Product maintenance information pushing method and system based on user data

Abstract

The invention discloses a product maintenance information pushing method and system based on user data, and relates to the technical field of maintenance information pushing. The method comprises the following steps: collecting product data of a user, obtaining user data of the user, and estimating a product fault type in combination with the product data and the user data; using data of the product is extracted according to the product data, and the fault probability of the product is evaluated according to the using data of the product; economic data of the user are obtained, and a maintenance willingness value of the user is obtained by combining the fault type and fault probability evaluation; and judging whether the maintenance willingness value reaches a preset willingness threshold value or not, and if the maintenance willingness value reaches the preset willingness threshold value, pushing product maintenance information corresponding to the fault type to the user. According to the invention, the product maintenance information pushing accuracy based on the user data is improved.

Description

Technical Field

[0001] This application relates to the technical field of repair information push, and in particular to a method and system for pushing product repair information based on user data. Background Art

[0002] Pushing product repair information is an effective way to ensure that customers can timely understand the repair status, maintenance suggestions and important updates of their equipment or products. Such information push not only improves the customer experience, enhances brand loyalty, but also helps manufacturers or service providers optimize service processes and improve efficiency. However, the content of the pushed information often lacks pertinence and cannot accurately meet the personalized needs of users. For valuable and important repair information, it usually needs to be sent to specific user groups or devices, but the existing technologies often fail to meet this requirement. This will lead to many push systems failing to accurately grasp the information that users are most interested in or most relevant to when pushing repair information to users, resulting in users receiving a large amount of useless or duplicate information. This not only reduces users' trust in the system, but may also affect users' enthusiasm for continuing to use the push. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and system for pushing product repair information based on user data to solve the problems raised in the above background art.

[0004] In the first aspect, the method for pushing product repair information based on user data provided by this application adopts the following technical solutions: Collect product data of users, obtain user data of users, and estimate the product failure type by combining the product data and the user data; Extract the usage data of the product according to the product data, and evaluate the failure probability of the product according to the usage data of the product; Obtain the economic data of the user, and evaluate the repair willingness value of the user by combining the failure type and the failure probability; Judge whether the repair willingness value reaches a preset willingness threshold. If it reaches the preset willingness threshold, push the product repair information corresponding to the failure type to the user.

[0005] Preferably, the step of collecting product data of users, obtaining user data of users, and estimating the product failure type by combining the product data and the user data is specifically as follows: Extract the search records of the user according to the user data, judge whether the product is definitely faulty according to the search records. If the product is definitely faulty, obtain the product failure type according to the search records; If the product is not definitely faulty, extract the product maintenance standard and the user usage duration according to the product data; Extract the user's product usage habits from the user data and record them as usage habits, and analyze the first type of failure in combination with the product maintenance standards; Obtain the historical failure records of the product, and obtain the second type of failure based on the historical failure records and the user's usage duration; Count the first type of failure and the second type of failure to obtain the product failure type.

[0006] Preferably, the step of extracting the user's product usage habits from the user data and recording them as usage habits, and analyzing the first type of failure in combination with the product maintenance standards is specifically as follows: Extract the user's product usage habits from the user data and record them as usage habits; Extract the maintenance habits according to the product maintenance standards, and compare to obtain the habit differences between the usage habits and the maintenance habits; Extract the product failures caused by not performing the maintenance habits according to the product maintenance standards to form a habit-failure table; Find the corresponding failures in the habit-failure table according to the habit differences to obtain the first type of failure.

[0007] Preferably, the step of extracting the user's product usage habits from the user data and recording them as usage habits is specifically as follows: Obtain the user's historical repair records, and extract the historical product failures according to the historical repair records; Find the corresponding habits in the habit-failure table according to the historical product failures, and form the user's usage habits by aggregating all the habits; If the user has no historical repair records, then obtain the usage habits of all users with historical repair records, and find the habit with the highest frequency in the same type of habits as the standard habit; Form a habit set from all types of standard habits as the usage habits of users without historical repair records.

[0008] Preferably, the step of obtaining the historical failure records of the product and obtaining the second type of failure based on the historical failure records and the user's usage duration is specifically as follows: Obtain the historical failure records of similar products, extract the product usage durations corresponding to different failures, calculate the average value of all corresponding product usage durations to obtain the average usage duration; Form a failure-duration table according to the average usage durations corresponding to different failures; Obtain the user's usage duration, and find the failure corresponding to the user's usage duration in the failure-duration table to obtain the second type of failure.

