be taken as an example to describe the product pain point recognition method provided by the embodiment of the application.

In step 110, the pain point recognition system may obtain the feedback corpus of the user on the target product in the preset period of time as the comment text of the target product. The target product comment text may contain positive emotion comment text, negative emotion comment text, etc., while comment text with positive emotion generally does not include a product pain point. The pain point recognition system can screen target product comment texts with negative emotions in the target product comment texts, for example, the screening of the antipodal emotion comment texts can be realized based on a mode of constructing an emotion dictionary or a mode based on machine learning, such as naive Bayes, maximum entropy and the like, and the negative emotion comment texts can contain target keywords for representing feedback to the target products, such as few, gift and side effect.

The pain point identification system can carry out pain point matching on the comment text of the target product with negative emotion in a preset time period based on the pain point label library, so that a pain point matching result of the target product is obtained. The target product matching result can include the number of pain spots of each product in addition to the product pain spots. For example, the pain point recognition result for a certain cosmetic may include not only "no return", "few gift", "allergic" and other product pain points, but also statistics of the number of comments on each product pain point.

The method for extracting the negative keywords from the feedback bad evaluation of the consumers is abandoned, the method is changed into a method taking trend life style as a guide, a large scene is split according to human demand levels, the method is further split into subdivision scenes according to the difference of consumption concepts and the difference of products, and then a pain point tag library is determined. As shown in fig. 2, taking the charging requirement as an example, based on the crowd-scene-pain point specific construction application, it can be seen that the characteristic requirements of different crowds on the charging products have obvious priority differences. Compared with the prior art, the method only has the defect that the general pain point can be identified, the pain point identification is clearer and more specific, and the association relation of people/products/time/events/space and the like related to the pain point can be clarified.

The method can refine the long-term trend continuous tracking result according to industry hot spot monitoring, qualitative trend signals according to analysis methods such as external-internal factor, PEST influence, venation evolution and the like, evaluate trend potential and then construct a pain spot label library. Specifically, for the tracked trend, further hierarchical disassembly and subdivision scenes and labels can be performed according to a Marlo demand model, domestic and foreign consumption concepts, life modes and the like, classification is performed according to the characteristics of the industry according to crowd, products, sites, time and the like, and an industry general label library and an industry characteristic label library are constructed. The industry universal tag library can be determined based on universal pain point words of the industry to which the target product belongs, and tags such as crowd and the like can be universal across industries and can be used as the universal pain point words of the industry universal tag library. The industry characteristic tag library can be determined based on characteristic pain point words of the industry to which the target product belongs, and the tag such as material has strong industry characteristics and can be used as the characteristic pain point words of the industry characteristic tag library.

The pain point label library can comprise an industry general label library and an industry characteristic label library, and also can comprise an inapplicable label library. The unsuitable tag library may be determined based on target product unsuitable pain spot words from universal pain spot words of the industry universal tag library. For example: for the washing and caring industry, the drying can be used as a general pain point word of an industry general label library of products, and the drying can be used for describing feedback of users using washing and caring products (such as facial cleanser). For dry skin crowds, dryness is a product pain spot; for the oil skin population, drying is not a product pain spot. For this case, the unsuitable tag library may be further determined based on the target product unsuitable pain spot words (facial cleanser drying unsuitable for describing the pain spots of the oil skin population) among the universal pain spot words of the industry universal tag library.

In step 120, the pain point identification system may determine a prevalence indicator for each pain point based on a ratio of the number of each pain point to the number of all pain points in the pain point matching result; and determining an urgency index for each pain point based on the rate of increase in the number of each pain point in the pain point matching result compared to the corresponding number of pain points in the previous preset time period. For example: the number of all pain points in the pain point matching result in the preset time period is A, wherein the number of a certain pain point is a, and the number of the pain point in the previous preset time period is b. The universality index of the pain points is equal to the number of the pain points in the preset time period/the total number of the pain points in the preset time period multiplied by 100 percent, namely a/A multiplied by 100 percent; the urgency index of the pain spot is equal to (the number of the pain spot in the preset time period-the number of the pain spot in the preset time period before)/the number of the pain spot in the preset time period before multiplied by 100%, namely (a-b)/b multiplied by 100%.

