CN107545457B - Automobile racing product type determination method and device - Google Patents
Automobile racing product type determination method and device Download PDFInfo
- Publication number
- CN107545457B CN107545457B CN201710075961.2A CN201710075961A CN107545457B CN 107545457 B CN107545457 B CN 107545457B CN 201710075961 A CN201710075961 A CN 201710075961A CN 107545457 B CN107545457 B CN 107545457B
- Authority
- CN
- China
- Prior art keywords
- vehicle type
- vehicle
- characteristic
- index
- determining
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the field of Internet data mining, in particular to a method and a device for determining automobile model of automobile racing products, wherein the method comprises the following steps: acquiring at least one characteristic index of each vehicle type from vehicle basic data and user behavior data of a plurality of vehicle related websites and applications; determining the weight of each characteristic index according to a principal component analysis rule, and generating an array containing numerical values corresponding to the characteristic indexes for each vehicle type; determining a characteristic point on a corresponding coordinate system by each array, and calculating the association degree of any two characteristic points so as to construct a competitive model prediction model; when a vehicle type is input, acquiring a target characteristic point of which the association degree of the characteristic point corresponding to the vehicle type is within a preset range according to the competitive product vehicle type prediction model, and determining a corresponding competitive product vehicle type according to the target characteristic point. The method and the device can determine the competitive product vehicle type of the vehicle type more quickly and accurately, and provide more accurate competitive product vehicle type information for the user.
Description
[ technical field ] A method for producing a semiconductor device
The invention relates to the field of internet data mining, in particular to a method and a device for determining automobile model of automobile racing products.
[ background of the invention ]
Digital marketing has become one of the main ways for enterprises to promote branding and sales. In recent years, with the rapid development of big data technology and application, the operation mode of digital marketing, from budget allocation to delivery adjustment to effect evaluation after delivery, has been profoundly changed.
One of the most important influences of big data on digital marketing is that the big data can be integrated with various source data to perform integrated analysis on the big data of a user, when online advertisement putting and product design of a certain automobile product are performed, a 'competitive model type' of the automobile is generally required to be obtained, advertisement putting and product design are facilitated, and generally speaking, the competitive model types of the automobile product in a product planning stage, a development stage and an on-sale stage are different.
The competitive automobile models in the product planning stage and the development stage are more subjective judgment of design and developers, and cannot be adjusted along with feedback of consumers, and the automobile products in the sale stage have feedback of the consumers, so that the competitive automobile models are identified more accurately. At present, the implementation mode of the competitive products motorcycle type is simpler, more operates through experience, and several common forms have:
firstly, the competitive products are defined by feeling or experience;
secondly, taking technical specifications, prices and the like as measuring indexes, and taking similar persons as competitive products;
thirdly, the first few name cards sold in the market segment are directly used as competitive products;
fourthly, the competitive product method is formed by combining a plurality of factors such as sales volume, budget, market positioning, enterprise competitive brands and the like, and is shown in figure 3.
The prior art has the following defects:
(1) different persons can give different competitive product vehicle types due to the difference of feeling and experience, and the accuracy is greatly fluctuated;
(2) the technical specification and the price measurement index are a static evaluation mode, and have larger difference with the real feeling of a user;
(3) judging the situation that the competitive product vehicle types cannot be distinguished from different grades according to the sales volume, wherein the situation has a great defect;
(4) the method for constructing the competitive product vehicle type through dimensions such as sales volume, budget, market positioning, enterprise competitive brands and the like cannot correct the competitive product vehicle type according to the feedback of a user and cannot dynamically adjust.
[ summary of the invention ]
The invention aims to provide a method and a device for determining automobile competitive product models.
In order to realize the purpose, the invention adopts the following technical scheme:
in a first aspect, the invention provides a method for determining vehicle types of automobile competitive products, which comprises the following steps:
acquiring at least one characteristic index of each vehicle type from vehicle basic data and user behavior data of a plurality of vehicle related websites and applications;
determining the weight of each characteristic index according to a principal component analysis rule, and generating an array containing numerical values corresponding to the characteristic indexes for each vehicle type;
determining a characteristic point on a corresponding coordinate system by each array, and calculating the association degree of any two characteristic points so as to construct a competitive model prediction model;
when a vehicle type is input, acquiring a target characteristic point of which the association degree of the characteristic point corresponding to the vehicle type is within a preset range according to the competitive product vehicle type prediction model, and determining a corresponding competitive product vehicle type according to the target characteristic point.