[0009] Preferably, the step of extracting the usage data of the product according to the product data and evaluating the failure probability of the product is specifically as follows: Compare to obtain the similarity of usage habits and maintenance habits, and record it as the habit similarity; According to the preset correlation curve of habit similarity and failure probability, search to obtain the corresponding failure probability, and record it as the first failure probability; Judge whether the product has been repaired according to the historical maintenance record. If the product has been repaired, obtain the maintenance time point of the product; Obtain the real-time time point, calculate the duration between the real-time time point and the maintenance time point and record it as the real-time usage duration; If the product has not been repaired, use the user's usage duration as the real-time usage duration; Establish a correlation curve between the real-time usage duration and the failure probability, and search to obtain the corresponding failure probability according to the real-time usage duration, and record it as the second failure probability; Combine the first failure probability and the second failure probability to calculate the failure probability of the product.

[0010] Preferably, the step of obtaining the economic data of the user and evaluating the user's willingness to repair by combining the failure type and the failure probability is specifically as follows: Obtain the user's economic income, and record the people with the same economic income level as the user as the same-income population; Obtain the average purchase price of the same type of product of the same-income population, and record it as the average price level; Obtain the consumption price of the product purchased by the user, and calculate the price difference between the consumption price and the average price level; Collect the preferential information on the market, and search for the maximum preferential reduction price of the same type of product according to the preferential information and record it as the reduced price; Combine the economic income, the price difference and the reduced price to calculate the user's economic willingness value; Obtain the product information of the same type of product on the market, and analyze the user's demand willingness value according to the product information; Obtain the browsing information of the user, and analyze the user's replacement willingness value according to the browsing information; Respectively set the proportionality factors of the user's economic willingness value, demand willingness value and replacement willingness value, and calculate the user's willingness to repair according to the proportionality factors.

[0011] Preferably, the step of obtaining the product information of the same type of product on the market and analyzing the user's demand willingness value according to the product information is specifically as follows: Obtain the product information of the same type of product on the market, and extract the product functions according to the information of the same type of product on the market and record them as the first functions; Obtain the product information of the user's product, and extract the product functions according to the user's product information and record them as the second functions; Compare the first function and the second function, find the upgraded function, and count the number of functions of the upgraded function; Count the proportion of the user's product in the market products and record it as the second proportion; Set the proportionality coefficients of the number of functions and the second proportion respectively, and calculate the demand willingness value according to the proportionality coefficients.

[0012] Preferably, the step of obtaining the browsing information of the user and analyzing the replacement willingness value of the user according to the browsing information is specifically as follows: Obtain the browsing information of the user, and judge whether the user browses the same type of products according to the browsing information; If the user browses the same type of products, count the browsing times and browsing duration of the user browsing the same type of products; Calculate the first replacement willingness value of the user according to the browsing times and browsing duration. If the user does not browse the same type of products, the first replacement willingness value is 0; Collect and count the total number of historical repairs of the user, and collect the replacement times of the user's same type of products; Calculate the ratio of the total number of historical repairs to the replacement times and record it as the user repair ratio; Statistically obtain the repair probability of the user for the failure type, and calculate the replacement willingness value of the user in combination with the first replacement willingness value, the repair ratio and the repair probability.

[0013] In a second aspect, the product repair information push system based on user data provided by the present application adopts the following technical solutions: The product repair information push system based on user data includes: The failure type module collects the product data of the user, obtains the user data of the user, and estimates the product failure type by combining the product data and the user data; The failure probability module extracts the usage data of the product according to the product data, and evaluates the failure probability of the product according to the usage data of the product; The repair willingness module obtains the economic data of the user, and evaluates the repair willingness value of the user by combining the failure type and the failure probability; The repair push module judges whether the repair willingness value reaches a preset willingness threshold. If it reaches the preset willingness threshold, it pushes the product repair information corresponding to the failure type to the user.

[0014] In summary, the present application includes at least one of the following beneficial technical effects: 1. Based on the failure type and probability of the user's product, combined with the repair willingness value obtained according to the user's economic willingness value, demand willingness value, and replacement willingness value for repairing the product, it is determined whether to push product repair information to the user. If product repair information is pushed, the product repair information corresponding to the failure type is pushed. More accurate pushing of product repair information can effectively reduce the user's boredom with repairing the product while pushing the product repair information in a timely manner, and improve the accuracy of pushing product repair information based on user data.