The pain point identification system can classify each pain point in the pain point matching result based on the calculated universality index of each pain point and the calculated urgency index of each pain point, so as to obtain a target product pain point identification result.

In one embodiment, classifying the pain point matching result based on the universality index of each pain point and the urgency index of each pain point to obtain a target product pain point identification result, including:

determining the average value of the universality index of all pain points in the pain point matching result based on the universality index of each pain point; based on the urgency index of each pain point, determining the average value of urgency indexes of all pain points in the pain point matching result;

classifying the pain points according to the pain point matching result based on the universality index of each pain point and the position of the urgency index of each pain point in the secondary pain point classification map to obtain a target product pain point identification result; the pain point identification result comprises a core pain point, a public pain point, an emergency pain point and an edge pain point; the core pain point represents a pain point with a universality index of the pain point being greater than or equal to the average value of the universality index of all the pain points, and with an urgency index of the pain point being greater than the average value of the urgency indexes of all the pain points; the general pain points represent the pain points with the universality index of the pain points being larger than or equal to the average value of the universality index of all the pain points and the urgency index of the pain points being smaller than or equal to the average value of the urgency index of all the pain points; the emergency pain points represent the pain points with the universality index of the pain points smaller than the average value of the universality index of all the pain points and the urgency index of the pain points larger than the average value of the urgency index of all the pain points; the marginal pain points represent the pain points whose prevalence index is less than the average of the prevalence index of all the pain points, and whose urgency index is less than or equal to the average of the urgency index of all the pain points.

The pain point identification system can calculate the universality index of each pain point, calculate the average value of the universal index, and serve as the average value of the universality index of all the pain points in the pain point matching result; and the urgency index of each pain point can be calculated, and the average value of the urgency indexes is calculated and used as the average value of urgency indexes of all pain points in the pain point matching result. And then determining a secondary pain point grading map based on the average value of the universality index of all the pain points and the average value of the urgency index of all the pain points. Specifically, the x-axis may represent the popularity index, the y-axis may represent the urgency index, and four quadrants may be divided according to the average value of the popularity index of all pain points and the average value of the urgency index of all pain points to obtain a secondary pain point classification map (as shown in fig. 3).

The pain point identification system can determine the quadrant where the pain point is located based on the position of the universality index of each pain point and the urgency index of each pain point in the secondary pain point classification map, so as to classify the pain point of the pain point matching result, and obtain the pain point identification result of the target product. Based on the popularity index and urgency index of each pain spot, the pain spot recognition result can be specifically classified into a core pain spot, a public pain spot, an emergency pain spot and an edge pain spot (as shown in fig. 4).

In particular, the core pain point may represent a pain point that is ubiquitous and has an increased negative impact on the user, and may be considered as a primary need to address the pain point. The general pain point may represent a pain point that is ubiquitous but has limited negative impact on the user. The general problem of the industry is common to the mass pain point, and planned and long-term progressive improvement is needed to avoid the problem of the mass pain point being converted into a core pain point due to poor treatment. Emergency pain spots may represent pain spots that occur only in a small number of situations, but once they occur have a large negative impact on the consumer. The emergency pain point can be selectively solved, and the proper treatment can be the focus of product differentiation. Edge pain points may represent pain points that occur in only a small number of situations and have limited negative impact on the consumer. The edge pain point may generally be treated without allocating resources first, the treatment priority is reduced, but tracking may be kept.

In one embodiment, based on the pain point tag library, performing pain point matching on a comment text of a target product with negative emotion in a preset time period to obtain a pain point matching result of the target product, including:

and filtering the unsuitable pain point words in the pain point matching result of the initial target product based on the unsuitable label library in the pain point label library to obtain the pain point matching result of the target product.