Specifically, the characteristic indexes of the vehicle type include one or more of a sales lead index, a price reduction index, an attention index and a contrast index.
Specifically, the sales lead index of the vehicle type is determined by the following steps:
respectively acquiring the sales lead amount submitted by each vehicle type of each city;
and counting the sum of sales leads of all cities in each vehicle type to serve as the sales lead index of each vehicle type.
Further, the price reduction index of the vehicle type is determined by the following steps:
obtaining the selling price average value and the price reduction average value of different vehicle types of the same vehicle type in all cities;
obtaining the price reduction percentage of the vehicle type according to the selling price mean value and the price reduction;
and multiplying the percentage of the price reduction by a preset value to serve as the price reduction index of the vehicle type.
Further, the attention index of the vehicle type is determined by the following steps:
respectively acquiring the page browsing amount of each vehicle type of each city;
and calculating the sum of the page browsing amount of each vehicle type of all cities as the attention index of each vehicle type.
Further, the comparison index of the vehicle type is determined by the following steps:
acquiring the times of each vehicle money compared with each other by the user;
and determining the corresponding vehicle type according to the vehicle money, and calculating the sum of the compared times of each vehicle type to serve as the comparison index of each vehicle type.
Preferably, the characteristic indexes are obtained from vehicle basic data and user behavior data of the vehicle-related websites and applications within a preset time.
Optionally, the degree of association between any two feature points includes a correlation coefficient or a euclidean distance between any two feature points.
In a second aspect, the present invention provides a vehicle type determination device for vehicle racing, comprising:
an acquisition module: the system comprises a plurality of automobile relevant websites, a plurality of application programs and a plurality of characteristic indexes, wherein the automobile relevant websites are used for acquiring automobile basic data and user behavior data of each automobile type;
a weight determination module: the weight of each characteristic index is determined according to principal component analysis rules, and an array containing numerical values corresponding to the characteristic indexes is generated for each vehicle type;
a relevance determination module: the system comprises a plurality of arrays, a model calculation module, a model calculation module and a model matching module, wherein the arrays are used for determining a characteristic point on a corresponding coordinate system and calculating the association degree of any two characteristic points so as to construct a competitive model prediction model;
a vehicle type determination module: and when a vehicle type is input, acquiring a target characteristic point of which the association degree of the characteristic point corresponding to the vehicle type is within a preset range according to the competitive product vehicle type prediction model, and determining the corresponding competitive product vehicle type according to the target characteristic point.
Compared with the prior art, the invention has the following advantages:
the characteristic indexes selected by the invention can well show the characteristics of the vehicle type, and the characteristics of the competitive products are similar, wherein the sales clue indexes of the vehicle type are important indexes reflecting the popularity of the vehicle to users; the price reduction index takes into account that the vehicle models of the competitive products have similar modes on the sale strategy; the attention index is approximately the same considering that the browsing condition of the user to the competitive product vehicle type; the comparison index is one of the most intuitive indexes for reflecting the competitive products, and the product vehicle concerned by the user and the related competitive product vehicle can be more intuitively seen through the comparison condition of the user to the vehicle type;
meanwhile, the weight of the characteristics is determined by the model by using a clustering idea and a principal component analysis method, so that the model has good mathematical theory support;
moreover, the model can flexibly and dynamically adjust the source of the characteristic index data, so that the obtained competitive product vehicle type can better adapt to the market environment and the user requirements.
It is to be understood that the foregoing general description of the advantages of the present invention is provided for illustration and description, and that various other advantages of the invention will be apparent to those skilled in the art from this disclosure.
Additional aspects and advantages of the invention 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 invention.
[ description of the drawings ]
FIG. 1 is a schematic flow chart illustrating an embodiment of a method for determining vehicle types of racing cars according to the present invention;
FIG. 2 is a diagram of the data associated with the 'Passat 2016 type' on the present automotive website;
FIG. 3 is a schematic diagram of a prior art method for analyzing racing vehicle models;
fig. 4 is a schematic diagram of an embodiment of the vehicle model determination device for the automobile racing product of the present invention.