[0015] 2. Estimate the failure type of the product based on the product's maintenance habits and user habits, and combine it with the failure type brought about by the product's usage duration to obtain the failure type of the product. More accurate estimation of the product's failure type is conducive to pushing more accurate product repair information to the user, and improves the accuracy of pushing product repair information based on user data.

[0016] 3. Evaluate the user's economic willingness value for repairing the product using the product price and the user's economic income, evaluate the user's demand willingness value for repairing the product according to the product's function upgrade situation, combine the user's product browsing situation and the user's historical repair situation to evaluate the user's replacement willingness value for repairing the product, and finally comprehensively obtain the user's repair willingness. Pushing product repair information according to the repair willingness can effectively reduce the interference of product repair information to the user, and improve the pertinence of pushing product repair information based on user data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic diagram of the specific steps of an embodiment of the method for pushing product repair information based on user data of the present invention.

[0018] Figure 2 is a schematic diagram of the module connection of an embodiment of the system for pushing product repair information based on user data of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0019] The following combines the embodiments and Figure 1 - Figure 2 further elaborates on the present invention in detail, but the embodiments of the present invention are not limited thereto.

[0020] The present invention discloses a method for pushing product repair information based on user data, specifically including the following steps: Step S1, collect the product data of the user, obtain the user data of the user, and estimate the product failure type by combining the product data and the user data.

[0021] Collect the product data and user data of the user through channels such as the user's authorization and active filling.

[0022] Step S2, extract the usage data of the product according to the product data, and evaluate the failure probability of the product according to the usage data of the product.

[0023] Step S3: Obtain the user's economic data, and combine the fault type and fault probability assessment to obtain the user's repair willingness value.

[0024] Step S4: Determine whether the repair willingness value reaches the preset willingness threshold. If it reaches the preset willingness threshold, push the product repair information corresponding to the fault type to the user.

[0025] If it reaches the preset willingness threshold, use the fault type with the fault probability reaching the preset fault threshold as the push type, and push the product repair information corresponding to the push type to the user.

[0026] In actual application, frequent pushing of product repair information will cause interference to users, resulting in users getting bored with product repair information, which is not conducive to the conversion of product repair information. Whether users need product repair information is related to whether their products need repair on the one hand and whether they have the willingness to repair on the other hand. If the user's product is in good condition and does not need repair, then the user does not need product repair information. Similarly, if the user has no willingness to repair, for example, the user is preparing to replace the product, then the user also does not need product repair information. Pushing when the user needs product repair information can not only improve the conversion rate of product repair information, increase the sales volume of product repair, but also reduce the interference of product repair information to users and improve the accuracy of product repair information push.

[0027] The steps of collecting the user's product data, obtaining the user's user data, and combining the product data and user data to estimate the product fault type are specifically as follows: Step S11: Extract the user's search records according to the user data, and judge whether the product is definitely faulty according to the search records. If the product is definitely faulty, obtain the product fault type according to the search records.

[0028] Search for the platform search records authorized by the user, and confirm whether the user is searching for relevant information such as product repair according to keywords such as "repair" and "broken". If the user confirms that there are the set keywords in the search, it is judged that the product is definitely faulty.

[0029] Step S12: If the product is not definitely faulty, extract the product maintenance standard and the user's usage duration according to the product data.

[0030] The fault situation of the product is highly related to the user's usage. Different products have different maintenance standards. For example, refrigerators and washing machines need to be cleaned regularly, and hair dryers cannot be used continuously for a long time, etc. Maintenance suggestions can be extracted according to the product information as the maintenance standard. The user's usage duration can be calculated according to the user's purchase date.

[0031] Step S13: Extract the user's product usage habits from the user data, record them as usage habits, and analyze the first type of failure in combination with the product maintenance standards.

[0032] Step S14: Obtain the historical failure records of the product, and obtain the second type of failure based on the historical failure records and the user's usage duration.

[0033] Step S15: Count the first type of failure and the second type of failure to obtain the product failure type.

[0034] In actual application, the types of product failures usually come from two aspects. On the one hand, they come from the product itself. For example, a washing machine is prone to the failure of dirty clothes during use. On the other hand, they come from user usage. Improper user usage will increase additional types of failures. Therefore, the first type of failure can be obtained based on the user's usage habits, and the second type of failure can be obtained based on the historical failure records of the product, so as to estimate the possible types of failures that may occur during the user's use of the product. Analyzing according to specific users is conducive to the targeted push of subsequent product repair information and improves the accuracy of product repair information push.