When pain point matching is performed, the pain point recognition system can perform pain point matching on the target product comment text with negative emotion in a preset time period based on an industry general tag library and an industry characteristic tag library in the pain point tag library, for example, a naive Algorithm, a KMP (Knuth-Morris-Pratt Algorithm) Algorithm, a BM (Bayer-Moore Algorithm) Algorithm and the like are used for matching, so that an initial target product pain point matching result is obtained.

The pain point matching result of the initial target product may contain labels matched to an industry general label library, but the labels are nonsensical, and the labels in the part may be mismatched due to word segmentation relationship, so that the construction relationship is disordered. The pain point identification system can match the pain point matching result of the initial target product based on the inapplicable label library in the pain point label library, and filter inapplicable pain point words matched with the inapplicable label library in the pain point matching result of the initial target product, so as to obtain the pain point matching result of the target product. For example: if the target product is a facial cleanser, the population of oil skin matches the pain spot word dry, and facial cleanser dry is not a pain spot for oil skin, so it can be filtered by an inapplicable tag library.

According to the product pain point identification method, product pain points are matched through the industry universal label library, the industry characteristic label library and the inapplicable label library, not only can product pain points be matched, but also meaningless pain points can be filtered, the accuracy of product pain point identification can be further improved, the readability of the pain points is improved, and the problems of no matching, high cost of manual interpretation and correction are avoided.

In one embodiment, performing pain point matching on a target product comment text with negative emotion in a preset time period based on an industry universal tag library and an industry characteristic tag library in a pain point tag library comprises:

generating a first vector based on product pain point words corresponding to an industry universal label library and an industry characteristic label library in the pain point label library; generating a second vector based on target keywords of target product comment texts with negative emotions in a preset time period; the target keywords represent feedback to the target product;

and carrying out regular matching on the first vector and the second vector.

The industry general label library in the pain point label library can contain general product pain point words, and the industry characteristic label library can contain characteristic product pain point words. When product pain point words corresponding to the industry universal tag library and the industry characteristic tag library are matched with target keywords of target product comment texts with negative emotions in a preset time period, vectorization can be performed on the product pain point words and the target keywords, and then regular matching is performed.

The pain point identification system can generate a first vector based on product pain point words corresponding to an industry universal label library and an industry characteristic label library in the pain point label library; and generating a second vector based on target keywords of the target product comment text with negative emotions in a preset time period, and then carrying out regular matching on the first vector and the second vector, so as to obtain an initial target product pain point matching result. Wherein the target keyword may represent feedback to the target product. For example: by regular matching, "unsuitable" can be matched with "less suitable".

According to the product pain point identification method, the initial target product pain point matching result can be obtained through vectorization processing and regular matching of the product pain point words and the target keywords, and the accuracy of product pain point identification is further improved.

In one embodiment, the method for identifying pain points of a product further comprises:

In some cases, the following phenomenon may exist: the method comprises the steps that a target keyword of a target product comment text with negative emotion in a preset time period is inconsistent with a product pain point word corresponding to an industry universal tag library and an industry characteristic tag library in a pain point tag library, but the target keyword has the same or similar meaning as the product pain point word to be expressed, and the target keyword is matched through the product pain point word, so that the target keyword cannot be counted in a product pain point matching result, and the accuracy of product pain point identification is affected.

Based on this, the product pain point recognition method provided by the embodiment of the application can calculate the similarity between the target keyword and the product pain point words, and map the target keyword to the corresponding product pain point words when the similarity between the target keyword and a certain product pain point word is greater than the similarity threshold.

For example, vocabulary similarity calculations may be made based on semantic dictionaries, based on corpus statistics, based on the number of pages retrieved, and so forth. The similarity threshold may be set according to a specific situation, and the application is not specifically limited. When the pain point recognition system calculates that the similarity between the target keyword and a certain product pain point word is larger than a similarity threshold, the target keyword can be mapped into the corresponding product pain point word. For example, if a certain brand of lipstick is provided, the pain point recognition system stores the product pain point words as "uncomfortable", the similar target keywords are "not silky", the similarity between the uncomfortable words and the non-silky is larger than a threshold value, and then the pain point recognition system can map the target keywords as "not silky" into the corresponding product pain point words as "uncomfortable".

