WO2020062657A1 - 用户风险画像生成方法、装置、设备及相关设备 - Google Patents

用户风险画像生成方法、装置、设备及相关设备 Download PDF

Info

Publication number
WO2020062657A1
WO2020062657A1 PCT/CN2018/123218 CN2018123218W WO2020062657A1 WO 2020062657 A1 WO2020062657 A1 WO 2020062657A1 CN 2018123218 W CN2018123218 W CN 2018123218W WO 2020062657 A1 WO2020062657 A1 WO 2020062657A1
Authority
WO
WIPO (PCT)
Prior art keywords
risk
poi
user
type
probability
Prior art date
Application number
PCT/CN2018/123218
Other languages
English (en)
French (fr)
Inventor
张海文
蔡健
Original Assignee
深圳壹账通智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳壹账通智能科技有限公司 filed Critical 深圳壹账通智能科技有限公司
Publication of WO2020062657A1 publication Critical patent/WO2020062657A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0202—Market predictions or forecasting for commercial activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Definitions

  • the present application mainly relates to the field of information technology, and in particular, to a method, an apparatus, a device, and a related device for generating a user risk profile.
  • the risk data reflects the possible risks of the user, and recommends insurance product information corresponding to the risk to the user to meet the user's demand.
  • This recommendation of insurance product information that meets user needs has a great correlation with the accuracy of user risk data, so that the accuracy of the determined user risk data occupies a very important position in the recommendation process of insurance product information.
  • the user risk data determined in the prior art cannot accurately reflect the user's risks in the process of use.
  • the recommended insurance product information based on the risk cannot meet the user's needs on the one hand, and it makes the financial institution insurance The success rate of product promotion is low. How to accurately reflect the specific risks of users is a technical problem that needs to be solved in the existing technology.
  • the main purpose of this application is to provide a method, device, equipment and related equipment for generating a user risk profile, which aims to solve the problem that the risks of the user cannot be accurately reflected in the prior art.
  • the present application provides a user risk profile generation method.
  • the POI type of interest includes a risk type POI.
  • the user risk profile generation method includes the following steps:
  • Probability product pairs are formed for each of the sub-risk POI probabilities and corresponding insurance products, and each probabilistic product pair is set as a user feature label and added to a preset portrait template to generate a risk portrait of the user in the preset period .
  • the present application also proposes a user risk profile generation device.
  • the POI type of interest points includes a risk type POI.
  • the user risk profile generation device includes:
  • a reading module configured to read the POI probability of the user going to each POI within a preset period, and statistics the POI probability according to the type of the POI to which the POI belongs, and determine the total risk POI probability of the user going to all risk types POI;
  • a determining module configured to determine whether the total risk POI probability is greater than a preset value, and if greater than the preset value, classify the attributes of insurance products for each of the risk type POIs, and determine whether the user is relatively The sub-risk POI probability of insurance products;
  • a generating module configured to form a probability product pair for each of the sub-risk POI probabilities and corresponding insurance products, and set each of the probability product pairs as a user feature label and add it to a preset portrait template to generate a user Set cycle risk portraits.
  • the present application also proposes a user risk profile generation device, where the user risk profile generation device includes: a memory, a processor, a communication bus, and computer-readable instructions stored on the memory;
  • the communication bus is used to implement connection and communication between the processor and the memory
  • the processor is configured to execute the computer-readable instructions to implement the steps of the user risk profile generating method.
  • the present application further provides a non-volatile readable storage medium, where the non-volatile readable storage medium stores one or more programs, and the one or more programs may be stored by one Or, when more than one processor executes, the steps of the method for generating a user risk profile are implemented.
  • the user risk profile generation method of this embodiment first determines the total risk POI probability of a user going to all risk types POI according to the POI probability of the user going to each POI within a preset period; this total risk POI probability reflects the user's daily life The risk situation involved, where the greater the probability of the total risk POI, the higher the risk involved in the daily life of the user; when it is judged that the probability of the total risk POI is greater than the preset value, the user is involved in the daily life The risk is higher, and the success rate of promoting insurance products to this user is higher according to the type of risk involved; thus further classifying the attributes of insurance products for each type of risk POI to determine the sub-risk POI of the user relative to various types of insurance products Probability, this sub-risk POI probability indicates the user's demand for various types of insurance products; furthermore, each sub-risk POI probability and its corresponding insurance product form multiple probability product pairs, and each probability product pair is added as a user feature label to In the preset portrait template, a risk
  • the POI probability of each sub-risk is determined by the POI probability of a risk type POI with the same insurance product attributes, and each risk type POI reflects the various risks involved in the daily life of the user, making the sub-risk POI probability and the user's various types Risk correlation; making the generated user risk portrait accurately reflect the risk situation of the user in daily life; avoiding the use of pure user risk data to reflect the user's risk, and achieving the accuracy of the user's risk reflect.
  • FIG. 1 is a schematic flowchart of a first embodiment of a user risk profile generating method of the present application
  • FIG. 2 is a functional module schematic diagram of a first embodiment of a user risk profile generating device of the present application
  • FIG. 3 is a schematic diagram of a device structure of a hardware operating environment involved in a method according to an embodiment of the present application.
  • This application provides a user risk profile generation method.
  • FIG. 1 is a schematic flowchart of a first embodiment of a method for generating a risk profile of a user of this application.
  • the POI type of the interest point includes a risk type POI
  • the user risk profile generation method includes:
  • Step S10 Read the POI probability of the user going to each POI within a preset period, and calculate the POI probability of each POI according to the type of POI to which the POI belongs, to determine the total risk POI probability of the user going to all risk types POI;
  • the method for generating a user risk profile of the present application is applied to a server, and is suitable for generating a risk profile representing a risk that a user may have in daily life through the server.
  • the server generates a user to go to each POI (Point in a preset period) in advance.
  • POI Point in a preset period
  • POI is a geographic object that can be abstracted as a point, especially some geographical entities that are closely related to people's lives, such as schools, banks, restaurants, gas stations, hospitals and supermarkets. Users going to different POIs reflect different needs of users. Among them, going to some POIs can reflect the risks that users have.
  • users who go to equestrian parks, rock climbing grounds, airports, and other POIs more frequently may have higher accident risks.
  • Users with more POI visits to hospitals, clinics, and health care stores may have higher health risks; users with more POI visits to car repair shops, gas stations, etc. may have higher vehicle insurance; and
  • the user's demand for each POI and the risk they have can be determined by the user's POI probability of going to each POI and the POI probability of going to a risk-related POI.
  • the server detects the user's location information within a preset period, determines the POI that exists around the location information, and then generates the POI probability that the user will go to each POI within the preset period; among them, the greater the POI probability of removing certain POIs related to risk , The greater the risk the user has.
  • the preset period is a period of time set in advance according to requirements, such as days, weeks, months, and years. The longer the preset period is, the more accurately the generated POI probability can reflect the user's needs.
  • each POI may belong to the same type of POI or different types of POI; for example, two banks located in different locations belong to the same type of POI, The gyms and combat clubs located in different locations belong to different types of POIs.
  • users have different POI requirements, they also have different needs for different types of POIs, which makes the POI probability of users going to various types of POIs different.
  • the user may go to a location with such a POI more times, and the probability of such a POI generated is greater. Take the risk-related POI as the risk type POI, and its POI probability is related to user needs.
  • the risk type POI is a characterization user such as an equestrian field, a climbing field, an airport, a hospital, a clinic, a health care store, a car repair shop, and a gas station POI of possible risks.
  • a characterization user such as an equestrian field, a climbing field, an airport, a hospital, a clinic, a health care store, a car repair shop, and a gas station POI of possible risks.
  • the statistics of each POI probability according to the type of POI to which it belongs, and the step of determining the total risk POI probability of the user going to all risk types POI includes:
  • Step S11 Read the identifier carried in each of the POI probabilities, and filter out the risk identifier belonging to the risk type from each of the identifiers;
  • the identifier f1 is set for public type POIs such as libraries and parks, and scenery types such as mountains and rivers
  • the POI sets an identifier f2, and sets the identifier f3 to a risk type POI such as a hospital, an airport, or a car repair shop.
  • an identifier corresponding to the type is added to its POI probability according to the type of each POI.
  • this POI Probability a adds identifier f1.
  • the identifiers carried in each probability are further read, and then the risk identifiers classified as risk types are screened from each identifier. Because the risk identifier that represents the type of risk is preset, compare the read identifier with this preset risk identifier to determine whether the two are consistent.
  • the read identifier is Risk identifiers, and filter them out of each identifier; if they are not the same, it means that the identifier read is not a risk identifier, and it is filtered out. After all the read identifiers are judged with the risk identifier, the selected identifier is the risk identifier belonging to the risk type in the identifier.
  • step S12 the POI probabilities with the risk identifiers are accumulated to obtain a total risk POI probability of the user going to all risk types POI.
  • the POI probability that carries the selected risk type identifier is the probability that the user will remove the risk type POI within a preset period; thus, the POI probability with this risk type identifier is accumulated to accumulate all The obtained result is the total risk POI probability that the user will go to all risk type POIs within a preset period.
  • step S20 it is judged whether the total risk POI probability is greater than a preset value, and if it is greater than the preset value, the attribute classification of the insurance product is performed for each of the risk type POIs, and the user is determined relative to each type of the insurance.
  • Product's sub-risk POI probability
  • a preset value is set in advance, and the preset value is set according to historical data, such as when multiple users are included in historical data
  • the probability of the total risk POI is greater than a certain value, the proportion of the risk type POI in the lives of the multiple users is greater, and the risks of the multiple users are greater; and using this value as a preset value, In order to determine the risk of other users through preset values in the future.
  • the total risk POI probability After generating the total risk POI probability of the user in a preset period through accumulation, compare the total risk POI probability with a preset value to determine whether the total risk POI probability is greater than the preset value; if it is determined that the total risk POI probability is not greater than this
  • the preset value indicates that the user has a small risk; if a risk situation is generated for this user to characterize the user's risk, and it is used by financial institutions to promote insurance product information based on the risk profile; due to the risk inherent in the user, It makes the reference of risk profile not significant, which will lead to invalid promotion of financial institutions; therefore, risk profiles are not generated for users whose total risk POI probability is less than or equal to a preset value.
  • the total risk POI probability is compared with a preset value and it is judged that the total risk POI probability is greater than the preset value, it means that the user has a greater risk, and a risk portrait is generated for it.
  • different types of risk correspond to different types of risk.
  • POIs such as equestrian grounds, climbing grounds, and airports correspond to accidental risks
  • POIs such as hospitals, clinics, and health care stores correspond to health risks
  • vehicle maintenance POIs such as stores and gas stations correspond to vehicle insurance.
  • Different types of risk correspond to different types of insurance products.
  • the type of insurance product corresponding to accidental risk is accident insurance
  • the type of insurance product corresponding to health risk is health insurance
  • the type of insurance product corresponding to vehicle insurance is car insurance.