[0035] The steps of extracting the user's product usage habits from the user data, recording them as usage habits, and analyzing the first type of failure in combination with the product maintenance standards are specifically as follows: Step S131: Extract the user's product usage habits from the user data and record them as usage habits.

[0036] Step S132: Extract the maintenance habits according to the product maintenance standards, and compare to obtain the habit differences between the usage habits and the maintenance habits.

[0037] For example, the maintenance habit of a refrigerator is to clean the inside of the refrigerator once a month and defrost it once a year. The user's usage habit is to clean the inside of the refrigerator once every two months and defrost it once a year. Then the habit difference between the usage habit and the maintenance habit is cleaning the inside of the refrigerator.

[0038] Step S133: Extract the product failures caused by not performing the maintenance habits according to the product maintenance standards to form a habit-failure table.

[0039] The maintenance standards given in the product information are all summarized from historical failures. Therefore, if a certain maintenance standard is not performed, there will be corresponding possible types of failures. For example, not cleaning the inside of the refrigerator may cause failure types such as drainage system failure and increased power consumption. According to the failures caused by not performing the maintenance habits, they are in one-to-one correspondence with the maintenance habits to form a habit-failure table, that is, not performing what maintenance habit will cause the corresponding type of failure.

[0040] Step S134: Search in the habit-fault table according to the habit difference to obtain the corresponding fault, and obtain the first fault type.

[0041] In actual application, according to the habit difference, the corresponding fault can be searched in the habit-fault table, so as to obtain the first fault type. For example, if the habit difference is that the user does not clean the washing machine regularly, the corresponding faults are dirt accumulation and drainage faults, so the first fault type is dirt accumulation and drainage faults.

[0042] The step of extracting the user's habit of using the product from the user data and recording it as the usage habit is specifically as follows: Step S1311: Obtain the user's historical maintenance records, and extract the historical faults of the product according to the historical maintenance records.

[0043] According to the historical maintenance records, it can be known which faults have occurred in the products used by the user before.

[0044] Step S1312: Search in the habit-fault table according to the historical faults of the product to obtain the corresponding habits, and form the user's usage habits by aggregating all the habits.

[0045] The corresponding fault can be inferred according to the habit, and the user's usage habit can also be inferred according to the fault. The habit-fault tables of different products are different. Find the corresponding habit-fault table of the product according to the historical faults of the product, and find the corresponding habit from it according to the fault. For example, the user has repaired a washing machine and an air conditioner. The fault of the washing machine is that the bearing is broken, and the corresponding habit is improper use. Then the specific operation of improper use is the usage habit of the user's washing machine product. The fault of the air conditioner is that the cooling effect drops, and the corresponding habit is not cleaning the product in time. Then take the user's cleaning frequency as the usage habit of the user's air conditioner product. Different products have different usage habits, and form usage habits for the user's existing products. For example, for a TV, there is no need to clean the inside, and the searched usage habits only target maintenance habits, and find the corresponding usage habits according to what kind of maintenance habits are needed.

[0046] Step S1313: If the user has no historical maintenance records, obtain the usage habits of all users with historical maintenance records, and search for the habit with the highest frequency in the same type of habits as the standard habit.

[0047] Different habits also correspond to different types, and some users do not have historical maintenance records, so it is difficult to extract the user's usage habits. In this case, the general usage habits of the public are used as standard habits. For example, among the user's usage habits that can be found, the most frequent cleaning of the air conditioner is once a year, and it is recommended to clean the air conditioner at least once every six months. Therefore, cleaning the air conditioner once a year is used as the user's usage habit of cleaning the air conditioner type. If the user does not have the habit of complying with the usage rules, the habit with the highest frequency among the same type of habits will also be used as the user's standard habit of complying with the usage rules type.

[0048] Step S1314, forming a habit set of all types of standard habits as the usage habits of users without historical maintenance records.

[0049] In actual use, all types of standard habits of user products are formed into a habit set as the usage habits of users without historical maintenance records. For users with historical maintenance records but incomplete habit types, the missing standard habits are substituted in, and combined with the existing usage habits to form the user's usage habits. Different products have corresponding usage habits, and corresponding usage habits are formed according to the product maintenance habits. For example, the maintenance habits of a washing machine are regular cleaning and usage time, so what is searched is the usage habits of the washing machine as the user's usage habits. If the user has not repaired the washing machine, the most frequent habits of other users in the historical maintenance of the washing machine are used as the user's usage habits.