[ detailed description ] embodiments
The present invention is further described with reference to the drawings and the exemplary embodiments, wherein like reference numerals are used to refer to like elements throughout. In addition, if a detailed description of the known art is not necessary to show the features of the present invention, it is omitted.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
Before the detailed description of the present embodiment, the following explanations will be made for the vehicle type and the vehicle model in the present embodiment:
the vehicle models are divided on the basis of vehicle brands, for example, popular brands under Germany vehicles, and vehicle models such as Passat, Tourist and Huiteng are arranged under the popular brands.
The vehicle types refer to different styles of the same vehicle type, such as 2015 and 2016 in Passat and 1 to 7 generations in Honda Accord, which belong to the vehicle types.
Step S100: and acquiring at least one characteristic index of each vehicle type from the vehicle basic data and the user behavior data of the plurality of vehicle related websites and applications.
The purpose of this embodiment is to establish a model for predicting vehicle models of vehicle racing products, so as to find out the vehicle models of racing products corresponding to a certain vehicle model more accurately. In the embodiment, at least one characteristic index of each vehicle type is acquired from internet websites, vehicle basic data of application programs and user behavior data related to a plurality of automobiles, such as websites of 'easy driving', 'home of automobile' and the like, which provide a large amount of vehicle basic data, the vehicle basic data includes the style, configuration, manufacturer guide price, merchant sales price and the like of the vehicle, a user can acquire various basic information of the vehicle from the websites, meanwhile, various operation behaviors of the user on the websites are recorded as the user behavior data, and the acquired vehicle basic data and the user behavior data provide data support for the competitive product vehicle type prediction model of the embodiment after being converted into corresponding characteristic indexes by the following method.
Specifically, the characteristic index into which the vehicle basic data and the user behavior data are converted includes one or more of a sales lead index, a price reduction index, a degree of attention index, and a comparison index.
The Sales Lead, also called Sales Lead, is the head end of the customer's opportunity in the Sales management system, for example, a user may leave his/her own contact after browsing a car advertisement, which is recorded as a Sales Lead of one time. For example, the user browses the Accord 7 generation on the "easy-to-drive network", fills in the own contact way to reserve to a local 4S car store to take a car and test a car, and the sales lead amount of the Accord 7 generation is recorded as 1 time. In this embodiment, the sales lead amount submitted by each vehicle type of each city is first obtained from a plurality of vehicle-related websites and applications, and then the total sales lead of each vehicle type of all cities is counted as the sales lead index of each vehicle type.
In this embodiment, the average value of the price reduction of all the vehicle money of a vehicle type is defined as the price reduction value of the vehicle type, the average value of all the vehicle money of the vehicle type is the vehicle value of the vehicle type, the price reduction percentage of the vehicle type is obtained according to the price reduction value and the vehicle value, and then the price reduction percentage is multiplied by a preset value to serve as the price reduction index of the vehicle type. As shown in fig. 2, the data related to "passat 2016 type" on an automobile website, and the price of the automobile is marked with "price reduction 2.5 ten thousand yuan, and the percentage of price reduction is 14%", generally, the automobile related website will indicate the price reduction information of an automobile, in this embodiment, the average value of the price reduction of all the vehicle types of the automobile type is firstly counted, then the percentage of price reduction is calculated with the manufacturer-directed average value of all the vehicle types of the automobile type, and the percentage of price reduction is multiplied by a preset value to serve as the price reduction index of the automobile type, and in this embodiment, the preset value is set to 10000.
The attention degree refers to a click amount/a browsing amount/a search amount of a certain vehicle type in user behavior data, in this embodiment, taking the browsing amount as an example for explanation, a user enters an information display interface of "XX vehicle type" in different ways, the information display interface is recorded as a browsing amount (or referred to as PV flow) every time the information display interface is refreshed, and the attention degree of each vehicle type can be intuitively obtained through the PV flow.
Meanwhile, most automobile websites/applications provide a function of comparing automobile types and automobile money, and a user can better compare different information of two or more types of automobiles, generally, after the user selects a plurality of compared automobile money, relevant websites/applications can record automobile money fields compared by the user, then the plurality of automobile money compared by the user each time can be obtained by splitting the fields, then the corresponding automobile type can be obtained by the automobile money, the automobile money compared by the user each time is marked as 1 comparison amount of the corresponding automobile type, and in the embodiment, the comparison amount of all automobile types is counted to be integrated as the comparison index of each automobile type.
Preferably, the characteristic index is obtained from vehicle basic data and user behavior data of the vehicle-related website and the application within a preset time, and the preset time may be 1 month, 2 months, half a year, and the like.