The implementation of the technical solution provided in the embodiments of the present application may also be based on the following manner: and directly matching the product pain point words with the target keywords, and for the target keywords which are not matched with the product pain point words, further matching the similarity between the target keywords and the product pain point words, and when the similarity between the target keywords and a certain product pain point word is greater than a preset similarity threshold, mapping the target keywords into the product pain point words.

The pain point recognition system can match the stored product pain point words with the target keywords, and the vocabulary of the product pain point words stored by the pain point recognition system is limited, so that the situation that the target keywords are different from the product pain point words and cannot be matched may exist. The unmatched target keywords may be words having the same meaning or similar meaning to the pain point words of the product, and the pain point recognition system may perform similarity matching on the unmatched target keywords. For example, a certain brand of bath lotion is characterized in that a product pain point word is smell bad, a target keyword is smell strange, and because the target keyword is smell strange and the product pain point word cannot be matched when the target keyword is smell strange when the target keyword and the product pain point word are directly matched by the pain point recognition system, the pain point recognition system can further match the similarity between the target keyword, smell strange and the product pain point word, so that the similarity between the smell strange and the smell strange is larger than a similarity threshold, and the pain point recognition system can map the target keyword, smell strange, into the corresponding product pain point word, smell strange.

According to the product pain point identification method, the range of pain point identification can be further enlarged through similarity matching and mapping of the target keywords and the product pain point words, the situation that comment text pain points are not identified due to limited product pain point words is effectively avoided, and therefore accuracy of product pain point identification is improved.

determining that the ratio of the total pain points in the pain point matching result of the target product to the number of the comment texts of the target product with negative emotion in a preset time period is smaller than the preset ratio, and updating the product pain point words corresponding to the industry universal tag library and the industry characteristic tag library in the pain point tag library based on the target keywords in the comment texts of the target product with negative emotion in the preset time period.

When the ratio of the total pain points in the pain point matching result of the target product to the number of the comment texts of the target product with negative emotion in the preset time period is smaller than the preset ratio, the product pain point words corresponding to the target product can be updated based on the target keywords.

The preset ratio may be specifically 40%, 50%, etc., and the size thereof may be adjusted according to actual conditions, which is not specifically limited in the present application.

After the pain point recognition system obtains the pain point matching result of the target product, the ratio of the total number of the pain points in the pain point matching result of the target product to the number of comment texts of the target product with negative emotion in a preset time period can be calculated. When the ratio of the number of all pain points to the number of the comment texts of the target product with negative emotion is smaller than a certain value, the condition that the pain point words of the product are not abundant may exist, and at the moment, the pain point words of the product corresponding to the target product can be updated based on the target keywords.

When the product pain point words are updated based on the target keywords, certain standards can be referred, such as the frequency or the number of the target keywords can be referred, and when the frequency or the number of the target keywords is high, the product pain point words can be updated.

For example, the target keywords are "strange smell" which is a word describing pain points of the product, and the frequency of occurrence is high (for example, the frequency of occurrence in all the target keywords reaches 10 percent, or the frequency of occurrence is high in a certain time period, for example, comment texts obtained by a pain point recognition system are 2-6 months, and the frequency of occurrence of "strange smell" in 6 months is high); or the number of the target keywords of the smell strange is large (for example, 15 target keywords are smell strange in 100 target keywords), the smell strange can be updated into an industry general tag library and an industry characteristic tag library in the pain point tag library.

According to the product pain point identification method, the product pain point words are updated and perfected continuously, so that the robustness of the product pain point identification method can be improved effectively, and the accuracy of product pain point identification is improved continuously.

The target product comment text with the negative emotion is screened from the target product comment text within a preset time period, and the target product comment text can be subjected to negative emotion word matching based on the negative emotion words of the industry to which the target product belongs, wherein the comment text comprising the negative emotion words is the target product comment text with the negative emotion.