[0050] The steps of obtaining the historical fault records of the product and obtaining the second fault type according to the historical fault records and the user's usage time are specifically as follows: Step S141, obtaining historical fault records of similar products, extracting product usage times corresponding to different faults, calculating the average usage time of all corresponding products, and obtaining the average usage time.

[0051] Sometimes product failures do not come from user use, but sometimes from the product itself. As the product is used, it will also fail due to its own aging. The duration of different failures is also different. For example, failures such as control panel failure and loose door seals occur in the first 3-5 years of the washing machine, and problems such as water seepage, water leakage, heating and vibration occur in the 5-8 years of the washing machine. As the product usage time changes, the time nodes of different failures are also different. Calculate the average duration of the failure. For example, the panel failure of device A lasts for 5 years, the panel failure of device B lasts for 1 year, and the panel failure of device C lasts for 3 years. The average duration of the product with panel failure is 3 years. According to the existing fault records on the market, the average of the calculated product usage time is used as the average usage time.

[0052] Step S142: Form a fault-duration table based on the average usage duration corresponding to different faults.

[0053] Since the nodes of the usage durations at which different faults occur are different, a one-to-one correspondence is formed according to the average values to obtain the fault-duration table. For example, the duration corresponding to the malfunction of the washing machine panel is 3 years, the duration corresponding to water seepage is 6 years, and the duration corresponding to electric leakage is 9 years.

[0054] Step S143: Obtain the user's usage duration, and find the fault corresponding to the user's usage duration according to the fault-duration table to get the second fault type.

[0055] In actual application, according to the user's usage duration, find in the fault-duration table the faults that the product may have due to the change of the usage years. Estimating the possible fault types of the user's product according to the usage duration is beneficial to promoting more user-suitable product maintenance information subsequently.

[0056] The steps of extracting the usage data of the product from the product data and evaluating the fault probability of the product according to the usage data of the product are specifically as follows: Step S21: Compare to obtain the similarity of usage habits and maintenance habits, and record it as the habit similarity.

[0057] Use the existing similarity model to compare the similarity of usage habits and maintenance habits. For example, the maintenance habit is to clean the washing machine once a month. User A cleans the washing machine once every two months, and User B cleans the washing machine once every six months. Therefore, the habit similarity of User A is greater than that of User B.

[0058] Step S22: According to the preset correlation curve between the habit similarity and the fault probability, find the corresponding fault probability and record it as the first fault probability.

[0059] When the habit similarity is greater, it means that the user uses the product more in accordance with the maintenance standard, and the probability of failure is lower.

[0060] Step S23: Judge whether the product has been repaired according to the historical repair records. If the product has been repaired, obtain the repair time point of the product.

[0061] Step S24: Obtain the real-time time point, calculate the duration between the real-time time point and the repair time point and record it as the real-time usage duration.

[0062] Step S25: If the product has not been repaired, use the user's usage duration as the real-time usage duration.

[0063] Step S26: Establish a correlation curve between the real-time usage duration and the failure probability, and find the corresponding failure probability according to the real-time usage duration, which is denoted as the second failure probability.

[0064] Moreover, the longer the product usage duration is, the greater the probability of failure will be as the product ages. If the product has been repaired, the failures caused by aging have been dealt with, and it will be more accurate to calculate the real-time usage duration starting from the repair point.

[0065] Step S27: Combine the first failure probability and the second failure probability to calculate the failure probability of the product.

[0066] In practical applications, the weight ratios of the first failure probability and the second failure probability are set respectively, and the failure probability of the product is calculated according to the weight ratios. The failure sources of the product are the aging of the product itself and the user's usage. Therefore, it will be more accurate to evaluate the failure probability from these two aspects. For example, originally the probability of the product failing in the second year of use is 50%, but the user has poor usage habits and causes some damage to the product imperceptibly. Therefore, the probability of the product failing in the second year of use will increase and be greater than 50%.

[0067] The steps of obtaining the user's economic data and evaluating the user's repair willingness value in combination with the failure type and the failure probability are as follows: Step S31: Obtain the user's economic income, and group the people with the same economic income level as the user as the same-income group.

[0068] The user's economic income is obtained through user input or public information.

[0069] Step S32: Obtain the average purchase price of the same type of product of the same-income group, which is denoted as the average price level.

[0070] Step S33: Obtain the consumption price of the product purchased by the user, and calculate the price difference between the consumption price and the average price level.

[0071] Step S34: Collect the preferential information in the market, and find the maximum preferential reduction price of the same type of product according to the preferential information, which is denoted as the reduced price.