Step S200: and determining the weight of each characteristic index according to a principal component analysis rule, and generating an array containing numerical values corresponding to the characteristic indexes for each vehicle type.
The influence degree of different characteristic indexes on the forecast of the competitive product vehicle type is different, the embodiment determines the weight of each characteristic index on the forecast of the competitive product vehicle type through a principal component analysis rule, the principal component analysis is a multivariate statistical method for investigating the correlation among a plurality of variables, the research is carried out on how to disclose the internal structure among the plurality of variables through a few principal components, namely, the few principal components are derived from the original variables, so that the information of the original variables is kept as much as possible and is not correlated with each other, and the original P indexes are subjected to linear combination to be used as a new comprehensive index through general mathematical processing; the most classical way is to express the variance of F1 (the first linear combination selected, i.e. the first comprehensive index), i.e. the larger Var (F1) is, the more information F1 contains. Therefore, the variance of the selected F1 in all linear combinations should be the largest, so the first principal component is called F1. If the first principal component is not enough to represent the original information of P indexes, F2 is selected, namely, the second linear combination is selected, in order to effectively reflect the original information, the information existing in F1 does not need to appear in F2, the information expressed by the mathematical language requires that Cov (F1, F2) is 0, then F2 is called the second principal component, and so on, the third, fourth, … …, P-th principal component can be constructed. Specifically, the principal component analysis comprises the following steps:
1. standardized acquisition of raw index data with p-dimensional random vector X ═ X (X)1，X2，X3，...Xp)TN samples xi＝(xi1，xi2，...，xip)TN, n > p, constructing a sample array, and carrying out the following standardized transformation on sample array elements:
wherein, the first and second guide rollers are arranged in a row,
2. matrix of correlation coefficients for normalized matrix Z
3. solving the characteristic square | R-lambada I of the sample correlation matrix RpI features p, determining principal components according to
Determining m value to make the utilization rate of information reach above 85%, and determining each lambdajJ 1, 2,.. m, solving the system of equations Rb λj. Obtaining unit feature vector
4. Converting the normalized index variable into a principal component
U1Referred to as the first principal component, U2Is called the second principal component, UpReferred to as the pth principal component.
5. Comprehensive evaluation of m principal components
And carrying out weighted summation on the m principal components to obtain a final evaluation value, wherein the weight is the variance contribution rate of each principal component.
Through the calculation, the weights of the sales lead index, the price reduction index, the attention index and the comparison index are respectively 0.33, 0.04, 0.31 and 0.32, an array containing the corresponding numerical value of the characteristic index is generated for each vehicle type, and the array is multiplied by the respective numerical value of the characteristic index through the weight to obtain an array containing 4 variables.
Step S300: and determining a characteristic point on the corresponding coordinate system by each array, and calculating the association degree of any two characteristic points so as to construct a competitive model prediction model.
For example, the coordinates of one feature point obtained above is X (X, X, X, X), and the coordinates of another feature point is Y (Y, Y, Y, Y), and then the association degree of any two feature points is calculatedThe closeness of the relation between the figures can be indicated by a distance or a correlation coefficient, in one embodiment, points with smaller distance or points with larger similarity coefficient are classified into the same class, and points with larger distance or points with smaller similarity coefficient are classified into different classesi＝(xi1，xi2…，xik)、Xj＝(xj1，xj2…，xjk) N, where k is 1, 2, 3.. n, and the euclidean distance between these two points is
Therefore, each characteristic point on the coordinate system represents a different vehicle type, so that a competitive product vehicle type prediction model is constructed, and the competitive product vehicle type of a vehicle type can be predicted more accurately through the model.
Step S400: when a vehicle type is input, acquiring a target characteristic point of which the association degree of the characteristic point corresponding to the vehicle type is within a preset range according to the competitive product vehicle type prediction model, and determining a corresponding competitive product vehicle type according to the target characteristic point.
The method determines a model for predicting vehicle models, when some 'vehicle model' information is input into the model, the input vehicle model information is converted into corresponding characteristic points on the model, then target characteristic points with the association degree of the characteristic points within a preset range are searched, and then corresponding competitive product vehicle models determined by the target characteristic points are output. Further, the competitive product vehicle types are arranged from large to small according to the association degree, and the higher the association degree is, the higher the possibility that the competitive product vehicle types are mutually increased is. For example, when the input vehicle type information is "pasait", the obtained racing vehicle types are "toyota chemeril", "honda attern", "beck june", and the like.