  • This different type of insurance product has different type attributes to characterize the type of risk to which it applies.
  • the type attribute of health insurance is health, which characterizes the risk to which it applies to health.
  • the attributes of the insurance product are classified for each type of risk POI to which the user goes in a preset period, and each type of risk POI is attributed to the insurance product with each type of attribute. Because insurance products with different types of attributes correspond to different types of risks, users will be classified into a type of risk type POI corresponding to the same type of insurance products; for example, if they go to clinics and medical institutions, the clinic and type attributes are healthy. Insurance products correspond, and medical institutions also correspond to insurance products whose type attributes are healthy; the type attributes of the corresponding insurance products are the same, and clinics and medical institutions are classified as a type of risk type POI.
  • the user's risk level in this type of risk POI can be determined.
  • the steps of classifying the attributes of insurance products for each risk type POI and determining the user's sub-risk POI probability relative to various types of insurance products include:
  • Step S21 Invoke the type attribute tag of each type of the insurance product, and compare each of the risk identifiers with each of the type attribute tags one by one to determine a target type attribute tag corresponding to each of the risk identifiers;
  • the type attributes of different types of insurance products are set in advance to have For different type attribute labels, for example, if the type attribute is a health insurance product, the applicable risk type is health risk, and the corresponding type attribute label is g3.
  • the type attribute labels of each insurance product are first called, and then the risk identifiers that belong to the risk type are compared with the attribute labels of each type one by one. It should be noted that although the risk identifier is read from the POI probability, the POI probability carrying this risk identifier is possessed by the risk type POI.
  • the POI probability, the risk type POI, and the risk type identifier have correspondence and can be
  • the specific risk type of the risk type POI is characterized by the risk identifier. Specifically, different characters can be added after the risk identifier to characterize the specific risk type of the risk type POI. For example, for the identifier f3 that represents the risk type POI, the risk identifier is f3, and a different one can be added after this f3.
  • Character to characterize the specific risk type of the risk type POI for example, f3-31 is used to characterize hospitals and health care stores to indicate that the specific risk type of the risk type POI with the risk identifier f3-31 is health risk; it is characterized by f3-32 For racetracks and rock climbing fields, the specific risk type of the risk type POI with the risk identifier f3-32 is accidental risk.
  • type attribute labels In order to characterize the consistency of risk types, there is a corresponding relationship between type attribute labels and risk identifiers, such as g3 and f3-31, which characterize health risks, as described above; thus, when comparing each risk identifier with each type of attribute label, when When it is determined that there is a corresponding relationship between the risk identifier currently being compared with a certain type of attribute label, it can be determined that this type of attribute label corresponds to the current risk identifier; this type of attribute label is used as the target type attribute label to indicate that it has the current risk.
  • the specific risk type of the identifier's risk type POI is consistent with the risk type applicable to insurance products with this target type attribute label. All risk type POIs with risk identifiers consistent with this target type attribute label can be classified as the same class.
  • Step S22 classify each of the risk type POIs with the same target type attribute tag to form a set of risk type POIs corresponding to various types of the insurance products;
  • the target type attribute labels corresponding to each risk identifier can be determined; and then the POI of each risk type is classified according to the target type attribute labels. , Classify each risk type POI with the same target type attribute label. It should be noted that because the type attribute tag and the risk identifier have a corresponding relationship, the target type attribute tag is a type attribute tag corresponding to the risk identifier, so that the risk identifier carried by the risk type POI can be characterized by the corresponding target type attribute tag .
  • a risk type POI carries a risk identifier, and this risk identifier corresponds to a target type attribute tag, and this target type attribute tag can be used to characterize the risk type identifier carried in the risk type POI, which can be described as this risk Type POI has this target type attribute tag.
  • the risk type POIs with the same target type attribute labels are classified to form a set of risk type POIs corresponding to various types of insurance products, and the corresponding relationship with various types of insurance products is determined by the corresponding target type attribute labels and insurance products. Relationship decision.
  • the risk type POI for users includes M1, M2, M3 ,, M4, and M5, respectively, where the target type attribute label of M1, M2, and M4 is g2, and the target type attribute label of M3 and M5 is g3; Classify M1, M2, and M4 to form a POI set of risk types of [M1, M2, M4]; classify M3 and M5 to form a POI set of [M3, M5] risk types; due to g2 and accident insurance types
  • the products correspond to, and g3 corresponds to health insurance products, so that [M1, M2, M4] corresponds to accident insurance products, and [M3, M5] corresponds to health insurance products.
  • step S23 the POI probabilities of each of the risk type POIs in each of the risk type POI sets are accumulated to generate a sub-risk POI probability of the user with respect to each type of the insurance product.
  • each risk type POI has a different POI probability; the POIs of all risk type POIs included in this risk type POI set The probabilities are accumulated one by one. When all the risk type POI sets are accumulated, the obtained result is the overall POI probability of each risk type POI set, which represents the user's risk level in various types of risks.
  • the overall POI probability is essentially the user's sub-risk probability relative to various types of insurance products, which represents the user's demand for various types of insurance products; when a user has a certain type of risk POI When the probability of a set of sub-risk POIs is greater, it indicates that the user has a greater risk of the type of risk represented by this type of risk POI set, and further indicates that the user has greater demand for insurance products suitable for such risks;
  • the sub-risk POI probability of the insurance products determines the user's demand for various insurance products.
  • Step S30 forming a probability product pair for each of the sub-risk POI probabilities and corresponding insurance products, and setting each probability product pair as a user feature tag to add to a preset portrait template to generate a user in the preset period Risk portrait.
  • the POI probability and The corresponding insurance products form a probability product pair.
  • the probability product pair reflects the user's risk level on various types of risks, and the type of insurance product corresponding to this type of risk.
  • a preset portrait template for generating a risk portrait is set in advance, and various types of feature label variables of the user are used in the preset portrait template to represent various types of characteristics of the user; each probability product pair formed is used as a user characteristic label Add it to the preset portrait template, replace the feature label variables therein, and use the preset portrait template after the addition and replacement operation as the user's risk portrait in a preset period.
  • the risk profile the risk distribution of users in daily life is displayed intuitively, and the risk profile is used as a recommendation basis for insurance products, so that the recommended insurance products are more in line with the actual situation of users.
  • the user risk profile generation method of this embodiment first determines the total risk POI probability of a user going to all risk types POI according to the POI probability of the user going to each POI within a preset period; this total risk POI probability reflects the user's daily life The risk situation involved, where the greater the probability of the total risk POI, the higher the risk involved in the daily life of the user; when it is judged that the probability of the total risk POI is greater than the preset value, the user is involved in the daily life The risk is higher, and the success rate of promoting insurance products to this user is higher according to the type of risk involved; thus further classifying the attributes of insurance products for each type of risk POI to determine the sub-risk POI of the user relative to various types of insurance products Probability, this sub-risk POI probability indicates the user's demand for various types of insurance products; furthermore, each sub-risk POI probability and its corresponding insurance product form multiple probability product pairs, and each probability product pair is added as a user feature label to In the preset portrait template, a risk
  • the POI probability of each sub-risk is determined by the POI probability of a risk type POI with the same insurance product attributes, and each risk type POI reflects the various risks involved in the daily life of the user, making the sub-risk POI probability and the user's various types Risk correlation; making the generated user risk portrait accurately reflect the risk situation of the user in daily life; avoiding the use of pure user risk data to reflect the user's risk, and achieving the accuracy of the user's risk reflect.
  • a second embodiment of the user risk profile generation method of the present application is proposed.
  • the user is generated in the preset period.
  • the risk profile steps include:
  • Step S40 a ratio of each of the sub-risk POI probability and the total risk POI probability is generated to generate a user's risk ratio with respect to each type of the insurance product;
  • the user's risk is different in different types of risks, and the risk can be reflected by the sum of the POI probabilities of the classified risk types POI in the total risk POI probability of the entire risk type POI, that is, The proportion of each sub-risk POI probability in the total risk POI probability is reflected; the total risk POI probability represents the total risk of the user in the preset period, and each sub-risk POI probability represents the user's various risks in the preset period.
  • the magnitude of the risk when the proportion of the POI probability of a certain sub-risk is greater, it means that the sub-risk POI probability occupies a larger value in the total risk POI probability, and the main source of the user's risk in a preset period is the same as this sub-risk POI
  • the type of risk corresponding to the probability Specifically, the ratio of each sub-risk POI probability and the total risk POI probability is taken as a ratio, and the magnitude of the ratio is used to represent how much the risk type corresponding to each sub-risk POI probability occupies the user's total risk in daily life.
  • the types of insurance products that are applicable to different types of risk are different.
  • the corresponding risk types occupy the total risk of the user in daily life, which can indicate the user's demand for various types of insurance products.
  • the risk types occupy the total risk in the daily life of the user, More, it means that users have greater demand for insurance products suitable for this type of risk. Therefore, the numerical result of the ratio is taken as the user's risk ratio to various types of insurance products, which characterizes the user's demand for various types of insurance products.
  • the ratio value result generated by the sub-risk POI probability is larger, the corresponding risk ratio is compared. Large, users have greater demand for insurance products applicable to this sub-risk POI probability corresponding to the type of risk.
  • Step S50 comparing each of the risk ratios, determining a target risk ratio with the largest value, and determining an insurance product type corresponding to the target risk ratio;
  • each risk ratio has a different numerical value, and the larger the value, the user has a higher risk for the type of risk applicable to the source of the insurance product. Greater risk.
  • the type of risk that the user has the greatest risk that is, the most important source of risk for the user in daily life; compare the proportions of each risk, determine the largest proportion of risk, and use this risk proportion as the target risk proportion; Because the risk proportions belong to a certain type of insurance product, the type of insurance product to which the target risk proportion belongs is taken as the type of insurance product corresponding to the target risk proportion to characterize the type of insurance product most likely to be required by the user.
  • Step S60 Acquire product information corresponding to the type of insurance product, and transmit the product information to a terminal held by a user.
  • various types of insurance products are suitable for different types of risks, and different types of risks have different characteristics; for these different characteristics, various types of insurance products have different product information.
  • health insurance for health-type risks and accident insurance for accident-type risks because of the differences in the characteristics of health-type risks and accident-type risks, product information differences exist between health insurance and accident insurance.
  • the product information of various types of insurance products is correspondingly stored in the storage unit of the financial institution's server, which can be retrieved from the storage.
  • the unit reads the product information corresponding to the identified insurance product type; assigns a consistent identifier to each insurance product and its product information, and after determining the insurance product type, reads the identifier of this insurance product type, and according to the identification Find the corresponding product information to read.