[0072] Step S35: Combine the economic income, the price difference, and the reduced price to calculate the user's economic willingness value.

[0073] Set the proportionality coefficients of economic income, price difference, and price reduction respectively, and calculate the economic willingness value based on the proportionality coefficients. When the user's economic income is higher, the user is more willing to directly replace the product. Therefore, from an economic perspective, the willingness to repair is lower, and the economic willingness value is even lower. When the price difference is larger, it means that the product exceeds the user's consumption level. Therefore, the willingness to repair will be greater, that is, the economic willingness value is greater. For example, if a user has a monthly income of 5,000 and purchases a refrigerator worth 10,000 yuan, which far exceeds the current income level of consumption, then when the product fails, the user will give priority to repair. If the user has a monthly income of 20,000 and purchases a washing machine worth 200 yuan, the price difference from the average price level is negative, which is far lower than the current income level of consumption. Therefore, the willingness to replace is greater, and the economic willingness value is smaller. When there is a discount on the product, the user's willingness to replace the product will also increase, and the willingness to repair will decrease. Therefore, the economic willingness value is even lower.

[0074] Step S36: Obtain the product information of similar products on the market, and analyze the user's demand willingness value based on the product information.

[0075] Step S37: Obtain the user's browsing information, and analyze the user's replacement willingness value based on the browsing information.

[0076] Step S38: Set the proportionality factors of the user's economic willingness value, demand willingness value, and replacement willingness value respectively, and calculate the user's repair willingness value based on the proportionality factors.

[0077] The greater the economic willingness value and the demand willingness value, the greater the user's willingness to repair; the greater the replacement willingness value, the smaller the user's willingness to repair.

[0078] In actual application, whether the user needs product repair information depends on two aspects: on the one hand, whether the user's product fails, and on the other hand, whether the user has the willingness to repair. If the user has no willingness to repair, then even if the product repair information is pushed, the user will not choose to repair the product, and the product repair information will instead cause interference to the user. Therefore, pushing product repair information according to the willingness to repair can reduce the interference caused by product repair information to the user.

[0079] The steps of obtaining the product information of similar products on the market and analyzing the user's demand willingness value based on the product information are specifically as follows: Step S361: Obtain the product information of similar products on the market, and extract the product functions according to the information of similar products on the market and record them as the first functions.

[0080] Step S362: Obtain the product information of the user's product, and extract the product functions according to the user's product information and record them as the second functions.

[0081] Step S363: Compare the first functions and the second functions, find the upgraded functions, and count the number of the upgraded functions.

[0082] For example, the second functions of the user product are washing, rinsing, and dehydration, while the similar products on the market have functions of washing, rinsing, dehydration, and drying. Then the upgraded function is drying. If the first function can only wash ordinary clothes but not down jackets, while the market products can wash down jackets, then increasing the types of clothes that can be washed is also an upgraded function.

[0083] Step S364: Count the proportion of the user product in the market products and record it as the second proportion.

[0084] Step S365: Set the proportionality coefficients of the function quantity and the second proportion respectively, and calculate the demand willingness value according to the proportionality coefficients.

[0085] In actual application, the more the number of upgraded functions, it indicates that the existing user products can no longer meet the general user needs. Therefore, when the product fails, the user's willingness to replace is greater, and the willingness to repair is lower. So, the more the number of functions, the lower the demand willingness value. And the smaller the proportion of the user product in the market products, it indicates that the probability of this product being phased out is greater. Therefore, the probability of the user replacing is greater, so the demand willingness value for repair is lower.

[0086] The steps to obtain the user's browsing information and analyze the user's replacement willingness value based on the browsing information are as follows: Step S371: Obtain the user's browsing information and determine whether the user browses similar products according to the browsing information.

[0087] Step S372: If the user browses similar products, count the browsing times and browsing duration of the user browsing similar products.

[0088] Step S373: Calculate the user's first replacement willingness value according to the browsing times and browsing duration. If the user does not browse similar products, the first replacement willingness value is 0.

[0089] Set the proportionality coefficients of the browsing times and browsing duration respectively, and calculate the user's first replacement willingness value according to the proportionality coefficients. Through the browsing information on the user's public platform, obtain the browsing situation of the user for similar products. If the browsing times are more and the browsing duration is longer, it indicates that the user's willingness to purchase new products is stronger, then the replacement willingness will be greater, that is, the first replacement willingness value is greater. And if the user does not browse similar products, it is default that the user currently has no replacement willingness.