Accordingly, as shown in fig. 4, a schematic diagram of an embodiment of the device for determining vehicle types on vehicle racing according to the present invention includes:
the acquisition module 100: the system is used for acquiring at least one characteristic index of each vehicle type from vehicle basic data and user behavior data of a plurality of vehicle related websites and applications.
The obtaining module 100 of this embodiment obtains at least one characteristic index of each vehicle type from internet websites related to multiple automobiles, vehicle basic data of application programs, and user behavior data, where websites such as "easy to drive", "home of automobile", and the like all provide a large amount of vehicle basic data, where the vehicle basic data includes style, configuration, manufacturer-guided price, and merchant sales price of a vehicle, and a user may obtain various basic information of the vehicle from such websites, and meanwhile, various operation behaviors of the user on the websites are recorded as the user behavior data.
Specifically, the characteristic indexes converted from the vehicle basic data and the user behavior data by the obtaining module 100 include one or more of a sales lead index, a price reduction index, a degree of attention index, and a comparison index.
The Sales Lead, also called Sales Lead, is the head end of the customer's opportunity in the Sales management system, for example, a user may leave his/her own contact after browsing a car advertisement, which is recorded as a Sales Lead of one time. For example, the user browses the Accord 7 generation on the "easy-to-drive network", fills in the own contact way to reserve to a local 4S car store to take a car and test a car, and the sales lead amount of the Accord 7 generation is recorded as 1 time. The obtaining module 100 of this embodiment first obtains the sales lead amount submitted by each vehicle type of each city from a plurality of vehicle-related websites and applications, and then counts the total sales lead of each vehicle type of all cities as the sales lead indicator of each vehicle type.
In this embodiment, the average value of the price reduction of all the vehicle money of a vehicle type is defined as the price reduction value of the vehicle type, the average value of all the vehicle money of the vehicle type is the vehicle value of the vehicle type, the price reduction percentage of the vehicle type is obtained according to the price reduction value and the vehicle value, and then the price reduction percentage is multiplied by a preset value to serve as the price reduction index of the vehicle type. As shown in fig. 2, the data related to "passat 2016 type" on an automobile website, and the price of the automobile is marked with "price reduction 2.5 ten thousand yuan, and the percentage of price reduction is 14%", generally, the automobile related website will indicate price reduction information of an automobile, the obtaining module 100 of this embodiment first counts the average value of the price reduction of all the vehicle types of the automobile model, and then calculates the percentage of price reduction with the manufacturer-directed average value of all the vehicle types of the automobile model, and multiplies the percentage by a preset value to serve as a price reduction index of the automobile model, where in this embodiment, the preset value is set to 10000.
The attention degree refers to a click amount/a browsing amount/a search amount of a certain vehicle type in the user behavior data, in this embodiment, taking the browsing amount as an example for explanation, a user enters an information display interface of "XX vehicle type" in different ways, the information display interface is recorded as a browsing amount (or referred to as PV flow) every time the information display interface is refreshed, and the attention degree of each vehicle type can be intuitively obtained through the PV flow.
Meanwhile, most automobile websites/applications provide a function of comparing automobile types and automobile money, and a user can better compare different information of two or more types of automobiles, generally, after the user selects a plurality of compared automobile money, relevant websites/applications can record automobile money fields compared by the user, then the plurality of automobile money compared by the user each time can be obtained by splitting the fields, then the corresponding automobile type can be obtained by the automobile money, the automobile money compared by the user each time is recorded as 1 comparison amount of the corresponding automobile type, and in this embodiment, the acquisition module 100 counts the synthesis of the comparison amounts of all automobile types as the comparison index of each automobile type.
Preferably, the characteristic index is obtained from vehicle basic data and user behavior data of the vehicle-related website and the application within a preset time, and the preset time may be 1 month, 2 months, half a year, and the like.
Weight determination module 200: and the weight of each characteristic index is determined according to a principal component analysis rule, and an array containing numerical values corresponding to the characteristic indexes is generated for each vehicle type.