Before screening the target product comment text with negative emotion from the target product comment text, a comment text sample of the target product within a preset time period can be obtained, the comment text sample is preprocessed, and the preprocessed comment text sample can be used as the target product comment text.

The pain point recognition system can extract comment texts of the target product within a preset time period and serve as samples, and then preprocessing steps such as data duplication removal processing, symbol and expression processing, nonsensical text processing, text data normalization processing and the like are carried out on sample data. Where nonsensical text processing such as processing of default ratings, etc. And the pretreated comment text sample is the comment text of the target product.

The pain point recognition system can extract and store negative emotion words of industries to which the target product belongs, and match the negative emotion words with comment texts of the target product, if the negative emotion words are matched, comment texts comprising the negative emotion words can be used as comment texts of the target product with negative emotion, and if the negative emotion words are not matched, the comment texts of the target product are positive emotion, and then the comment texts of the target product do not comprise product pain points. For example, the comment text of the target product is "my skin dry skin can be used", in the prior art, the product pain point is usually contained in the comment text because the dry word is identified, the negative emotion word is firstly matched with the comment text, the comment text is identified as not being matched with the negative emotion word, the text emotion is positive emotion, and the identification result is that the product pain point is not contained.

According to the product pain point identification method, the negative emotion of the comment text of the target product is identified, and then pain point matching is carried out, so that the pain point identification of the comment text with positive emotion is avoided, and the product pain point identification efficiency and accuracy are improved.

The following further describes the technical solution of the present application with an example of a method for identifying pain points of products provided by the embodiments of the present application.

As shown in fig. 5, this example may include the steps of:

(1) Creating a pain point label library: splitting the scene pain points according to the group demand level, and clearly displaying the corresponding pain points after being associated with specific entities, wherein the pain points are shown in the following table:

and constructing a pain point label library based on the pain point splitting result.

(2) Text data input: public opinion data of a specific industry in the processing time range is extracted, and data deduplication, symbol and expression processing, nonsensical text processing (such as default evaluation and the like) and text data normalization processing are carried out to obtain comment texts.

(3) Judging text emotion: and extracting negative words of an emotion word bank of the industry to which the target product belongs, splicing a text array, vectorizing and matching the comment text by using the text array, and returning a result to a list. The list length corresponds to how many negative terms the comment text matches. And each element value in the list is the number of each passive word matched with the comment text. Therefore, the comment text with negative emotion is obtained when the list element value is larger than 0.

(4) Outputting negative emotion text: element numbers with extraction list element values greater than 0 form a vector m. And indexing the comment text by using the vector m, and extracting the first comment text df with negative emotion.

(5) Pain point matching is carried out: and carrying out structural construction according to the form of entity-concept-relation, extracting a product pain point word splicing text array, carrying out vectorization matching on a target keyword of a first comment text by using a pain point tag library, and returning a result to a list. The list length corresponds to how many pain point phrases the first comment text target keyword is matched with. The values of the elements in the list are the number of pain point words matched with each product.

(6) Outputting a matching phrase: the element serial numbers with the number of the elements of the extraction result list being more than 0 form a vector n. And indexing the first comment text df by using the vector n to obtain text data res. And splicing the text data res and the corresponding list elements to obtain a two-dimensional data table containing text contents and matched pain point phrases.

(7) Matching mapping relation: and performing similarity matching on target keywords which are not matched with the pain point words of the product in the first comment text and the pain point word groups, and correcting the pain point result of the product based on the number of the target keywords when the similarity of the target keywords and the pain point word groups is larger than a threshold value.

(8) And (5) outputting pain points of the product: filtering the inapplicable label and outputting the pain point matching result of the product.

(9) Classifying pain points of the product: calculating the universality index and the urgency index of a preset time period, constructing a four-quadrant diagram of pain point classification, dividing edge pain points, public pain points, emergency pain points and core pain points, and providing pain point classification treatment suggestions.

The product pain point identification device provided in the embodiment of the present application is described below, and the product pain point identification device described below and the product pain point identification method described above may be referred to correspondingly.

Fig. 6 is a schematic structural diagram of a pain point identification device for a product according to an embodiment of the present application. Referring to fig. 6, the apparatus may include:

the matching module 610 is configured to perform pain point matching on the comment text of the target product with negative emotion in the preset time period based on the pain point tag library, so as to obtain a pain point matching result of the target product; the pain point tag library comprises an industry general tag library, an industry characteristic tag library and an inapplicable tag library; the industry universal tag library is a tag library determined based on universal pain point words of the industry to which the target product belongs; the industry characteristic tag library is a tag library determined based on characteristic pain point words of the industry to which the target product belongs; the inapplicable tag library is a tag library determined based on inapplicable pain spot words of target products in general pain spot words of the industry general tag library;

the grading module 620 is configured to grade the pain point matching result based on the universality index of each pain point and the urgency index of each pain point, so as to obtain a target product pain point identification result; the universality index of each pain point is determined based on the ratio of the number of each pain point to the number of all pain points in the pain point matching result; the urgency index of each pain point is determined based on the rate of increase of the number of each pain point in the pain point matching result compared to the corresponding number of pain points in the previous preset time period.

In one embodiment, the classification module 620 is specifically configured to:

In one embodiment, the matching module 610 is specifically configured to:

In one embodiment, the matching module 610 is further configured to:

and carrying out regular matching on the first vector and the second vector.

In one embodiment, the matching module 610 is further configured to:

It should be noted that, the product pain point identification device provided in the embodiment of the present application can implement all the method steps implemented in the method embodiment and achieve the same technical effects, and the same parts and beneficial effects as those of the method embodiment in the embodiment are not described in detail herein.

Fig. 7 illustrates a physical schematic diagram of an electronic device, as shown in fig. 7, which may include: processor 710, communication interface (Communications Interface) 720, memory 730, and communication bus 740, wherein processor 710, communication interface 720, memory 730 communicate with each other via communication bus 740. Processor 710 may call logic instructions in memory 730 to perform the following method:

Further, the logic instructions in the memory 730 described above may be implemented in the form of software functional units and may be stored in a computer readable storage medium when sold or used as a stand alone product. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the related art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

In another aspect, embodiments of the present application disclose a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, are capable of performing the methods provided by the method embodiments described above, for example comprising:

In yet another aspect, embodiments of the present application further provide a non-transitory computer readable storage medium having stored thereon a computer program that, when executed by a processor, is implemented to perform the method provided by the above embodiments, for example, comprising:

The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.

From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the related art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the method described in the respective embodiments or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only for illustrating the present application, and are not limiting of the present application. While the present application has been described in detail with reference to the embodiments, those skilled in the art will understand that various combinations, modifications, or equivalents of the technical solutions of the present application may be made without departing from the spirit and scope of the technical solutions of the present application, and all such modifications are intended to be covered by the claims of the present application.

Claims

1. A method for identifying pain points of a product, comprising:

2. The method for identifying pain spots of a product according to claim 1, wherein the step of classifying the pain spots based on the prevalence index of each pain spot and the urgency index of each pain spot to obtain the identification result of the pain spot of the target product comprises:

3. The method for identifying pain spots of products according to claim 1, wherein the performing pain spot matching on the comment text of the target product with negative emotion in the preset time period based on the pain spot tag library to obtain a pain spot matching result of the target product comprises the following steps:

4. The method for identifying pain spots of products according to claim 3, wherein the performing pain spot matching on the comment text of the target product with negative emotion in the preset time period based on the industry universal tag library and the industry characteristic tag library in the pain spot tag library comprises:

And carrying out regular matching on the first vector and the second vector.

5. The method of claim 4, further comprising:

6. The method of claim 4, further comprising:

7. The method of claim 1, further comprising:

8. A product pain point identification device, comprising:

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of product pain point identification according to any one of claims 1 to 7 when the computer program is executed by the processor.

10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the product pain point identification method according to any of claims 1 to 7.