[0090] Step S374: Collect and count the total number of historical repairs of the user, and collect the replacement times of the user's similar products.

[0091] Step S375: Calculate the ratio of the total number of historical repairs to the replacement times and record it as the user repair ratio.

[0092] Step S376: Statistically obtain the repair probability of the user for the fault type, and calculate the replacement willingness value of the user by combining the first replacement willingness value, the repair ratio, and the repair probability.

[0093] In practical applications, the repair probability of the user for different fault types can be statistically obtained based on the user's historical repair records, because different users have different handling methods for different faults. Some users will directly replace the product without repairing for major faults. According to the evaluated fault type, accumulate to obtain the repair probability of the user for possible fault types. If there is no repair probability in the historical repair records, then use the average repair probability of other users for this fault type as the user's repair probability.

[0094] Set the weight ratios of the first replacement willingness value, the repair ratio, and the repair probability respectively, and calculate the replacement willingness value of the user according to the weight ratios. The larger the repair ratio, the more willing the user is to repair, and the lower the replacement willingness value. The larger the repair probability, the more willing the user is to repair, and the smaller the replacement willingness value.

[0095] A product repair information push system based on user data, by applying the product repair information push method based on user data as described above, includes: A fault type module, which collects the product data of the user, obtains the user data of the user, and estimates the product fault type by combining the product data and the user data.

[0096] A fault probability module, which extracts the usage data of the product according to the product data, and evaluates the fault probability of the product according to the usage data of the product.

[0097] A repair willingness module, which obtains the economic data of the user, and evaluates the repair willingness value of the user by combining the fault type and the fault probability.

[0098] A repair push module, which determines whether the repair willingness value reaches a preset willingness threshold. If it reaches the preset willingness threshold, then push the product repair information corresponding to the fault type to the user.

[0099] The above are all the preferred embodiments of this application, and the protection scope of this application is not limited thereby. Therefore, all equivalent changes made according to the structure, shape, and principle of this application should be covered within the protection scope of this application.

Claims

1. A method for pushing product maintenance information based on user data, characterized in that: The following steps are involved: Collect the user's product data, obtain the user's user data, and combine the product data and user data to estimate the product failure type; Extract product usage data based on product data, and evaluate the failure probability of the product based on the product usage data; Obtain the user's economic data, and combine the fault type and fault probability assessment to obtain the user's maintenance willingness value; Determine whether the maintenance willingness value reaches the preset willingness threshold. If it reaches the preset willingness threshold, push product maintenance information corresponding to the fault type to the user.

2. The method for pushing product maintenance information based on user data according to claim 1, characterized in that: The steps of collecting the user's product data, obtaining the user's user data, and combining the product data and the user data to estimate the product failure type are specifically as follows: Extracting the user's search history based on the user data, judging whether the product is confirmed to be faulty based on the search history, and if the product is confirmed to be faulty, obtaining the product fault type based on the search history; If the product is not sure of the fault, the product maintenance standard and user usage time are extracted based on the product data; Extracting the user's product usage habits from the user data and recording them as usage habits, and analyzing them in combination with the product maintenance standards to obtain the first fault type; Obtain historical fault records of the product, and obtain the second fault type based on the historical fault records and the user's usage time; The first failure type and the second failure type are counted to obtain the product failure type.

3. The method for pushing product maintenance information based on user data according to claim 2, characterized in that: The step of extracting the user's product usage habits from the user data and recording them as usage habits, and analyzing the first fault type in combination with the product maintenance standard is specifically as follows: Extract the user's product usage habits from the user data and record them as usage habits; Extract maintenance habits based on product maintenance standards, and compare the differences between usage habits and maintenance habits; Extract product failures caused by not executing maintenance habits according to product maintenance standards and form a habit-failure table; According to the habit difference, the corresponding fault is searched in the habit-fault table to obtain the first fault type.

4. The method for pushing product maintenance information based on user data according to claim 3, characterized in that: The step of extracting the user's product usage habits from the user data and recording them as usage habits is specifically as follows: Obtain the user's historical maintenance records and extract historical product failures based on the historical maintenance records; According to the historical faults of the product, the corresponding habits are found in the habit-fault table, and all habits are combined to form the user's usage habits; If the user has no historical maintenance record, the usage habits of all users with historical maintenance records are obtained, and the habit with the highest frequency among similar habits is found as the standard habit; All types of standard habits are formed into a habit set as the usage habits of users without historical maintenance records.