The influence degree of different characteristic indexes on the forecast competitive model is different, the weight determining module 200 of this embodiment determines the weight of each characteristic index on the forecast competitive model through a principal component analysis rule, the principal component analysis is a multivariate statistical method for investigating the correlation among a plurality of variables, and how to disclose the internal structure among the plurality of variables through a few principal components is studied, that is, a few principal components are derived from the original variables, so that the information of the original variables is kept as much as possible and are not correlated with each other, and the general mathematical processing is to linearly combine the original P indexes to serve as a new comprehensive index. Specifically, the principal component analysis comprises the following steps:
1. standardized acquisition of raw index data with p-dimensional random vector X ═ X (X)1，X2，X3，...Xp)TN samples xi＝(xi1，xi2，...，xip)TN, n > p, constructing a sample array, and carrying out the following standardized transformation on sample array elements:
wherein, the first and second guide rollers are arranged in a row,
2. matrix of correlation coefficients for normalized matrix Z
3. solving the characteristic square | R-lambada I of the sample correlation matrix RpI features p, determining principal components according toDetermining m value to make the utilization rate of information reach above 85%, for each timeA lambdajJ 1, 2,.. m, solving the system of equations Rb λj. Obtaining unit feature vector
4. Converting the normalized index variable into a principal component
U1Referred to as the first principal component, U2Is called the second principal component, UpReferred to as the pth principal component.
5. Comprehensive evaluation of m principal components
And carrying out weighted summation on the m principal components to obtain a final evaluation value, wherein the weight is the variance contribution rate of each principal component.
The weight determining module 200 obtains the sales lead index, the price reduction index, the attention index and the comparison index with weights of 0.33, 0.04, 0.31 and 0.32 respectively through the above calculation, and then generates an array containing the corresponding values of the characteristic indexes for each vehicle type, wherein the array is obtained by multiplying the weights by the respective values of the characteristic indexes to obtain an array containing 4 variables.
The association degree determination module 300: and the method is used for determining a characteristic point on the corresponding coordinate system by each array and calculating the association degree of any two characteristic points so as to construct a competitive model prediction model.
In this embodiment, the association degree determining module 300 describes the association degree of two points by using cluster analysis, in the cluster analysis, the closeness degree of the association between commonly used research objects may be indicated by "distance" or "correlation coefficient", in one embodiment, points with smaller distance "or points with larger similarity coefficient" are classified into the same class, points with larger distance "or points with smaller similarity coefficient are classified into different classes, in this embodiment, the association degree between objects is researched by using distance, in one embodiment, the method for measuring distance selects euclidean distance, and for any two points X, the association degree of two points X is described by using distancei＝(xi1，xi2...，xik)、Xj＝(xj1，xj2…，xjk) N, where k is 1, 2, 3.. n, and the euclidean distance between these two points is
At this moment, each feature point on the coordinate system represents a different vehicle type, so that a competitive product vehicle type prediction model is constructed.
The vehicle type determination module 400: and when a vehicle type is input, acquiring a target characteristic point of which the association degree of the characteristic point corresponding to the vehicle type is within a preset range according to the competitive product vehicle type prediction model, and determining the corresponding competitive product vehicle type according to the target characteristic point.
When certain 'vehicle type' information is input into the model, the vehicle type determining module 400 converts the input vehicle type information into corresponding feature points on the model, then searches for target feature points with the association degree of the feature points within a preset range, and then outputs corresponding competitive product vehicle types determined by the target feature points. Further, the competitive product vehicle types are arranged from large to small according to the association degree, and the higher the association degree is, the higher the possibility that the competitive product vehicle types are mutually increased is. For example, when the input vehicle type information is "pasait", the obtained racing vehicle types are "toyota chemeril", "honda attern", "beck june", and the like.
The characteristic indexes selected by the invention can well show the characteristics of the vehicle type, and the characteristics of the competitive products are similar, wherein the sales clue indexes of the vehicle type are important indexes reflecting the popularity of the vehicle to users; the price reduction index takes into account that the vehicle models of the competitive products have similar modes on the sale strategy; the attention index is approximately the same considering that the browsing condition of the user to the competitive product vehicle type; the comparison index is one of the most intuitive indexes for reflecting the competitive products, and the product vehicle concerned by the user and the related competitive product vehicle can be more intuitively seen through the comparison condition of the user to the vehicle type; meanwhile, the weight of the characteristics is determined by the model by using a clustering idea and a principal component analysis method, so that the model has good mathematical theory support; moreover, the model can flexibly and dynamically adjust the source of the characteristic index data, so that the obtained competitive product vehicle type can better adapt to the market environment and the user requirements.