  • the read product information is transmitted to the terminal held by the user, which is convenient for the user to understand the product information of the insurance product applicable to the type of risk of the source of the greatest risk; the terminal held by the user may be the mobile phone, tablet, etc. used by the user There are no restrictions on this.
  • the step of obtaining product information corresponding to the type of insurance product includes:
  • Step S61 Read the basic user information, and read the product attributes of each insurance product corresponding to the insurance product type;
  • K1 and K2 both like outdoor sports, and both types of risks are accident insurance.
  • K1 is older and K2 is younger, which makes K2 have better physical strength and better risk coping ability.
  • the accidental risk of K2 is smaller than that of K1.
  • various insurance products for different types of risks include multiple insurance products; the overall types of risks targeted by multiple insurance products are the same.
  • product attributes due to this type of risk applicable to different users.
  • the insurance products X1 and X2 for users with K1 and K2 types of risks are all accidental types of risks.
  • X1 applies to younger users, while X2 applies to older users, making X1 applicable.
  • X2 applies to user K1.
  • This user's basic information includes the user's gender, age, city of residence, and nature of work
  • read the product attributes of each insurance product included in the insurance product type This product attribute characterizes the population to which each insurance product applies, such as the applicable age range, city, monthly income, etc.
  • Step S62 Determine a target insurance product according to the matching degree of the user basic information and each of the product attributes, and determine the product information of the target insurance product as product information corresponding to the type of the insurance product.
  • the basic information and product attributes are compared one by one to determine the degree of matching between the user's basic information and the product attributes; where the degree of matching is that the user's basic information meets each
  • the degree of product attribute requirements such as the applicable age in the product attribute is between 20 and 30, and the applicable city is Shenzhen, and when the basic information of the user is 45 years old, the user does not match this product attribute.
  • the basic user information meets the requirements of product attributes, the higher the degree of matching between the two, the more suitable an insurance product with this product attribute is for the user.
  • the method before the step of reading the POI probability of the user going to each POI within a preset period, the method includes:
  • Step q1 Read each position information of the user within a preset period, and determine a POI within a preset range centered on each of the position information;
  • each POI probability is generated according to the user's various location information in a preset period, the POIs around different location information are different, and the locations where users go at different times are different; thus generating users
  • the GPS (Global Positioning System (Global Positioning System) positioning service or base station positioning service obtains each location information of the user in a preset period, and determines a POI within a preset range centered on each location information according to the identifier.
  • the operation of reading position information and statistical POI can be a real-time operation in a preset period or a unified operation after the preset period is completed.
  • GPS or a base station locates the user's position information in real time, and Store the position information of the position; read the position information of the position in real time in a preset period to determine the POI within a preset range centered on the position information; for unified operation, it is completed in the time of the preset period Then, the location information to which the user goes in a preset period is uniformly read, and a POI within a preset range centered on each location information is determined.
  • the POIs that users go to are limited, and the POIs that are farther away from each location information, the less likely they are to go; therefore, based on multiple historical data, the user uses location information in advance Set this preset range for the distance range from the center to the POI, so that the preset range meets the travel needs of most people.
  • the POIs within this preset range of the location information are determined, and characterize each POI that the user may go to from the location corresponding to the location information.
  • Step q2 Count the number of POIs of the POI, and determine the POI probability of a user going to each of the POIs from each of the location information according to the number of POIs, and generate a POI probability of a user going to each POI within a preset period.
  • the user can go to the POI he needs according to his needs.
  • the probability of the user going to each POI is the same, and the sum of the total probability of going to each POI is 1. Therefore, for the POI probability of each POI, the number of POIs within a preset range is accumulated and counted, and the result of the statistics is the number of POIs, and then the POI probability of the user from each location information to each POI is determined according to the number of POIs. Because the probability that the user goes to each POI from each location is the same, and the total is 1.
  • the total number of POIs can be evenly divided by the sum 1, that is, divide the number of POIs by 1, and use the inverse value of the number of POIs to characterize the user.
  • the probability of each POI is the same, and each probability is an inverse value of the number of POIs, and the POI probability of the user from a certain position information of a preset period to each POI within a preset range of this position information is generated. If there are 5 POIs within a preset range of location information, the probability of going to each POI is 1/5. In this way, the POI probability of the user from each location information to each POI within a preset range is calculated until the POI probability of each POI within the preset range of all location information is calculated and determined.
  • the location information is stored at the POI probability of each POI within a preset range, so that the user's need to go to each POI is determined according to this POI probability.
  • different distances have an impact on the probability that the user will go from location information to each POI; where the POI farther from the location information, the user goes The smaller the probability, the lower the probability that a user will go to this POI; so that the POI probability that a user will go to each POI from location information is related to the distance between each POI and location information in addition to the number of POIs.
  • determining a POI probability that a user goes to each of the POIs from each of the location information, and generating a POI probability that a user goes to each POI within a preset period includes:
  • Step q21 divide the preset range into a plurality of near and far areas according to a distance relationship between each of the position information and the preset range;
  • the probability that the user goes to each POI is related to the distance between each POI and the location information; therefore, when determining the POI probability of each POI, the preset relationship is first set according to the distance relationship between each location information and a preset range.
  • the range is divided into multiple near and far areas, and the distances between the different near and far areas and the position information are different.
  • the position represented by the position information is used as a starting point, and the preset range is divided into a plurality of near and far areas every certain distance length; the distance length is a length set in advance through historical data, and the multiple near and far areas are within Areas where different starting points extend by this distance.
  • the preset range is a 1000m circular range with A as the center, and the distance is 500m; a circular area 500m forward from A as the starting point will be used as a near and far area interval, and The boundary of the circular section, that is, the boundary with a distance of 500m is the starting point, and the circle area between 500m and 1000m is used as another distance area.
  • Step q22 Assign a regional probability to each of the near and far areas according to the number of areas in the near and far areas;
  • the distance between the near and far area information is relatively large, and the probability of the user going to this near and far area is relatively large, while the probability of going to the near and far area with relatively long distance information is relatively small;
  • the distance relationship assigns different regional probabilities to each of the near and far regions, and assigns a larger regional probability to the near and far regions that are closer to the location information, and a smaller regional probability to the far and near regions that are closer to the location information.
  • the sum of the regional probability of a region is 1, and the different regional probabilities are used to represent the probability that the user will go to each of the near and far regions within a preset range.
  • Step q23 Use the product of the reciprocal value of the number of POIs and the area probability to determine the POI probability of the user going from each of the position information to the POI in each of the near and far areas, and generate a user to go to each within a preset period. POI probability of POI.
  • the POI probability of the user going to each POI within a preset range is determined. Specifically, the area probability of the near and far areas is multiplied by the reciprocal value of the number of POIs of the POIs within the preset range; the result of the multiplication is the POI probability of the user from the location information to each POI in each of the near and far areas, that is, The user goes from the location information to the POI probability of each POI within a preset range; in this way, the POI probability of the user from each location information to each POI in each of the far and near areas within the preset range is calculated until the preset range of all location information Each POI probability within the calculation is completed.
  • the determined POI probability is more accurate; after the calculation of the POIs in the preset range is completed, the calculated POI probability of each POI is stored, Generate the POI probability that the user goes to each POI in a preset period, so that the user's demand for each POI is accurately reflected according to this POI probability.
  • this application provides a user risk profile generation device.
  • POI types of interest include risk type POI.
  • the user risk profile generation device includes:
  • the reading module 10 is configured to read the POI probability of the user going to each POI within a preset period, and to calculate the POI probability of each POI according to the type of the POI to which the POI belongs, to determine the total risk POI probability of the user going to all risk types. ;
  • a determining module 20 is configured to determine whether the total risk POI probability is greater than a preset value, and if it is greater than the preset value, classify the attributes of insurance products for each of the risk type POIs, and determine the user relative to the type The sub-risk POI probability of the insurance product;
  • a generating module 30 is configured to form a probability product pair for each of the sub-risk POI probabilities and corresponding insurance products, and set each of the probability product pairs as a user feature tag and add it to a preset portrait template to generate a user Risk profile for a preset period.
  • Each of the virtual function modules of the user risk profile generating device is stored in the memory 1005 of the user risk profile generating device shown in FIG. 3.
  • the processor 1001 executes computer-readable instructions, each module in the embodiment shown in FIG. 2 is implemented.
  • the storage medium may be a read-only memory, a magnetic disk, or an optical disk.
  • FIG. 3 is a schematic structural diagram of a device related to a hardware operating environment according to a method according to an embodiment of the present application.
  • the user risk profile generating device may be a PC (personal computer, personal computer). ), Or a terminal device such as a smart phone, tablet, e-book reader, or portable computer.
  • the user risk profile generation device may include: a processor 1001, such as a CPU (Central Processing Unit (central processing unit), memory 1005, and communication bus 1002.
  • the communication bus 1002 is used to implement connection and communication between the processor 1001 and the memory 1005.
  • the memory 1005 may be a high-speed RAM (random access memory (random access memory) or non-volatile memory memory), such as disk storage.
  • the memory 1005 may optionally be a storage device independent of the foregoing processor 1001.
  • FIG. 3 does not constitute a limitation on the user risk profile generating device.
  • the memory 1005 as a non-volatile readable storage medium may include an operating system, a network communication module, and computer-readable instructions.
  • the operating system is a program that manages and controls the hardware and software resources of the user risk profile generation device, and supports the operation of computer-readable instructions and other software and / or programs.
  • the network communication module is used to implement communication between components in the memory 1005 and to communicate with other hardware and software in the user risk profile generation device.
  • the processor 1001 is configured to execute computer-readable instructions stored in the memory 1005 to implement steps in the embodiments of the above-mentioned user risk profile generating method.
  • the present application provides a non-volatile readable storage medium.
  • the non-volatile readable storage medium is preferably a computer-shortable storage medium, and the computer non-volatile readable storage medium stores one or more than one.
  • a program, the one or more programs may also be executed by one or more processors for implementing steps in the embodiments of the user risk profile generation method described above.

Landscapes

  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Data Mining & Analysis (AREA)
  • Game Theory and Decision Science (AREA)
  • Technology Law (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)

Abstract

本申请公开一种用户风险画像生成方法、装置、设备及相关设备，所述方法包括：读取用户在预设周期内去各POI的POI概率，并对各POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；判断总风险POI概率是否大于预设值，若大于预设值，则对各风险类型POI进行保险产品的属性归类，并确定用户相对于各类保险产品的子风险POI概率；将各子风险POI概率及对应的保险产品形成概率产品对，并将各概率产品对设为用户特征标签添加到预设画像模板中，生成用户在预设周期的风险画像。本方案基于用户行为的大数据分析，生成用户风险画像，来实现对用户日常生成中所涉及到的风险情况的精准反映。

Description

用户风险画像生成方法、装置、设备及相关设备
本申请要求于2018年09月27日提交中国专利局、申请号为201811135650.1、发明名称为“用户风险画像生成方法、装置、设备及可读存储介质”的中国专利申请的优先权，其全部内容通过引用结合在申请中。
技术领域
本申请主要涉及信息技术领域，具体地说，涉及一种用户风险画像生成方法、装置、设备及相关设备。
背景技术
目前各金融机构在进行保险产品营销时，为了实现精准营销，通常结合用户的风险数据进行，由风险数据反映用户可能存在的风险，而向用户推荐与风险对应的保险产品信息，以满足用户的需求。此满足用户需求的保险产品信息的推荐，与用户风险数据的精准性具有很大的相关性，使得所确定的用户风险数据的精准性在保险产品信息推荐过程中占有非常重要的地位。现有技术中所确定的用户风险数据，在使用过程中并不能精准的体现用户所具有的风险，而导致依据风险所推荐的保险产品信息一方面不能满足用户需求，另一方面使得金融机构保险产品推广的成功率低，如何精准的体现用户所具体的风险是现有技术中亟待解决的技术问题。
发明内容
本申请的主要目的是提供一种用户风险画像生成方法、装置、设备及相关设备，旨在解决现有技术中不能精准的反映用户所具有风险的问题。
为实现上述目的，本申请提供一种用户风险画像生成方法，兴趣点POI类型包括风险类型POI，所述用户风险画像生成方法包括以下步骤：
读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
此外，为实现上述目的，本申请还提出一种用户风险画像生成装置，兴趣点POI类型包括风险类型POI，所述用户风险画像生成装置包括：
读取模块，用于读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
确定模块，用于判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
生成模块，用于将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
此外，为实现上述目的，本申请还提出一种用户风险画像生成设备，所述用户风险画像生成设备包括：存储器、处理器、通信总线以及存储在所述存储器上的计算机可读指令；
所述通信总线用于实现处理器和存储器之间的连接通信；
所述处理器用于执行所述计算机可读指令，以实现上述用户风险画像生成方法的步骤。
此外，为实现上述目的，本申请还提供一种非易失性可读存储介质，所述非易失性可读存储介质存储有一个或者一个以上程序，所述一个或者一个以上程序可被一个或者一个以上的处理器执行时，以实现上述用户风险画像生成方法的步骤。
本实施例的用户风险画像生成方法，先根据用户在预设周期内去各个POI的POI概率，确定用户去所有风险类型POI的总风险POI概率；此总风险POI概率反映了用户日常生活中所涉及到的风险情况，其中总风险POI概率越大，说明用户日常生活中所涉及到的风险越高；当判断出此总风险POI概率大于预设值时，说明用户日常生活中所涉及到的风险较高，依据所涉及到的风险类型向此用户推广保险产品的成功率较高；从而进一步对各风险类型POI进行保险产品的属性归类，确定用户相对于各类保险产品的子风险POI概率，此子风险POI概率表明用户对各类保险产品的需求情况；进而将各子风险POI概率与其对应的保险产品形成多个概率产品对，将该各概率产品对设为用户特征标签添加到预设画像模板中，生成用户在预设周期的风险画像。因各子风险POI概率由具有相同保险产品属性的风险类型POI的POI概率确定，而各风险类型POI体现了用户日常生活中所涉及到的各类风险，使得子风险POI概率与用户的各类风险相关；使得所生成的用户风险画像，可准确反映用户在日常生活中所涉及到的风险情况；避免使用单纯的用户风险数据来反映用户所具有的风险，实现了对用户所具有风险的精准反映。
附图说明
图1是本申请的用户风险画像生成方法第一实施例的流程示意图；
图2是本申请的用户风险画像生成装置第一实施例的功能模块示意图；
图3是本申请实施例方法涉及的硬件运行环境的设备结构示意图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
具体实施方式
应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
本申请提供一种用户风险画像生成方法。
请参照图1，图1为本申请用户风险画像生成方法第一实施例的流程示意图。在本实施例中，兴趣点POI类型包括风险类型POI，所述用户风险画像生成方法包括：
步骤S10，读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
本申请的用户风险画像生成方法应用于服务器，适用于通过服务器生成表征用户日常生活中可能所具有风险的风险画像。具体地，服务器中预先生成有用户在预设周期内去各个POI（Point of Interest，兴趣点）的POI概率，POI是可以抽象为点的地理对象，尤其是一些与人们生活密切相关的地理实体，如学校、银行、餐馆、加油站、医院和超市等。用户去不同的POI反映了用户的不同需求，其中去某些POI可反映用户所具有的风险，如去马术场、攀岩场、机场等POI次数较多的用户，可能具有较高的意外风险；而去医院、诊所、医疗保健店等POI次数较多的用户，可能具有较高的健康风险；而去汽车维修店、加油站等POI次数较多的用户，可能具有较高的车辆险；且用户对各POI的需求大小，以及所具有的风险大小可由用户去各POI的POI概率，以及去与风险相关POI的POI概率决定。服务器对用户在预设周期内的位置信息检测，确定位置信息周围所具有的POI，进而生成用户在预设周期内去各个POI的POI概率；其中去某些与风险相关POI的POI概率越大，用户所具有的风险越大。
需要说明的是，对于同一个POI，因用户在预设周期内所去到的时间不同，对应此不同的时间需要生成不同的POI概率；如对于某一POI，用户分别在预设周期中某天的上午和下午均去到此POI，则需要针对上午和下午的时间点分别生成两个POI概率；即各个POI的POI概率不具有叠加性，而与预设周期中所去的不同时间相关。其中预设周期为根据需求所预先设置的一段时间范围，如天、周、月、年等，预设周期的时间越长，生成的POI概率越能准确的反映用户需求，本实施优选“年”作为预设周期，以“年”为例进行说明；即在此以年作为预设周期的时间范围内实时检测用户的位置信息，通过一年内用户的各个位置信息，生成反映用户需求的各个POI概率，并通过与风险相关POI的POI概率反映用户的风险大小。
在生成反映用户预设周期内对各个POI需求的POI概率后，因各个POI可能属于同一类型的POI，也可能属于不同类型的POI；如位于不同位置的两个银行则属于同一类型的POI，而位于不同位置的健身房和搏击俱乐部则属于不同类型的POI；用户对各个POI需求不一样的同时，对不同类型POI的需求也不一样，而使得用户去各类POI的POI概率大小不一样；对于用户需求高的POI类型，可能用户去具有此类POI的位置的次数较多，所生成的此类POI概率较大。将与风险相关POI作为风险类型POI，其POI概率相应的与用户需求相关，其中风险类型POI为马术场、攀岩场、机场、医院、诊所、医疗保健店、汽车维修店、加油站等表征用户可能所存在的风险的POI。为了表征用户风险大小，在生成用户在预设周期内所去到各个POI的POI概率后，需要从此各个POI的POI概率中，确定属于风险类型POI的所有POI概率，进而根据此所有POI概率判定用户在风险类型POI中的风险大小。具体地，先读取用户预设周期内去各个POI的POI概率，再对此各个POI概率按照所归属的POI类型进行统计，即统计归属为风险类型POI的POI概率，以确定用户在预设周期内去所有风险类型POI的总风险POI概率，表征用户在预设周期内所具有的风险大小。具体地，对各POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率的步骤包括：
步骤S11，读取各所述POI概率中所携带的标识符，并从各所述标识符中筛选出归属为风险类型的风险标识符；
考虑到市面上的POI类型众多，为了对不同类型POI进行区分，对不同类型POI预先设置不同的标识符，如对图书馆、公园等公共类型POI设置标识符f1，对山川、河流等风景类型POI设置标识符f2，对医院、机场、汽车维修店等风险类型POI设置标识符f3。在生成用户去各个POI的POI概率时，根据各个POI的类型，为其POI概率添加与类型对应的标识符，如当用户在预设周期内去图书馆的POI概率为a，则对此POI概率a添加标识符f1。在读取到用户在预设周期内去各POI的POI概率后，进一步读取各个概率中所携带的标识符，再从各个标识符中筛选出归属为风险类型的风险标识符。因表征风险类型的风险标识符为预先设定的，将读取的标识符和此预先设定的风险标识符对比，判断两者是否一致；若两者一致，则说明读取的标识符为风险标识符，而将其从各标识符中筛选出来；若两者不一致，则说明读取的标识符不是风险标识符，而将其过滤掉。在所有读取的标识符均和风险标识符进行判断后，筛选出来的标识符为标识符中归属于风险类型的风险标识符。
步骤S12，将具有所述风险标识符的所述POI概率进行累加，得到用户去所有风险类型POI的总风险POI概率。
可理解地，携带有所筛选出来的风险类型标识符的POI概率，即为用户在预设周期内去风险类型POI的概率；从而对具有此风险类型标识符的POI概率进行累加操作，累加所得到的结果即为用户在预设周期内去所有风险类型POI的总风险POI概率。
步骤S20，判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
进一步地，为了表征用户在预设周期去所有风险类型POI的总风险POI概率的大小，预先设置有预设值，此预设值为根据历史数据进行设定，如历史数据中当多个用户的总风险POI概率大于某一值时，风险类型POI在此多个用户生活中所占的比重较大，此多个用户所具有的风险较大；而将此某一值作为预设值，以便于后续通过预设值判定其他用户的风险大小。在通过累加生成用户在预设周期的总风险POI概率后，将此总风险POI概率和预设值对比，判断总风险POI概率是否大于此预设值；若判断出总风险POI概率不大于此预设值，则说明用户所具有的风险不大；如果针对此用户生成用于表征用户风险的风险情况，而供各金融机构根据风险画像进行保险产品信息推广；因用户本身所具有的风险，使得风险画像的参考意义不大，会导致金融机构的无效推广；从而不针对总风险POI概率小于或等于预设值的用户生成风险画像。而当将总风险POI概率和预设值对比，判断总风险POI概率大于预设值时，则说明用户具有较大的风险，而针对其生成风险画像。
可理解地，不同的风险类型POI所对应的风险类型不一样，如马术场、攀岩场、机场等POI与意外风险对应；而医院、诊所、医疗保健店等POI与健康风险对应；而汽车维修店、加油站等POI与车辆险对应。而不同类型的风险又对应不同类型的保险产品，如意外风险对应的保险产品类型为意外险、健康风险对应的保险产品类型为健康险、而车辆险对应的保险产品类型为车险。此不同类型的保险产品具有不同的类型属性，以表征其所适用的风险类型，如健康险的类型属性为健康，表征其所适用于在健康方面的风险。为了确定用户在各个类型风险上的风险大小，对用户在预设周期所去到的各个风险类型POI进行保险产品的属性归类，将各个风险类型POI归属到具有各个类型属性的保险产品中。因具有不同类型属性的保险产品与不同的风险类型对应，从而将用户去与相同类型属性保险产品对应的风险类型POI归为一类；如去诊所与医疗机构，因诊所与类型属性为健康的保险产品对应，而医疗机构也与类型属性为健康的保险产品对应；两者所对应保险产品的类型属性相同，而将诊所与医疗机构归为一类风险类型POI。通过各归类风险类型POI中各个风险类型POI的POI概率，可确定用户在此类风险类型POI的风险大小。在对各个风险类型POI均进行保险产品的属性归类，将各个风险类型POI均归为与具有各类型属性的保险产品对应后，需要进一步确定用户相对于各类保险产品的子风险POI概率；因各类型保险产品适用不同的风险类型，使得各子风险POI概率表征用户在各个类型风险上的风险大小，其可用各归类风险类型POI的POI概率之和表示。具体地，对各风险类型POI进行保险产品的属性归类，并确定用户相对于各类保险产品的子风险POI概率的步骤包括：
步骤S21，调用各类所述保险产品的类型属性标签，并将各所述风险标识符逐一和各所述类型属性标签对比，确定各所述风险标识符对应的目标类型属性标签；
进一步地，因各类保险产品用类型属性表示其所适用的风险类型，为了对各类保险产品所适用的风险类型进行区分，即类型属性进行区分；预先针对不同类型保险产品的类型属性设置有不同的类型属性标签，如类型属性为健康的保险产品，其所适用的风险类型为健康风险，对应的类型属性标签为g3。在对各风险类型POI进行属性归类时，先调用各保险产品的类型属性标签，再将各个表征归属为风险类型的风险标识符逐一和此各类型属性标签进行对比。需要说明的是，风险标识符虽然从POI概率中所读取，但携带有此风险标识符的POI概率为风险类型POI所具有，POI概率、风险类型POI、风险类型标识符具有对应性，可通过风险标识符表征风险类型POI的具体风险类型。具体地，可在风险标识符后面增加不同字符的方式来表征风险类型POI的具体风险类型，如对于表征风险类型POI的标识符f3，风险标识符即为f3，可在此f3后面增加不同的字符来来表征风险类型POI的具体风险类型；如用f3-31表征医院、医疗保健店，以说明具有风险标识符f3-31的风险类型POI的具体风险类型为健康风险；用f3-32表征赛马场、攀岩场，以说明具有风险标识符f3-32的风险类型POI的具体风险类型为意外风险。
为了表征风险类型的一致性，类型属性标签和风险标识符之间具有对应关系，如上述表征健康风险的g3与f3-31；从而在将各风险标识符逐一和各类型属性标签对比时，当判断出当前正在对比的风险标识符和某一类型属性标签具有对应关系时，则可判定此类型属性标签与当前的风险标识符对应；将此类型属性标签作为目标类型属性标签，表征具有当前风险标识符的风险类型POI的具体风险类型，与具有此目标类型属性标签的保险产品所适用的风险类型一致，可将所有具有与此目标类型属性标签一致的风险标识符的风险类型POI归为同一类。
步骤S22，对具有相同所述目标类型属性标签的各所述风险类型POI进行归类，形成与各类所述保险产品对应的风险类型POI集合；
更进一步地，在所有风险标识符均完成和各类型属性标签对比后，即可确定各个风险标识符所对应的目标类型属性标签；进而根据此目标类型属性标签，对各风险类型POI进行归类，将具有相同目标类型属性标签的各风险类型POI进行归类。需要说明的是，因类型属性标签和风险标识符具有对应关系，目标类型属性标签为与风险标识符对应的类型属性标签，使得风险类型POI所携带的风险标识符可用对应的目标类型属性标签表征。即当风险类型POI携带某一风险标识符，而此风险标识符与某一目标类型属性标签对应，而可用此目标类型属性标签表征风险类型POI中所携带的风险类型标识，可描述为此风险类型POI具有此目标类型属性标签。对具有相同目标类型属性标签的各个风险类型POI进行归类，形成与各类保险产品对应的风险类型POI集合，且与各类保险产品的对应关系，由各目标类型属性标签与保险产品的对应关系决定。如对于用户的风险类型POI分别包括M1、M2、M3,、M4和M5，其中M1、M2和M4所具有的目标类型属性标签为g2，M3和M5所具有的目标类型属性标签为g3；从而对M1、M2、M4进行归类，形成[M1、M2、M4]的风险类型POI集合；对M3和M5进行归类，形成[M3、M5]的风险类型POI集合；因g2与意外险类型的产品对应，而g3与健康险类型的产品对应，而使得[M1、M2、M4]与意外险产品对应，[M3、M5]与健康险产品对应。
步骤S23，对各所述风险类型POI集合中的各所述风险类型POI的所述POI概率进行累加，生成用户相对于各类所述保险产品的子风险POI概率。
可理解地，不同风险类型POI集合中涉及到多个具有相同类型风险的风险类型POI，而各个风险类型POI具有不同的POI概率；将此各个风险类型POI集合中包括的所有风险类型POI的POI概率逐一进行累加，当所有风险类型POI集合均累加完成后，所得到的结果为各个风险类型POI集合所具有的整体POI概率，表征用户在各种类型风险中的风险大小。因不同类型的风险对应不同类型的保险产品，此各个整体POI概率其实质为用户相对于各类保险产品的子风险概率，表征用户对各类保险产品的需求大小；当用户某个风险类型POI集合的子风险POI概率较大时，表明用户在此类风险类型POI集合所表征类型风险的风险较大，进而表明用户对适用于此类风险的保险产品的需求较大；通过用户相对于各类保险产品的子风险POI概率，确定用户对各类保险产品的需求情况。
步骤S30，将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
更进一步地，在确定用户在预设周期所涉及到的各个类型的风险，并用各保险产品表征各类型风险，以及用子风险POI概率表征各类型风险的大小后，将各子风险POI概率及对应的保险产品形成概率产品对，通过概率产品对反映用户在各类型风险上的风险大小，以及与此类型风险对应的保险产品类型。具体地，预先设置有用于生成风险画像的预设画像模板，且预设画像模板中用用户的各类特征标签变量表征用户所具有的各类特征；将形成的各个概率产品对作为用户特征标签添加到该预设画像模板中，替换其中的特征标签变量，将经添加替换操作后的预设画像模板作为用户在预设周期的风险画像。通过风险画像，直观的展现用户在日常生活中的风险分布情况，将风险画像作为保险产品的推荐依据，使得推荐的保险产品更贴合用户的实际情况。
本实施例的用户风险画像生成方法，先根据用户在预设周期内去各个POI的POI概率，确定用户去所有风险类型POI的总风险POI概率；此总风险POI概率反映了用户日常生活中所涉及到的风险情况，其中总风险POI概率越大，说明用户日常生活中所涉及到的风险越高；当判断出此总风险POI概率大于预设值时，说明用户日常生活中所涉及到的风险较高，依据所涉及到的风险类型向此用户推广保险产品的成功率较高；从而进一步对各风险类型POI进行保险产品的属性归类，确定用户相对于各类保险产品的子风险POI概率，此子风险POI概率表明用户对各类保险产品的需求情况；进而将各子风险POI概率与其对应的保险产品形成多个概率产品对，将该各概率产品对设为用户特征标签添加到预设画像模板中，生成用户在预设周期的风险画像。因各子风险POI概率由具有相同保险产品属性的风险类型POI的POI概率确定，而各风险类型POI体现了用户日常生活中所涉及到的各类风险，使得子风险POI概率与用户的各类风险相关；使得所生成的用户风险画像，可准确反映用户在日常生活中所涉及到的风险情况；避免使用单纯的用户风险数据来反映用户所具有的风险，实现了对用户所具有风险的精准反映。
进一步地，在本申请用户风险画像生成方法第一实施例的基础上，提出本申请用户风险画像生成方法第二实施例，在第二实施例中，所述生成用户在所述预设周期的风险画像的步骤之后包括：
步骤S40，对各所述子风险POI概率和所述总风险POI概率做比值，生成用户相对于各类所述保险产品的风险占比；
可理解地，用户在不同类型风险上的风险大小不一样，且风险大小可用所归类的各风险类型POI的POI概率之和在整个风险类型POI的总风险POI概率中的占比体现，即各子风险POI概率在总风险POI概率中的占比体现；其中总风险POI概率表征用户在预设周期中的总风险大小，而各子风险POI概率表征用户在预设周期中各类风险的风险大小；当某一子风险POI概率的占比越大，则说明此子风险POI概率在总风险POI概率中占有较大的数值，用户在预设周期中的风险主要来源与此子风险POI概率对应的风险类型。具体地，将各子风险POI概率逐一和总风险POI概率做比值，通过比值大小表征各子风险POI概率所对应风险类型占据用户在日常生活中总风险的多少。因不同类型的风险所适用的保险产品类型不一样，对应风险类型占据用户在日常生活中总风险多少，可表征用户对各类保险产品的需求大小；当风险类型占据用户日常生活中总风险较多，则说明用户对适用于此风险类型的保险产品的需求较大。从而将比值的数值结果作为用户相对于各类保险产品的风险占比，表征用户对各类保险产品的需求情况；当由子风险POI概率所生成的比值数值结果较大，则相应的风险占比较大，用户对此子风险POI概率对应风险类型所适用保险产品的需求较大。
步骤S50，将各所述风险占比对比，确定数值最大的目标风险占比，并确定与所述目标风险占比对应的保险产品类型；
进一步地，在生成用户在各类保险产品中的风险占比后，各个风险占比具有不同的数值大小，其中数值越大，说明用户在此风险占比来源保险产品所适用的类型风险上具有较大风险。为了确定用户所具有最大风险的类型风险，即日常生活中用户最主要的风险来源；将各个风险占比进行对比，确定数值最大的风险占比，并将此风险占比作为目标风险占比；因风险占比均归属于某类保险产品，而将目标风险占比所归属的保险产品类型作为与目标风险占比对应的保险产品类型，以表征用户最有可能需要的保险产品类型。
步骤S60，获取与所述保险产品类型对应的产品信息，并将所述产品信息传输到用户所持有的终端。
可理解地，各类保险产品适用于不同类型的风险，而不同类型的风险具有不同的特性；针对此不同特性，各类保险产品具有不同的产品信息。如针对健康类型风险的健康险，以及针对意外类型风险的意外险；因健康类型风险和意外类型风险的特征差异性，使得健康险与意外险之间存在产品信息差异性。在确定与目标风险占比对应的保险产品类型，表征用户最可能对此保险产品类型有需求时，因各类保险产品的产品信息均对应存储在金融机构服务器的存储单元中，从而可从存储单元中读取与此确定的保险产品类型对应的产品信息；对各保险产品和其产品信息分配一致的标识符，在确定保险产品类型后，读取此保险产品类型的标识符，而依据标识符查找到对应的产品信息进行读取。将此读取的产品信息传输到用户所持有的终端，方便用户了解其最大风险来源的风险类型所适用保险产品的产品信息；用户所持有的终端可以是用户使用的手机、平板电脑等，对此不做限制。
进一步地，在本申请用户风险画像生成方法另一实施例中，所述获取与所述保险产品类型对应的产品信息的步骤包括：
步骤S61，读取用户基本信息，并读取与所述保险产品类型对应的各保险产品的产品属性；
可理解地，不同用户虽然在各个类型风险上具有相似性，但各用户之间还具有个体化的差异；如用户K1和K2均喜欢户外运动，而使得两者的类型风险均为意外险，但K1年龄较大、而K2的年龄较小，而使得K2具有较好的体力，以及较好的风险应对能力；在两者面对同样的户外运动时，K2的意外风险相对于K1较小。为了更好的贴合各用户之间的个体差异需求，针对于不同类型风险的各类保险产品中包括有多个保险产品；多个保险产品整体所针对的风险类型相同，但各个保险产品之间因适用于不同用户的此类型风险，而具有不同的产品属性。如上述针对用户K1和K2类型风险的保险产品X1和X2，均属于意外类型风险，但对于X1和X2之间存在差异，X1适用年龄小的用户，而X2适用年龄大的用户，使得X1适用于用户K2，而X2适用于用户K1。在确定适用于用户的保险产品类型后，为了从保险产品类型中确定贴合用户个体需求的保险产品，需要进一步读取用户基本信息，此用户基本信息包括用户性别、年龄、居住城市、工作性质等；同时读取保险产品类型中所包括的各保险产品的产品属性，此产品属性表征各保险产品所适用的人群，如适用的年龄段、城市、月收入等。通过读取的用户基本信息，以及产品信息，可确定与用户基本信息所匹配的产品信息。
步骤S62，根据所述用户基本信息和各所述产品属性的匹配程度，确定目标保险产品，并将所述目标保险产品的产品信息确定为与所述保险产品类型对应的产品信息。
进一步地，在读取到用户基本信息以及各保险产品的产品属性后，将基本信息和产品属性逐一比对，确定用户基本信息和各产品属性的匹配程度；其中匹配程度为用户基本信息满足各产品属性需求的程度，如产品属性中所适用的年龄在20到30岁之间，且适用的城市为深圳，而当用户基本信息为年龄45岁，则用户与此产品属性不匹配。当用户基本信息越满足产品属性的需求，则两者之间的匹配程度越高，具有此产品属性的保险产品越适合用户。确定与用户基本信息匹配程度最高的产品属性，而将具有此产品属性的保险产品作为目标保险产品；读取此目标保险产品的产品信息，将此产品信息确定为与保险产品类型对应的产品信息，而将此产品信息传输到用户所持有的终端上进行目标保险产品的推荐。因所推荐的目标保险产品的产品信息与用户基本信息的匹配程度最高，可使得所推荐的目标保险产品在最大程度上对用户日常生活中所具有的风险进行保障，满足用户的需求。
进一步地，在本申请用户风险画像生成方法另一实施例中，所述读取用户在预设周期内去各POI的POI概率的步骤之前包括：
步骤q1，读取用户在预设周期内的各个位置信息，并确定以各所述位置信息为中心的预设范围内的POI；
可理解地，因各个POI概率依据用户在预设周期内的各个位置信息进行生成，而不同位置信息周围所具有的POI不相同，且用户在不同时间所去的位置不相同；从而在生成用户在预设周期中的各个POI概率时，通过GPS（Global Positioning System，全球定位系统）定位服务或基站定位服务获取用户在预设周期中的各个位置信息，并根据标识符确定以各个位置信息为中心的预设范围内的POI。其中读取位置信息和统计POI操作可以是在预设周期中的实时操作，也可以是在预设周期完成后的统一操作；对于实时操作，GPS或基站对用户的位置信息进行实时定位，并将所定位的位置信息进行存储；在预设周期中实时读取此定位的位置信息，确定以此位置信息为中心的预设范围内的POI；对于统一操作，则在预设周期的时间完成后，统一读取用户在预设周期所去到的位置信息，确定以各位置信息为中心的预设范围内的POI。考虑到虽然各位置信息周围的POI众多，但用户所去的POI有限，且距离各位置信息越远的POI，用户所去的可能性越小；从而预先根据多个历史数据中用户以位置信息为中心所去到POI的距离范围设定此预设范围，以使预设范围满足大部分人的出行需求。对各位置信息的此预设范围内的POI进行确定，表征用户从各位置信息对应位置所可能去到的各个POI。
步骤q2，统计所述POI的POI数量，并根据所述POI数量，确定用户从各所述位置信息去各所述POI的POI概率，生成用户在预设周期内去各POI的POI概率。
进一步地，对于在预设范围内的各个POI，用户可根据需求去到其需要去的POI，用户去各个POI的概率相同，且去各个POI的总概率之和为1。从而对于各个POI的POI概率，先对预设范围内的所具有POI的数量进行累加统计，统计得到的结果为POI数量，再根据此POI数量确定用户从各个位置信息去各个POI的POI概率。因用户从各位置信息去各个POI的概率相同，且总和为1；可用总和1对所具有的POI数量进行均分，即用1除以POI数量，用POI数量的倒数值的方式表征用户去各个POI的概率相同，且各概率均为POI数量的倒数值，生成用户从预设周期的某位置信息去此位置信息预设范围内中各个POI的POI概率。如某位置信息预设范围内有5个POI，则去各个POI的概率均为1/5。依照此方式对用户从各位置信息去到预设范围内各个POI的POI概率进行计算，直到所有位置信息的预设范围内的各个POI的POI概率均经计算确定，将此计算的用户从各位置信息去预设范围内各个POI的POI概率进行存储，以便后续根据此各个POI概率确定用户去各个POI的需求大小。此外，考虑到预设范围内不同POI与位置信息之间的距离不一样，不同的距离对用户从位置信息去各POI的概率有影响；其中距离位置信息越远的POI，用户所去到的可能性越小，从而降低了用户去此POI的概率；使得用户从位置信息去各个POI的POI概率除了与POI数量相关外，还与各个POI和位置信息之间的距离相关。具体地，根据POI数量，确定用户从各所述位置信息去各所述POI的POI概率，生成用户在预设周期内去各POI的POI概率的步骤包括：
步骤q21，根据各所述位置信息与所述预设范围的距离关系，将所述预设范围划分为多个远近区域；
更进一步地，因用户去各POI的概率与各POI和位置信息之间的距离远近相关；从而在确定各POI的POI概率时，先根据各位置信息与预设范围的距离关系，将预设范围划分为多个远近区域，不同远近区域与位置信息的距离不相同。具体地，以位置信息所表征的位置作为起始点，每间隔某一距离长度，将预设范围划分为多个远近区域；其中距离长度为预先通过历史数据设置的长度，多个远近区域为在不同起始点以此距离长度进行延伸的区域。如当前位置为A，而预设范围为以A为中心向外的1000m圆形范围、距离长度为500m；则将以A为起始点向前500m的圆形区域作为一个远近区域区间，并以圆形区间的边界，即距离A为500m的边界为起始点，将500m到1000m之间的圆环区域作为另一个远近区域。
步骤q22，根据所述远近区域的区域数量，为各所述远近区域分配区域概率；
可理解地，距离位置信息较近的远近区域，用户去到此远近区域的概率相对较大，而去到距离位置信息较远的远近区域的概率相对较小；从而依据远近区域与位置信息的距离关系，为各远近区域分配不同的区域概率，且与位置信息距离较近的远近区域分配较大的区域概率，而与位置信息距离较远的远近区域则分配较小的区域概率，各个远近区域的区域概率之和为1，通过此各不同的区域概率来表征用户去预设范围内各远近区域的可能性大小。
步骤q23，用所述POI数量的倒数值和所述区域概率做乘积，确定用户从各所述位置信息去各所述远近区域中所述POI的POI概率，生成用户在预设周期内去各POI的POI概率。
进一步地，在对各远近区域的区域概率分配后，在此各区域概率的基础上，确定用户去预设范围内各POI的POI概率。具体地，用远近区域所具有的区域概率与预设范围内所具有POI的POI数量的倒数值相乘；相乘的结果即为用户从位置信息去各远近区域中各个POI的POI概率，即用户从位置信息去预设范围内各个POI的POI概率；依照此方式对用户从各位置信息去到预设范围内各远近区域的各个POI的POI概率进行计算，直到所有位置信息的预设范围内的各POI概率均经计算完成。通过将预范围继续细分为多个远近区域，使所确定的POI概率更为准确；当预设范围内所具有的POI均计算完成后，将计算所得到的各个POI的POI概率进行存储，生成用户在预设周期内去各POI的POI概率，以便后续根据此各个POI概率精准的反映用户去各个POI的需求大小。
此外，请参照图2，本申请提供一种用户风险画像生成装置，兴趣点POI类型包括风险类型POI，在本申请用户风险画像生成装置第一实施例中，所述用户风险画像生成装置包括：
读取模块10，用于读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
确定模块20，用于判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
生成模块30，用于将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
其中，上述用户风险画像生成装置的各虚拟功能模块存储于图3所示用户风险画像生成设备的存储器1005中，处理器1001执行计算机可读指令时，实现图2所示实施例中各个模块的功能。需要说明的是，本领域普通技术人员可以理解实现上述实施例的全部或部分步骤可以通过硬件来完成，所述的程序可以存储于一种非易失性可读存储介质中，上述提到的存储介质可以是只读存储器，磁盘或光盘等。
参照图3，图3是本申请实施例方法涉及硬件运行环境的设备结构示意图。
本申请实施例用户风险画像生成设备可以是PC( personal computer，个人计算机 )，也可以是智能手机、平板电脑、电子书阅读器、便携计算机等终端设备。
如图3所示，该用户风险画像生成设备可以包括：处理器1001，例如CPU（Central Processing Unit，中央处理器），存储器1005，通信总线1002。其中，通信总线1002用于实现处理器1001和存储器1005之间的连接通信。存储器1005可以是高速RAM（random access memory，随机存取存储器），也可以是稳定的存储器（non-volatile memory），例如磁盘存储器。存储器1005可选的还可以是独立于前述处理器1001的存储装置。本领域技术人员可以理解，图3中示出的用户风险画像生成设备结构并不构成对用户风险画像生成设备的限定。
如图3所示，作为一种非易失性可读存储介质的存储器1005中可以包括操作系统、网络通信模块以及计算机可读指令。操作系统是管理和控制用户风险画像生成设备硬件和软件资源的程序，支持计算机可读指令以及其它软件和/或程序的运行。网络通信模块用于实现存储器1005内部各组件之间的通信，以及与用户风险画像生成设备中其它硬件和软件之间通信。
在图3所示的用户风险画像生成设备中，处理器1001用于执行存储器1005中存储的计算机可读指令，实现上述用户风险画像生成方法各实施例中的步骤。
本申请提供了一种非易失性可读存储介质，所述非易失性可读存储介质优选为计算机可短存储介质，所述计算机非易失性可读存储介质存储有一个或者一个以上程序，所述一个或者一个以上程序还可被一个或者一个以上的处理器执行以用于实现上述用户风险画像生成方法各实施例中的步骤。
以上所述仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是在本申请的构思下，利用本申请说明书及附图内容所作的等效结构变换，或直接/间接运用在其他相关的技术领域均包括在本申请的专利保护范围内。

Claims (20)

  1. 一种用户风险画像生成方法，其特征在于，兴趣点POI类型包括风险类型POI，所述用户风险画像生成方法包括以下步骤：
    读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
    判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
    将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
  2. 如权利要求1所述的用户风险画像生成方法，其特征在于，所述对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率的步骤包括：
    读取各所述POI概率中所携带的标识符，并从各所述标识符中筛选出归属为风险类型的风险标识符；
    将具有所述风险标识符的所述POI概率进行累加，得到用户去所有风险类型POI的总风险POI概率。
  3. 如权利要求2所述的用户风险画像生成方法，其特征在于，所述对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率的步骤包括：
    调用各类所述保险产品的类型属性标签，并将各所述风险标识符逐一和各所述类型属性标签对比，确定各所述风险标识符对应的目标类型属性标签；
    对具有相同所述目标类型属性标签的各所述风险类型POI进行归类，形成与各类所述保险产品对应的风险类型POI集合；
    对各所述风险类型POI集合中的各所述风险类型POI的所述POI概率进行累加，生成用户相对于各类所述保险产品的子风险POI概率。
  4. 如权利要求3所述的用户风险画像生成方法，其特征在于，所述生成用户在所述预设周期的风险画像的步骤之后包括：
    将各所述子风险POI概率和所述总风险POI概率做比值，生成用户相对于各类所述保险产品的风险占比；
    将各所述风险占比对比，确定数值最大的目标风险占比，并确定与所述目标风险占比对应的保险产品类型；
    获取与所述保险产品类型对应的产品信息，并将所述产品信息传输到用户所持有的终端。
  5. 如权利要求4所述的用户风险画像生成方法，其特征在于，所述获取与所述保险产品类型对应的产品信息的步骤包括：
    读取用户基本信息，并读取与所述保险产品类型对应的各保险产品的产品属性；
    根据所述用户基本信息和各所述产品属性的匹配程度，确定目标保险产品，并将所述目标保险产品的产品信息确定为与所述保险产品类型对应的产品信息。
  6. 如权利要求1所述的用户风险画像生成方法，其特征在于，所述读取用户在预设周期内去各POI的POI概率的步骤之前包括：
    读取用户在预设周期内的各个位置信息，并确定以各所述位置信息为中心的预设范围内的POI；
    统计所述POI的POI数量，并根据所述POI数量，确定用户从各所述位置信息去各所述POI的POI概率，生成用户在预设周期内去各POI的POI概率。
  7. 如权利要求6所述的用户风险画像生成方法，其特征在于，所述根据所述POI数量，确定用户从各所述位置信息去各所述POI的POI概率，生成用户在预设周期内去各POI的POI概率的步骤包括：
    根据各所述位置信息与所述预设范围的距离关系，将所述预设范围划分为多个远近区域；
    根据各所述远近区域与所述位置信息的距离关系，为各所述远近区域分配区域概率；
    用所述POI数量的倒数值和所述区域概率做乘积，确定用户从各所述位置信息去各所述远近区域中所述POI的POI概率，生成用户在预设周期内去各POI的POI概率。
  8. 一种用户风险画像生成装置，其特征在于，兴趣点POI类型包括风险类型POI，所述用户风险画像生成装置包括：
    读取模块，用于读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
    确定模块，用于判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
    生成模块，用于将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
  9. 一种用户风险画像生成设备，其特征在于，所述用户风险画像生成设备包括：存储器、处理器、通信总线以及存储在所述存储器上的计算机可读指令；
    所述通信总线用于实现处理器和存储器之间的连接通信；
    所述处理器用于执行所述计算机可读指令，以实现以下步骤：
    读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
    判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
    将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
  10. 如权利要求9所述的用户风险画像生成设备，其特征在于，所述对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率的步骤包括：
    读取各所述POI概率中所携带的标识符，并从各所述标识符中筛选出归属为风险类型的风险标识符；
    将具有所述风险标识符的所述POI概率进行累加，得到用户去所有风险类型POI的总风险POI概率。
  11. 如权利要求10所述的用户风险画像生成设备，其特征在于，所述对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率的步骤包括：
    调用各类所述保险产品的类型属性标签，并将各所述风险标识符逐一和各所述类型属性标签对比，确定各所述风险标识符对应的目标类型属性标签；
    对具有相同所述目标类型属性标签的各所述风险类型POI进行归类，形成与各类所述保险产品对应的风险类型POI集合；
    对各所述风险类型POI集合中的各所述风险类型POI的所述POI概率进行累加，生成用户相对于各类所述保险产品的子风险POI概率。
  12. 如权利要求11所述的用户风险画像生成设备，其特征在于，所述生成用户在所述预设周期的风险画像的步骤之后，所述处理器用于执行所述计算机可读指令，以实现以下步骤：
    将各所述子风险POI概率和所述总风险POI概率做比值，生成用户相对于各类所述保险产品的风险占比；
    将各所述风险占比对比，确定数值最大的目标风险占比，并确定与所述目标风险占比对应的保险产品类型；
    获取与所述保险产品类型对应的产品信息，并将所述产品信息传输到用户所持有的终端。
  13. 如权利要求12所述的用户风险画像生成设备，其特征在于，所述获取与所述保险产品类型对应的产品信息的步骤包括：
    读取用户基本信息，并读取与所述保险产品类型对应的各保险产品的产品属性；
    根据所述用户基本信息和各所述产品属性的匹配程度，确定目标保险产品，并将所述目标保险产品的产品信息确定为与所述保险产品类型对应的产品信息。
  14. 如权利要求9所述的用户风险画像生成设备，其特征在于，所述读取用户在预设周期内去各POI的POI概率的步骤之前，所述处理器用于执行所述计算机可读指令，以实现以下步骤：
    读取用户在预设周期内的各个位置信息，并确定以各所述位置信息为中心的预设范围内的POI；
    统计所述POI的POI数量，并根据所述POI数量，确定用户从各所述位置信息去各所述POI的POI概率，生成用户在预设周期内去各POI的POI概率。
  15. 一种非易失性可读存储介质，其特征在于，所述非易失性可读存储介质上存储有计算机可读指令，所述计算机可读指令被处理器执行，实现以下步骤：
    读取用户在预设周期内去各POI的POI概率，并对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率；
    判断所述总风险POI概率是否大于预设值，若大于所述预设值，则对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率；
    将各所述子风险POI概率及对应的保险产品形成概率产品对，并将各所述概率产品对设为用户特征标签添加到预设画像模板中，生成用户在所述预设周期的风险画像。
  16. 如权利要求15所述的非易失性可读存储介质，其特征在于，所述对各所述POI概率按照所归属的POI类型进行统计，确定用户去所有风险类型POI的总风险POI概率的步骤包括：
    读取各所述POI概率中所携带的标识符，并从各所述标识符中筛选出归属为风险类型的风险标识符；
    将具有所述风险标识符的所述POI概率进行累加，得到用户去所有风险类型POI的总风险POI概率。
  17. 如权利要求16所述的非易失性可读存储介质，其特征在于，所述对各所述风险类型POI进行保险产品的属性归类，并确定用户相对于各类所述保险产品的子风险POI概率的步骤包括：
    调用各类所述保险产品的类型属性标签，并将各所述风险标识符逐一和各所述类型属性标签对比，确定各所述风险标识符对应的目标类型属性标签；
    对具有相同所述目标类型属性标签的各所述风险类型POI进行归类，形成与各类所述保险产品对应的风险类型POI集合；
    对各所述风险类型POI集合中的各所述风险类型POI的所述POI概率进行累加，生成用户相对于各类所述保险产品的子风险POI概率。
  18. 如权利要求17所述的非易失性可读存储介质，其特征在于，所述生成用户在所述预设周期的风险画像的步骤之后，所述计算机可读指令被处理器执行，实现以下步骤：
    将各所述子风险POI概率和所述总风险POI概率做比值，生成用户相对于各类所述保险产品的风险占比；
    将各所述风险占比对比，确定数值最大的目标风险占比，并确定与所述目标风险占比对应的保险产品类型；
    获取与所述保险产品类型对应的产品信息，并将所述产品信息传输到用户所持有的终端。
  19. 如权利要求18所述的非易失性可读存储介质，其特征在于，所述获取与所述保险产品类型对应的产品信息的步骤包括：
    读取用户基本信息，并读取与所述保险产品类型对应的各保险产品的产品属性；
    根据所述用户基本信息和各所述产品属性的匹配程度，确定目标保险产品，并将所述目标保险产品的产品信息确定为与所述保险产品类型对应的产品信息。
  20. 如权利要求15所述的非易失性可读存储介质，其特征在于，所述读取用户在预设周期内去各POI的POI概率的步骤之前，所述计算机可读指令被处理器执行，实现以下步骤：
    读取用户在预设周期内的各个位置信息，并确定以各所述位置信息为中心的预设范围内的POI；
    统计所述POI的POI数量，并根据所述POI数量，确定用户从各所述位置信息去各所述POI的POI概率，生成用户在预设周期内去各POI的POI概率。
PCT/CN2018/123218 2018-09-27 2018-12-24 用户风险画像生成方法、装置、设备及相关设备 WO2020062657A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811135650.1A CN109447689A (zh) 2018-09-27 2018-09-27 用户风险画像生成方法、装置、设备及可读存储介质
CN201811135650.1 2018-09-27

Publications (1)

Publication Number Publication Date
WO2020062657A1 true WO2020062657A1 (zh) 2020-04-02

Family

ID=65545937

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/123218 WO2020062657A1 (zh) 2018-09-27 2018-12-24 用户风险画像生成方法、装置、设备及相关设备

Country Status (2)

Country Link
CN (1) CN109447689A (zh)
WO (1) WO2020062657A1 (zh)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111724218A (zh) * 2019-03-18 2020-09-29 河南省技术产权交易所有限公司 知识产权挂牌方法、装置、设备及计算机可读存储介质
CN110322274A (zh) * 2019-05-30 2019-10-11 深圳壹账通智能科技有限公司 基于数据分析的人群画像生成方法、装置和计算机设备
CN110766559B (zh) * 2019-10-30 2023-09-29 海腾保险代理有限公司 标签配置方法及装置
CN110782361A (zh) * 2019-10-30 2020-02-11 海腾保险代理有限公司 定制保险的推荐方法及装置
CN110689446A (zh) * 2019-10-30 2020-01-14 海腾保险代理有限公司 组合保险的推荐方法及装置
CN115099680B (zh) * 2022-07-14 2024-02-02 平安科技（深圳）有限公司 风险管理方法、装置、设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106651603A (zh) * 2016-12-29 2017-05-10 平安科技（深圳）有限公司 基于位置服务的风险评估方法及装置
CN106682394A (zh) * 2016-11-30 2017-05-17 北京拓明科技有限公司 一种生存风险的大数据分析方法及系统
CN107657527A (zh) * 2017-09-29 2018-02-02 平安科技（深圳）有限公司 贷款产品匹配方法、装置及计算机可读存储介质
CN107730389A (zh) * 2017-09-30 2018-02-23 平安科技（深圳）有限公司 电子装置、保险产品推荐方法及计算机可读存储介质
CN108288229A (zh) * 2018-03-02 2018-07-17 北京邮电大学 一种用户画像构建方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106656932B (zh) * 2015-11-02 2020-03-20 阿里巴巴集团控股有限公司 一种业务处理方法及装置
CN107566358B (zh) * 2017-08-25 2020-10-30 腾讯科技（深圳）有限公司 一种风险预警提示方法、装置、介质及设备
CN107818506A (zh) * 2017-09-30 2018-03-20 上海壹账通金融科技有限公司 电子装置、信贷风险控制方法及存储介质
CN108171610B (zh) * 2017-12-28 2020-10-30 中国平安人寿保险股份有限公司 页面生成方法、页面生成设备、存储介质及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106682394A (zh) * 2016-11-30 2017-05-17 北京拓明科技有限公司 一种生存风险的大数据分析方法及系统
CN106651603A (zh) * 2016-12-29 2017-05-10 平安科技（深圳）有限公司 基于位置服务的风险评估方法及装置
CN107657527A (zh) * 2017-09-29 2018-02-02 平安科技（深圳）有限公司 贷款产品匹配方法、装置及计算机可读存储介质
CN107730389A (zh) * 2017-09-30 2018-02-23 平安科技（深圳）有限公司 电子装置、保险产品推荐方法及计算机可读存储介质
CN108288229A (zh) * 2018-03-02 2018-07-17 北京邮电大学 一种用户画像构建方法

Also Published As

Publication number Publication date
CN109447689A (zh) 2019-03-08

Similar Documents

Publication Publication Date Title
WO2020062657A1 (zh) 用户风险画像生成方法、装置、设备及相关设备
US10038968B2 (en) Branching mobile-device to system-namespace identifier mappings
CN111933299A (zh) 传染病感染风险评估方法和装置、电子设备以及存储介质
CN108230206A (zh) 基于互联网的医养结合监管系统和方法
CN109542966B (zh) 数据融合方法、装置、电子设备及计算机可读介质
Stribling et al. Examining the UK Covid‐19 mortality paradox: Pandemic preparedness, healthcare expenditure, and the nursing workforce
WO2023050779A1 (zh) 预约服务的数量分析方法、装置、设备及存储介质
Zandbergen Influence of street reference data on geocoding quality
WO2022035018A1 (ko) 클리닉 매니저 서비스 플랫폼 및 이를 제공하는 시스템
Nikel et al. Perceived quality of bus transit services: a route-level analysis
JP7292368B2 (ja) デバイスからの属性および位置シグネチャを使用してデバイスを識別する方法、その方法のための一意に生成された識別子のサーバ、およびその方法のための命令シーケンスを記憶する非一時的コンピュータ可読記憶媒体
Wu et al. How do urban services facilities affect social segregation among people of different economic levels? A case study of Shenzhen city
CN113704306B (zh) 数据库的数据处理方法、装置、存储介质及电子设备
WO2019091003A1 (zh) 薪资计算方法、装置、设备及计算机可读存储介质
CN112561636A (zh) 一种推荐方法、装置、终端设备及介质
Zhou et al. Evaluating the nonlinear impacts of built environment on online car-hailing: Focus on the whole day
CN111933283A (zh) 一种健康监测方法及装置
CN110941638B (zh) 应用分类规则库构建方法、应用分类方法及装置
WO2022075560A1 (ko) 에너지 데이터 중개 시스템 및 방법
Murad Planning and location of health care services in Jeddah City, Saudi Arabia: Discussion of the constructive use of geographical information systems
CN115795525A (zh) 敏感数据识别方法、装置、电子设备、介质和程序产品
Mizuno et al. Optimizing intra-facility crowding in Wi-Fi environments using continuous-time Markov chains
Du et al. Efficiency Evaluation and Influence Factor Analysis of Medical Malls Taking Fukuoka City, Japan as a Case Study
Achabal et al. Designing and evaluating a health care delivery system through the use of interactive computer graphics
Hashtarkhani et al. Where to place emergency ambulance vehicles? Using a capacitated maximum covering location model with real call data

Legal Events

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

Ref document number: 18934748

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 08/07/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 18934748

Country of ref document: EP

Kind code of ref document: A1