5. The method for pushing product maintenance information based on user data according to claim 2, characterized in that: The step of obtaining the historical fault records of the product and obtaining the second fault type according to the historical fault records and the user's usage time is specifically as follows: Obtain historical fault records of similar products, extract the product usage time corresponding to different faults, calculate the average usage time of all corresponding products, and obtain the average usage time; A fault-duration table is formed based on the average usage duration corresponding to different faults; The user's usage time is obtained, and the fault corresponding to the user's usage time is searched according to the fault-time table to obtain the second fault type.

6. The method for pushing product maintenance information based on user data according to claim 5, characterized in that: The step of extracting product usage data based on product data and evaluating the product failure probability based on the product usage data is specifically as follows: The similarity between the usage habits and the maintenance habits is obtained by comparison and recorded as the habit similarity; According to the preset association curve between the habit similarity and the fault probability, the corresponding fault probability is found and recorded as the first fault probability; Determine whether the product has been repaired based on historical maintenance records. If the product has been repaired, obtain the repair time of the product; Obtain the real-time time point, calculate the duration between the real-time time point and the maintenance time point and record it as the real-time usage duration; If the product has not been repaired, the user's usage time will be used as the real-time usage time; Establish a correlation curve between real-time usage time and failure probability, find the corresponding failure probability according to the real-time usage time, and record it as the second failure probability; The failure probability of the product is calculated by combining the first failure probability and the second failure probability.

7. The method for pushing product maintenance information based on user data according to claim 1, characterized in that: The step of obtaining the economic data of the user and evaluating the maintenance willingness value of the user in combination with the fault type and the fault probability is specifically as follows: Obtain the economic income of the user, and record the people with the same economic income level as the user as the same income group; Get the average purchase price of the same type of products for people with the same income and record it as the average level price; Obtain the consumer price of the product purchased by the user, and calculate the price difference between the consumer price and the average level price; Collect the preferential information on the market, find the maximum preferential price reduction of similar products based on the preferential information and record it as the reduced price; The economic willingness value of the user is calculated by combining economic income, price difference and price reduction; Obtain product information of similar products on the market, and obtain the user's demand willingness value based on product information analysis; Obtain the user's browsing information, and obtain the user's replacement willingness value based on the browsing information analysis; The proportional factors of the user's economic willingness value, demand willingness value and replacement willingness value are set respectively, and the user's maintenance willingness value is calculated according to the proportional factors.

8. The method for pushing product maintenance information based on user data according to claim 7, characterized in that: The step of obtaining product information of similar products on the market and obtaining the user's demand willingness value based on product information analysis is specifically as follows: Obtain product information of similar products on the market, extract product functions based on the information of similar products on the market and record them as the first function; Obtain product information of the user's product, extract product functions based on the user's product information and record them as the second function; Compare the first function and the second function, find the upgraded function, and count the number of the upgraded functions; Count the proportion of user products in the market and record it as the second proportion; Set the proportional coefficients of the number of functions and the second proportion respectively, and calculate the demand willingness value based on the proportional coefficients.

9. The method for pushing product maintenance information based on user data according to claim 7, characterized in that: The step of obtaining the user's browsing information and obtaining the user's replacement willingness value based on the browsing information analysis is specifically as follows: Obtain the user's browsing information and determine whether the user browses similar products based on the browsing information; If the user browses similar products, the number of times and the browsing time of the user browsing similar products are counted; The user's first replacement intention value is calculated based on the number of views and the browsing time. If the user does not browse similar products, the first replacement intention value is 0; Collect and count the total number of historical repairs of users, and the number of replacements of similar products by users; Calculate the ratio of the total number of historical repairs to the number of replacements, and record it as the user repair ratio; The user's repair probability for the fault type is obtained by statistics, and the user's replacement willingness value is calculated by combining the first replacement willingness value, the repair ratio and the repair probability.

10. Product maintenance information push system based on user data, characterized in that: By applying the product maintenance information push method based on user data as described in any one of claims 1 to 9, the method comprises: The fault type module collects the user's product data, obtains the user's user data, and combines the product data and user data to estimate the product fault type; The failure probability module extracts product usage data based on product data and evaluates the product failure probability based on the product usage data; The maintenance willingness module obtains the user's economic data and combines the fault type and fault probability evaluation to obtain the user's maintenance willingness value; The maintenance push module determines whether the maintenance willingness value reaches a preset willingness threshold. If it reaches the preset willingness threshold, the product maintenance information corresponding to the fault type is pushed to the user.

Images