Although a few exemplary embodiments of the present invention have been shown and described, it would be appreciated by those skilled in the art that changes may be made in these exemplary embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the claims and their equivalents.
Claims (9)
1. A method for determining the type of an automobile competitive product is characterized by comprising the following steps:
acquiring at least one characteristic index of each vehicle type from vehicle basic data and user behavior data of a plurality of vehicle related websites and applications;
determining the weight of each characteristic index according to a principal component analysis rule, and generating an array containing numerical values corresponding to the characteristic indexes for each vehicle type;
determining a characteristic point on a corresponding coordinate system by each array, and calculating the association degree of any two characteristic points so as to construct a competitive model prediction model;
when a vehicle type is input, acquiring a target characteristic point of which the association degree of the characteristic point corresponding to the vehicle type is within a preset range according to the competitive product vehicle type prediction model, and determining a corresponding competitive product vehicle type according to the target characteristic point;
wherein the principal component analysis comprises:
obtaining a standardized array according to at least one characteristic index of each vehicle type;
calculating a correlation coefficient matrix according to the normalized matrix;
calculating a characteristic root of the correlation coefficient matrix, and determining a corresponding unit characteristic vector aiming at the characteristic root meeting the condition;
the normalized index variables are converted into principal components.
2. The method of claim 1, wherein the characteristic indicators of the vehicle model include one or more of a sales lead indicator, a price reduction indicator, a focus indicator, and a comparison indicator.
3. The method of claim 2, wherein the lead of sale indicator for the vehicle model is determined by:
respectively acquiring the sales lead amount submitted by each vehicle type of each city;
and counting the sum of sales leads of all cities in each vehicle type to serve as the sales lead index of each vehicle type.
4. The method of claim 2, wherein the price reduction indicator for the vehicle type is determined by:
obtaining the selling price average value and the price reduction average value of different vehicle types of the same vehicle type in all cities;
obtaining the price reduction percentage of the vehicle type according to the selling price average value and the price reduction average value;
and multiplying the percentage of the price reduction by a preset value to serve as the price reduction index of the vehicle type.
5. The method of claim 2, wherein the attention index of the vehicle type is determined by:
respectively acquiring the page browsing amount of each vehicle type of each city;
and calculating the sum of the page browsing amount of each vehicle type of all cities as the attention index of each vehicle type.
6. The method of claim 2, wherein the comparison index for the vehicle type is determined by:
acquiring the times of each vehicle money compared with each other by the user;
and determining the corresponding vehicle type according to the vehicle money, and calculating the sum of the compared times of each vehicle type to serve as the comparison index of each vehicle type.
7. The method according to any one of claims 1 to 6, wherein the characteristic index is obtained from vehicle basic data and user behavior data of the vehicle-related website and application within a preset time.
8. The method according to claim 1, wherein the degree of association between any two of the feature points comprises a correlation coefficient or a euclidean distance between any two of the feature points.
9. An automobile competitive product model determining device is characterized by comprising:
an acquisition module: the system comprises a plurality of automobile relevant websites, a plurality of application programs and a plurality of characteristic indexes, wherein the automobile relevant websites are used for acquiring automobile basic data and user behavior data of each automobile type;
a weight determination module: the weight of each characteristic index is determined according to principal component analysis rules, and an array containing numerical values corresponding to the characteristic indexes is generated for each vehicle type;
a relevance determination module: the system comprises a plurality of arrays, a model calculation module, a model calculation module and a model matching module, wherein the arrays are used for determining a characteristic point on a corresponding coordinate system and calculating the association degree of any two characteristic points so as to construct a competitive model prediction model;
a vehicle type determination module: when a vehicle type is input, acquiring a target characteristic point of which the association degree of the characteristic point corresponding to the vehicle type is within a preset range according to the competitive product vehicle type prediction model, and determining a corresponding competitive product vehicle type according to the target characteristic point;
wherein the principal component analysis comprises:
obtaining a standardized array according to at least one characteristic index of each vehicle type;
calculating a correlation coefficient matrix according to the normalized matrix;
calculating a characteristic root of the correlation coefficient matrix, and determining a corresponding unit characteristic vector aiming at the characteristic root meeting the condition;
the normalized index variables are converted into principal components.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710075961.2A CN107545457B (en) | 2017-02-13 | 2017-02-13 | Automobile racing product type determination method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710075961.2A CN107545457B (en) | 2017-02-13 | 2017-02-13 | Automobile racing product type determination method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107545457A CN107545457A (en) | 2018-01-05 |
| CN107545457B true CN107545457B (en) | 2020-01-21 |
Family
ID=60966194
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710075961.2A Active CN107545457B (en) | 2017-02-13 | 2017-02-13 | Automobile racing product type determination method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107545457B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537596B (en) * | 2018-04-16 | 2020-02-21 | 北京车慧科技有限公司 | Method, device and system for recommending vehicle type in search box and memory |
| CN108920178B (en) * | 2018-06-29 | 2021-11-23 | 北京汽车股份有限公司 | Method and device for adjusting vehicle type carrying configuration |
| CN110443290B (en) * | 2019-07-23 | 2022-02-18 | 广东数鼎科技有限公司 | Product competition relationship quantitative generation method and device based on big data |
| CN111597199A (en) * | 2020-05-18 | 2020-08-28 | 泰康保险集团股份有限公司 | Vehicle type query method, device, equipment and computer readable storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886432A (en) * | 2014-04-04 | 2014-06-25 | 天津市市政工程设计研究院 | Wharf function adjustment compressive evaluation method based on weighting principal component analysis |
| CN104809877B (en) * | 2015-05-14 | 2017-09-08 | 重庆大学 | The highway place traffic state estimation method of feature based parameter weighting GEFCM algorithms |
| CN104978411B (en) * | 2015-06-23 | 2019-03-22 | 中车青岛四方机车车辆股份有限公司 | A kind of automobile development method and apparatus of bullet train |
-
2017
- 2017-02-13 CN CN201710075961.2A patent/CN107545457B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107545457A (en) | 2018-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10354309B2 (en) | Methods and systems for selecting an optimized scoring function for use in ranking item listings presented in search results | |
| TWI539305B (en) | Personalized information push method and device | |
| Li et al. | A MIDAS modelling framework for Chinese inflation index forecast incorporating Google search data | |
| US9672554B2 (en) | Methods and systems for deriving a score with which item listings are ordered when presented in search results | |
| CN106251174A (en) | Information recommendation method and device | |
| CN107545457B (en) | Automobile racing product type determination method and device | |
| US8738436B2 (en) | Click through rate prediction system and method | |
| CN108921398B (en) | Shop quality evaluation method and device | |
| CN105426514A (en) | Personalized mobile APP recommendation method | |
| CN104391849A (en) | Collaborative filtering recommendation method for integrating time contextual information | |
| WO2011035298A2 (en) | Methods and apparatus to perform choice modeling with substitutability data | |
| CN106296257A (en) | A kind of fixation of advertisement position put-on method based on user behavior analysis and system | |
| Wijnhoven et al. | Sentiment analysis and Google trends data for predicting car sales | |
| US10140339B2 (en) | Methods and systems for simulating a search to generate an optimized scoring function | |
| WO2009014753A2 (en) | Method for scoring products, services, institutions, and other items | |
| Li et al. | Collaborative filtering based on user attributes and user ratings for restaurant recommendation | |
| KR20140053444A (en) | Apparatus for forecasting market volume, method of forecasting market volume, and storage medium for storing a program forecasting market volume | |
| CN113065062A (en) | News recommendation method and system based on user reading time behavior | |
| CN107292713A (en) | A kind of rule-based individual character merged with level recommends method | |
| CN112487283A (en) | Method and device for training model, electronic equipment and readable storage medium | |
| CN113407854A (en) | Application recommendation method, device and equipment and computer readable storage medium | |
| Jianjun | Research on collaborative filtering recommendation algorithm based on user behavior characteristics | |
| Zhao et al. | Online comments of multi-category commodities based on emotional tendency analysis | |
| CN113761084B (en) | POI search ranking model training method, ranking device, method and medium | |
| KR20150144916A (en) | system and method providing a suited shopping information by customer profiling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 100044 Tengda Building, 168 Xizhimenwai Street, Haidian District, Beijing, 2101-2103 and 2105-2111 on the 21st floor Applicant after: Beijing Chehui Technology Co., Ltd. Address before: 100044 01-07, 29 story, Tengda tower, 168 west gate, Haidian District, Beijing. Applicant before: Beijing Che Hui Interactive Advertising Co., Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |