CN112632411A - Target object data query method, device, equipment and storage medium - Google Patents

Target object data query method, device, equipment and storage medium Download PDF

Info

Publication number
CN112632411A
CN112632411A CN202011553754.1A CN202011553754A CN112632411A CN 112632411 A CN112632411 A CN 112632411A CN 202011553754 A CN202011553754 A CN 202011553754A CN 112632411 A CN112632411 A CN 112632411A
Authority
CN
China
Prior art keywords
current
rule
object data
target object
query
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011553754.1A
Other languages
Chinese (zh)
Inventor
程皓
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wuhan Kuangshi Jinzhi Technology Co ltd
Beijing Megvii Technology Co Ltd
Original Assignee
Wuhan Kuangshi Jinzhi Technology Co ltd
Beijing Megvii Technology Co Ltd
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 Wuhan Kuangshi Jinzhi Technology Co ltd, Beijing Megvii Technology Co Ltd filed Critical Wuhan Kuangshi Jinzhi Technology Co ltd
Priority to CN202011553754.1A priority Critical patent/CN112632411A/en
Publication of CN112632411A publication Critical patent/CN112632411A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9536—Search customisation based on social or collaborative filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation

Abstract

The application provides a target object data query method, a device, equipment and a storage medium, wherein the method comprises the following steps: receiving a query instruction for a target object, the query instruction comprising: the number of expected target objects and the rule content of the at least one activity rule; acquiring initial rule parameters; the initial rule parameters are contained in a query instruction, or the initial rule parameters are given by a system running the query method; taking the initial rule parameter as a current rule parameter; and (3) query step: constructing a current query condition based on the rule content and the current rule parameter; processing object data based on the current query condition to obtain current target object data; and obtaining final target object data according with the expected target object quantity based on the current target object data. The method and the device are based on data automation mining analysis, and actively push out the information of the target object which most possibly meets the query requirement, so that the object identification efficiency is improved.

Description

Target object data query method, device, equipment and storage medium
Technical Field
The application relates to the technical field of big data, in particular to a target object data query method, a device, equipment and a storage medium.
Background
Big data refers to a data set which cannot be captured, managed and processed by a conventional software tool within a certain time range, and massive, high-growth rate and diversified information assets which can have stronger decision-making power, insight discovery power and flow optimization capability only by a new processing mode are needed. With the development of informatization, large data has been applied to various fields. In the field of target object data query, related object activity rule information is generally obtained based on big data analysis, deep mining analysis is performed on the related object activity rule information, an activity rule of a target object is found, an object query condition is formed, and then an object with the activity rule meeting the object query condition is used as the target object.
In the object query scheme, a clear object query condition needs to be set. Typically, the user knows the order of magnitude of the target object based on a priori knowledge, although the exact number of target objects is not known. If the target object inquired through the inquiry condition set by the user meets the expected magnitude order of the user, the inquiry condition is reasonably set.
However, in an actual application environment, specific parameters in the object query condition are often difficult to set. For example, the object query condition is "occur N times and more somewhere; or how many consecutive days appear/do not appear in a designated place for more than N days ", it is difficult for a user to define the value of N at the beginning when querying the target object data, and only a certain N value can be set heuristically, which often causes the number of target objects returned after the certain N value is set heuristically to be not in line with expectations, and needs to adjust and try the N value many times, resulting in poor user experience.
Disclosure of Invention
The embodiments of the present application aim to provide a target object data query method, a target object data query device, and a storage medium, which can push target object data meeting an expected number of objects when target object data meeting the expected number of objects cannot be provided by using initial parameters based on automatic adjustment of a target object data query device on query conditions or adjustment of a result push manner, so that target object data query efficiency is improved, and user experience is improved.
A first aspect of the embodiments of the present application provides a target object data query method, including: receiving a query instruction for a target object, the query instruction comprising: the number of expected target objects and the rule content of the at least one activity rule; acquiring initial rule parameters; the initial rule parameters are contained in a query instruction, or the initial rule parameters are given by a system running the query method; taking the initial rule parameter as a current rule parameter; and (3) query step: constructing a current query condition based on the rule content and the current rule parameter; processing object data based on the current query condition to obtain current target object data; and obtaining final target object data according with the expected target object quantity based on the current target object data.
In an embodiment, the processing the object data based on the current query condition to obtain the current target object data includes: and taking the object data with the activity rule meeting the current query condition as the current target object data.
In an embodiment, the obtaining final target object data according to the expected target object quantity based on the current target object data includes: if the current target object quantity corresponding to the current target object data meets the expected target object quantity, taking the current target object data as the final target object data; if the number of the current target objects does not accord with the expected number of the target objects, the current rule parameter of at least one active rule is updated, and the query step is executed again.
In an embodiment, the processing the object data based on the current query condition to obtain the current target object data includes: sorting the object data according to the matching degree of the current query condition to obtain current target object data; obtaining final target object data according with the expected target object quantity based on the current target object data, wherein the final target object data comprises the following steps: and taking the current target object data of the expected target object number as final target object data.
In an embodiment, the sorting the object data according to the matching degree to the current query condition includes: according to the rule content, counting the object data to obtain actual parameters of each object in the object data; calculating the matching degree of each object and the current query condition according to the actual parameters and the current query parameters; and sequencing the objects according to the matching degree.
In one embodiment, the query instruction further includes: a rule range of the activity rule; the constructing of the current query condition based on the rule content and the current rule parameter includes: and constructing a current query condition based on the rule content, the current rule parameter and the rule range.
In one embodiment, the query instruction further includes: an object appearance time range and/or an object appearance place range; the processing object data based on the current query condition includes: and processing the object data of the objects with the appearance time within the object appearance time range and/or the appearance place within the object appearance place range based on the current query condition.
In one embodiment, the object data is pre-statistical snapshot data; the step of pre-counting the snapshot data comprises the following steps: grouping the snapshot data of the same object into a same object image set; the snapshot data correspond to the snapshot time and the snapshot place of the snapshot data; and counting the appearance positions of the objects in the object image set according to the snapshot positions.
A second aspect of the embodiments of the present application provides a target object data query apparatus, including: a receiving module, configured to receive a query instruction for a target object, where the query instruction includes: the number of expected target objects and the rule content of the at least one activity rule; the acquisition module is used for acquiring initial rule parameters; the initial rule parameters are contained in a query instruction, or the initial rule parameters are given by a system running the query method; the determining module is used for taking the initial rule parameter as a current rule parameter; a query module for querying: constructing a current query condition based on the rule content and the current rule parameter; processing object data based on the current query condition to obtain current target object data; and obtaining final target object data according with the expected target object quantity based on the current target object data.
In one embodiment, the query module is configured to: and taking the object data with the activity rule meeting the current query condition as the current target object data.
In one embodiment, the query module is configured to: if the current target object quantity corresponding to the current target object data meets the expected target object quantity, taking the current target object data as the final target object data; if the number of the current target objects does not accord with the expected number of the target objects, the current rule parameter of at least one active rule is updated, and the query step is executed again.
In one embodiment, the query module is configured to: sorting the object data according to the matching degree of the current query condition to obtain current target object data; obtaining final target object data according with the expected target object quantity based on the current target object data, wherein the final target object data comprises the following steps: and taking the current target object data of the expected target object number as final target object data.
In an embodiment, the sorting the object data according to the matching degree to the current query condition includes: according to the rule content, counting the object data to obtain actual parameters of each object in the object data; calculating the matching degree of each object and the current query condition according to the actual parameters and the current query parameters; and sequencing the objects according to the matching degree.
In one embodiment, the query instruction further includes: a rule range of the activity rule; the query module is configured to: and constructing a current query condition based on the rule content, the current rule parameter and the rule range.
In one embodiment, the query instruction further includes: an object appearance time range and/or an object appearance place range; the query module is configured to: and processing the object data of the objects with the appearance time within the object appearance time range and/or the appearance place within the object appearance place range based on the current query condition.
In one embodiment, the object data is pre-statistical snapshot data; the device also comprises a pre-counting module, which is used for pre-counting the snapshot data and comprises: grouping the snapshot data of the same object into a same object image set; the snapshot data correspond to the snapshot time and the snapshot place of the snapshot data; and counting the appearance positions of the objects in the object image set according to the snapshot positions.
A third aspect of embodiments of the present application provides an electronic device, including: a memory to store a computer program; the processor is configured to execute the method of the first aspect and any embodiment of the present application to query target object data that meets the query instruction.
A fourth aspect of embodiments of the present application provides a non-transitory electronic device-readable storage medium, including: a program which, when run by an electronic device, causes the electronic device to perform the method of the first aspect of an embodiment of the present application and any embodiment thereof.
The target object data query method, device, equipment and storage medium provided by the application receive a query instruction input by a user, wherein the query instruction comprises: the number of expected target objects and the rule content of the at least one activity rule; then obtaining initial rule parameters; the initial rule parameters are contained in a query instruction, or the initial rule parameters are given by a system running the query method; taking the initial rule parameter as a current rule parameter; constructing a current query condition based on the rule content and the current rule parameter; processing object data based on the current query condition to obtain current target object data; and obtaining final target object data according with the expected target object quantity based on the current target object data. Therefore, the information of the target object which most possibly meets the query requirement is actively pushed out based on the mining analysis of data automation, and the object query efficiency is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a schematic structural diagram of an electronic device according to an embodiment of the present application;
FIG. 2 is a flowchart illustrating a target object data query method according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of a target object data query device according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. In the description of the present application, the terms "first," "second," and the like are used solely to distinguish one from another and are not to be construed as indicating or implying relative importance.
As shown in fig. 1, the present embodiment provides an electronic apparatus 1 including: at least one processor 11 and a memory 12, one processor being exemplified in fig. 1. The processor 11 and the memory 12 are connected by a bus 10. The memory 12 stores instructions executable by the processor 11, and the instructions are executed by the processor 11, so that the electronic device 1 may execute all or part of the processes of the methods in the embodiments described below, so as to identify the target object in the target area, which meets the determination rule.
In an embodiment, the electronic device 1 may be a mobile phone, a tablet computer, a notebook computer, a desktop computer, or an operating system formed by multiple computers.
Please refer to fig. 2, which is a method for querying target object data according to an embodiment of the present application, and the method may be executed by the electronic device 1 shown in fig. 1 and may be applied in an object query scenario based on big data, so as to query the target object data meeting requirements from the object data to be queried based on a query instruction.
The method comprises the following steps:
step S11: receiving a query instruction for a target object, the query instruction comprising: the number of target objects and the rule content of the at least one activity rule are expected.
The object corresponds to object data, the object data to be queried may be full object data which can be acquired, and the target object data may be target object data which is queried from the object data to be queried and meets requirements. In one example, the object data includes data corresponding to an object obtained by clustering captured images according to object features. For example, when the object is a person, a plurality of face images are captured in a city, the captured face images can be clustered according to face features extracted from the face images, and a data set corresponding to the person obtained by clustering is used as person data. In another example, the object data includes pre-statistical data of the object, and the pre-statistical data of the object is data obtained by pre-counting data corresponding to the object, which is obtained by clustering the captured images according to the object features. And taking the data corresponding to the object obtained by clustering as clustering data, wherein the clustering data comprises at least one snapshot image of the object, the pre-statistics can be preliminary statistics according to the snapshot location and the snapshot time of each snapshot image in the clustering data, and further statistics can be carried out according to the activity rule on the basis of the data obtained by the pre-statistics, so that the matching degree of the object data and the activity rule is judged. For example, the total number of the snap-shot images in the cluster data of the object a is 5, and the snap-shot time and the snap-shot place are respectively as follows:
1 month and 1 day at 21 o' clock in Hotel H
School with 1 month, 3 days and 1 point in S
Hotel H at 12 o' clock 2 months and 1 day
Hotel H at 22 o' clock in 1/3
5 month, 1 day, 2 o' clock in M museum
The pre-statistics is to count the times and the capturing time of capturing at each capturing place, and the pre-statistical data obtained by pre-counting the clustering data is as follows:
h Hotel 3 times of 21 o ' clock in 1 month and 1 day, 12 o ' clock in 1 month and 2 days and 22 o ' clock in 1 month and 3 days
S school
1 time of 1 month, 3 days and 1 point
M museum
1 time of 5 month, 1 day and 2 o' clock
The activity rule is 1 month and 1 day to 1 month and 20 days, and the number of the times of the stay in and the stay out of any hotel and the hotel from 21 to 22 points per day is more than or equal to 3 days.
On the basis of the data obtained by pre-statistics, further statistics can be carried out according to activity rules, 3 times meeting the condition that any hotel is photographed at the H hotel are screened out, and among the 3 times, the number of times of being caught from 1 month to 20 months from 1 month to 1 month and from 21 points to 22 points per day is further screened out to be 1 time, so that the number of days that the object is caught from 1 month to 20 months from 1 month to 1 month is judged, the number of days that the number of times of being caught from 21 points to 22 points per day at any hotel is not less than 3 is 0 day, and the object data is not matched with the activity rules.
Therefore, the clustering data are subjected to the statistics, partial operation can be performed before the query instruction is received, and the matching degree of the object data and the activity rule can be quickly determined after the query instruction is received.
In one example, the query instruction includes at least one activity rule, and the activity rule reflects an activity rule of the object, such as a presence condition, an appearance or departure condition, a peer condition, and the like of the object. Each activity rule corresponds to a rule content, and activity rules of the same kind correspond to the same rule content. Each activity rule of the at least one activity rule may be of the same kind or of a different kind. When the user selects one or more rules, the rule content corresponding to the one or more rules is determined. The activity rule contains the rule content.
For example, the activity rule may be: the number of days from 1 month and 1 day to 1 month and 20 days, from 21 to 22 points per day, when the number of times of the hotel is more than or equal to 3, is more than or equal to 3 days,
the rule content corresponding to the activity rule is as follows: the number of days that the number of times of absence is more than or equal to P times in a certain time period is more than or equal to Q days.
The query instruction also includes a desired target object number W. The expected number of target objects can generally be determined by the user based on a priori knowledge, and the number of target objects that the user wishes to query is close to or at least of the same order of magnitude as W. For example, a user knows that the number of stolen persons in the district is 100, and wants to find out who the stolen persons are. The user wishes to approximate 100 people by the people contained in the query results returned based on the query instructions.
Step S12: initial rule parameters are obtained.
The rule parameters are parameters as count thresholds in the activity rule, such as the number of days threshold Q and the number of times threshold P in the previous example. Some types of activity rules contain rule parameters and some types of activity rules do not. A rule such as "first time in town between 1 month and 20 days" does not contain a rule parameter. In each activity rule contained in the query instruction of the method, at least one activity rule contains a rule parameter.
The initial rule parameter may be set by the user, included in the query instruction, or may be a default parameter that does not need to be set by the user. The default parameters may be fixed or random initial values, or may be given by the system running the query method based on a priori information. In the prior art, the rule parameters are set by the user. However, the user does not determine how well the rule parameters are set, and the random setting often causes the number of returned results to be greatly different from the expected number, and at this time, the rule parameters need to be readjusted for query again. In the embodiment, the default rule parameters can be given by the system without setting the rule parameters by a user; even if the user sets the rule parameters in the query instruction, the rule parameters set by the user are only used as initial rule parameters, and the initial rule parameters can be automatically adjusted in the query process, so that the number of results meets the expectation of the user.
Step S13: and taking the initial rule parameter as the current rule parameter.
And taking the initial rule parameter as the current rule parameter before the query step aiming at the query instruction is executed for the first time.
Step S14: a query step is performed.
The querying step includes steps S141-S143.
Step S141: and constructing a current query condition based on the rule content and the current rule parameters.
First, a current activity rule is constructed based on the rule content and current rule parameters.
For example, after acquiring the rule content "a certain time slot, the number of days with the number of times of absence or presence at a certain location being equal to or greater than P is equal to or greater than Q days" and the current rule parameter "the number of days threshold Q is 4 and the number of times threshold P is 3", the current activity rule may be constructed as "a certain time slot, and the number of days with the number of times of absence or presence at a certain location being equal to or greater than 3 is equal to or greater than 4 days".
In one example, the current activity rule may also contain a rule range. Rule scope refers to parameters in an activity rule that define a time and/or place scope. For example, in the activity rule of "1 month 1 day to 1 month 20 days, 21 o 'clock to 22 o' clock each day in the days of more than or equal to 3 times of the number of times of the presence or absence of any hotel or hotel" more than or equal to 3 days, the parameter for defining the time range is "1 month 1 day to 1 month 20 days, 21 o 'clock to 22 o' clock each day, and the parameter for defining the location range is" any hotel or hotel ". The rule range may be included in the query instruction, and if the rule range is obtained from the query instruction, the current query condition may be constructed based on the rule content, the rule range, and the current rule parameter. For example, when rule content "a certain time slot, days with the number of times of absence or presence at a certain place being more than or equal to P times being more than or equal to Q days" is acquired, the rule range "21 o 'clock to 22 o' clock every day from 1 month 1 day to 1 month 20 days" and "any hotel" included in the query instruction are acquired, and after the current rule parameter "day threshold Q is 4 and number threshold P is 3", the current activity rule may be constructed as "21 o 'clock to 22 o' clock every day from 1 month 1 day to 1 month 20 days, and days with the number of times of absence or presence at any hotel being more than or equal to 3 times being more than or equal to 4 days".
Secondly, current query conditions are constructed based on the current activity rules and the relations among the activity rules.
If the query instruction contains a plurality of activity rules, the query instruction also contains the relationship among the plurality of activity rules. The relationship between the plurality of activity rules may be any logical operational relationship, such as AND, OR, NOT. After constructing the current activity rule A, B, C, the current query condition may be that the current activity rule a, the current activity rule B, and the current activity rule C are satisfied at the same time, or that the current query condition may be that the current activity rule a or the current B is satisfied and the current activity rule C is not satisfied.
Step S142: and processing the object data based on the current query condition to obtain current target object data.
As previously mentioned, the object data is cluster data or pre-statistical data of the object. After the current query condition is obtained, the cluster data or the pre-statistical data of the object to be queried are counted according to each current activity rule contained in the current query condition, so that the matching degree of the object data and each current activity rule is determined, and further the matching degree of the object data and the current query condition is determined.
The degree of match may be a boolean value. For example, the current activity rule is "21 to 22 points per day from 1 month 1 to 1 month 20 days, and the number of times of hotel presence or absence is 3 or more and 4 or more days". If the object data is counted to find that the object has 21 to 22 points per day from 1 month and 20 days, and the number of the days when the number of the hotel exits or the like is more than or equal to 3 is 2 days, the matching degree of the object and the current activity rule is 0. If the data of the object is counted, the object is found to be 21 to 22 points per day from 1 month and 1 month to 20 days, and the number of the days when the number of the hotel exits or exits is more than or equal to 3 in any hotel is 5 days, the matching degree of the object and the current activity rule is 1. The matching degree of the object data and the current activity rule is a boolean value, and further, the matching degree of the object data and the current query condition is also a boolean value, which can be obtained by performing a logical operation corresponding to the relationship between the activity rules between the matching degrees of the current activity rules.
The degree of matching may also be a continuous value calculated in a particular way. For example, the current activity rule is "21 to 22 points per day from 1 month and 1 to 20 months, and the number of times of hotel presence is 3 or more and 4 days. If the data of the object is counted and found that the object is 21 to 22 points per day from 1 month and 20 days, and the number of the days when the number of the hotel exits or the like is more than or equal to 3 is 2 days, the matching degree of the object and the current activity rule is 2/4 to 0.5. Or, if the object data is counted to find that the object is 21 to 22 points per day from 1 month 1 to 1 month 20, and when the number of times of the hotel presence or absence is not less than N and the number of days is 4 days, N is 2, the matching degree of the object and the current activity rule is 2/3 to 0.67. Of course, the calculation method of the matching degree is not limited to this. The matching degree of the object data and the current activity rule is a continuous value, and further, the matching degree of the object data and the current query condition is a continuous value.
When the matching degree is a boolean value, processing the object data based on the current query condition refers to screening out object data whose matching degree with the current query condition is 1, and taking the screened out object data as the current target object data.
And when the matching degree is a continuous value, processing the object data based on the current query condition, namely sorting the object data to be queried according to the matching degree with the current query condition, and taking the sorted object data as the current target object data.
Step S143: and obtaining final target object data according with the expected target object quantity based on the current target object data.
And when the matching degree is a Boolean value, determining the quantity of the current target object data after obtaining the current target object data, and if the quantity of the current target object data accords with the expected target object quantity, returning the current target object data serving as final target object data to the user, so that the user can obtain a result according with the expected quantity.
When the matching degree is a continuous value, after the current target object data is obtained, the current target object data with the number of the previous expected target objects with the highest matching degree is selected as the final target object data and returned to the user, so that the user can obtain a result according with the expected number.
Therefore, the user can obtain expected number of results under the condition of only setting the initial rule parameters once or not setting any initial rule parameters, and the user experience is greatly improved.
In one embodiment, the matching degree is a boolean value, and the method for querying the target object data from the object data to be queried includes:
step S21, receiving a query instruction aiming at the target object, wherein the query instruction comprises the following steps: the number of expected target objects and the rule content of the at least one activity rule;
step S22, obtaining initial rule parameters; the initial rule parameters are contained in a query instruction, or the initial rule parameters are given by a system running the query method;
step S23, taking the initial rule parameter as the current rule parameter;
step S24, the query step is performed. The querying step includes steps S241-S243.
Step S241, constructing a current query condition based on the rule content and the current rule parameter;
and step S242, screening the object data to be inquired based on the current inquiry condition to obtain the current target object data. That is, object data whose matching degree with the current query condition is 1 is screened out, and the screened-out object data is set as the current target object data.
Step S243, if the current target object quantity corresponding to the current target object data accords with the expected target object quantity, the current target object data is used as the final target object data; if the number of the current target objects does not meet the expected number of the target objects, the current rule parameter of at least one active rule is updated, and the query step is executed again in S24.
It can be understood that the current query condition is constructed by using the current rule parameters, and when the obtained number of the current target objects does not meet the expected number of the target objects, the current rule parameters and the actual situation are shown to be in or out, and at this time, the current rule parameters can be adjusted based on the difference value between the number of the current target objects and the expected number of the target objects. The rule parameters comprise positive correlation rule parameters positively correlated with the quantity of the target objects and/or negative correlation rule parameters negatively correlated with the quantity of the target objects. The positive correlation rule parameter means that the number of target objects is increased along with the increase of the rule parameter; a negatively correlated rule parameter means that as the rule parameter increases, the number of target objects decreases. If the current target number is smaller than the expected target number, the current rule parameters are updated in the following mode: at least one positively correlated current rule parameter is increased and/or at least one negatively correlated current rule parameter is decreased. If the current target number is larger than the expected target number, the current rule parameters are updated in the following mode: the at least one positively correlated current regulation parameter is decreased and/or the at least one negatively correlated current regulation parameter is increased.
The process of updating the rule parameters is automatically completed by the system without the participation of a user. When updating the rule parameters, each rule parameter can be updated one by one, or a plurality of rule parameters can be updated once, and a current query condition is generated once each update. Each rule parameter may have a corresponding change step size and/or update interval, and when the rule parameter is increased or decreased, the minimum unit of increase or decrease is the change step size, and the upper and lower limits of increase or decrease are the update interval. Each rule parameter may have its corresponding priority, e.g., a rule parameter with a higher priority may be updated preferentially. Each rule parameter may also have its corresponding sensitivity. For example, when the rule parameter with high sensitivity is changed, the current target object number variation is large, and when the rule parameter with low sensitivity is changed, the current target object number variation is small.
The updating process of the rule parameters can be various possible combinations of values of the rule parameters which are traversed one by one, or the current rule parameters to be updated and the updating step length can be determined by a gradient descent method and other methods based on the difference value between the number of the current target objects and the number of the expected target objects. For example, when the difference between the current number of target objects and the expected number of target objects is large, the parameter that is preferentially updated is a parameter with high sensitivity, and the update step size is large, whereas the parameter that is preferentially updated is a parameter with low sensitivity, and the update step size is small. In summary, it is desirable to quickly converge the current target object number to the desired target object number by updating the current rule parameters.
Therefore, the embodiment of the application can automatically update the current rule parameters when the number of the current target objects is not in line with the expectation, rebuild the current query conditions based on the updated current rule parameters, screen out the current target object data from the object data to be queried by using the current query conditions, and display the current target object data as the final target object data to the user when the number of the current target objects queried by using the current query conditions is in line with the number of the expected target objects. In this way, when the initial rule parameters given by the user or the system cannot obtain the expected target object number, the parameters can be automatically adjusted and the expected result can be returned without perception of the user, and the user does not need to try and adjust the parameters.
In another embodiment, the matching degree is a continuous value, and the method for querying the target object data from the object data to be queried comprises the following steps:
step S31, receiving a query instruction aiming at the target object, wherein the query instruction comprises the following steps: the number of expected target objects and the rule content of the at least one activity rule;
step S32, obtaining initial rule parameters; the initial rule parameters are contained in a query instruction, or the initial rule parameters are given by a system running the query method;
step S33, taking the initial rule parameter as the current rule parameter;
step S34, the query step is performed. The querying step comprises steps S341-S343.
Step S341, constructing a current query condition based on the rule content and the current rule parameter; the current query condition may include a plurality of current activity rules.
Step S342: and sequencing the object data to be queried according to the matching degree with the current query condition, and taking the sequenced object data to be queried as the current target object data.
Firstly, the matching degree of the object data to be inquired and the current inquiry condition is calculated.
In one example, the matching degree between the object data to be queried and the current query condition can be calculated as follows:
counting the data of the object to be inquired according to the rule content of the current activity rule contained in the current inquiry condition to obtain the actual parameters of the data of the object to be inquired under the rule content of the current activity rule.
Specifically, the data of the object to be queried can be counted according to the rule content and the rule range of the current activity rule, so as to obtain the actual parameters of the data of the object to be queried under the current activity rule. For example, the current activity rule is "21 to 22 points per day from 1 month and 20 days per day, the number of days for which the number of times of presence or absence of any hotel is greater than or equal to 3 is greater than or equal to 4 days", the statistics on the data of the object to be queried may be from 21 to 22 points per day from 1 month and 20 days per day, the number of days for which the number of times of presence or absence of any hotel is greater than or equal to 3 times, or from 21 to 22 points per day from 1 month and 20 days per day, and when the number of times of presence or absence of any hotel is greater than or equal to N is greater than or equal to 4 days, the maximum value of N is the actual parameter at this time, which is the number of days or the value.
Thus, the actual parameters corresponding to each current activity rule can be obtained.
Secondly, calculating the matching degree of the object data to be inquired and the current inquiry condition according to the actual parameters corresponding to the current activity rules and the current rule parameters corresponding to the current activity rules.
In one example, the actual parameter vector is formed by the actual parameter, the current rule parameter vector is formed by the current rule parameter, and the distance between the actual parameter vector and the current rule parameter vector is used as the matching degree of the object data to be queried and the current query condition. For example, the current query condition includes 3 current activity rules, thereby constructing an actual parameter vector (1, 2, 3) consisting of 3 actual parameters corresponding to the 3 current activity rules and a current rule parameter vector (3, 2, 3) consisting of 3 current rule parameters corresponding to the 3 current activity rules, according to the distance ((1-3) between the actual parameter vector and the current rule parameter vector2+(2-2)2+(3-3)2)1/2And determining the matching degree of the data of the object to be queried and the current query condition. The closer the distance, the higher the degree of matching.
In another example, the actual parameter and the current rule parameter may be mapped into a coordinate system, and a distance between a coordinate point corresponding to the actual parameter and a coordinate point corresponding to the current rule parameter in the coordinate system may be used as a matching degree between the object data to be queried and the current query condition. For example, if the actual parameter and the current rule parameter include three parameters, the actual parameter and the current rule parameter of each object in the object data may be projected to the three coordinate axes according to three coordinate axes X, Y, and Z, respectively, to obtain an actual parameter corresponding point and a current rule parameter corresponding point, and the matching degree between the object data to be queried and the current query condition may be determined according to the distance between the two points. The closer the distance, the higher the degree of matching.
In another example, the matching degree of the object data to be queried and each current activity rule is subjected to weighted summation to obtain the matching degree of the object data to be queried and the current query condition. For example, when the actual parameters corresponding to the 3 current activity rules are (1, 2, 3), the current rule parameters corresponding to the 3 current activity rules are (3, 2, 3), the matching degree between the object data to be queried and the current rule parameters is (1/3, 2/2, 3/3), and the weights of the three current activity rules are 2, 1, respectively, the three current activity rules are weighted and summed to obtain the matching degree of 2 × 1/3+1 × 2/2+1 × 3/3.
And secondly, sorting the object data to be inquired according to the matching degree of the object data to be inquired and the current inquiry condition, and taking the result obtained after sorting as the current target object data.
Step S243: and selecting current target object data with the number of previous expected target objects with the highest matching degree with the current query condition from the current target object data as final target object data.
Assuming that the expected number of target objects is E, the current target object data ranked in the top E is taken as the final target object data. For example, in the coordinate system, E objects corresponding to E actual parameter corresponding points closest to the current rule parameter corresponding point may be taken as target objects, and object data corresponding to the E objects may be final target object data.
In one example, the query instruction further comprises: the appearance time range and/or the appearance place range of the object to be inquired.
For example, the query instruction includes the appearance time range T1 and/or the appearance place range P1 of the object to be queried.
The appearance time range of the object may be understood as a range within which the snapshot time of the snapshot image of the object falls. The appearance time range of the object can be understood as a range within which the snapshot location of the snapshot image of the object falls.
In one example, objects appearing in the range of T1 and P1 can be used as the objects to be queried. That is, if an object appears at T1 or P1, but also appears at other times or places, the object is an object to be queried. If the object does not appear at T1, P1, then the object is not the object to be queried.
In another example, the data of each object in the range of T1 and P1 may be used as the object data to be queried. For example, if some objects appear at T1 and P1, and also appear at other times and places, the data of the objects captured at T1 and P1 are used as the data of the objects to be queried, and the data of the objects captured at other times and places are used as the data of the objects not to be queried.
Therefore, the occurrence time and the occurrence place of the object to be inquired can be limited in the inquiry instruction, so that the data range of the data to be inquired can be reduced during inquiry, and the data processing amount is reduced.
In one example, the current rule parameters are stored in response to obtaining final target object data that meets the expected number of target objects.
It is understood that the current rule parameter at this time is a rule parameter that is verified to have an expected number of results, and has a certain reasonableness, and may be stored for later use as an initial rule parameter, a system default rule parameter, or a proposed rule parameter.
In one example, when the number of updates to the current rule parameter exceeds a preset threshold, the query is stopped. The number of updates may be determined according to the number of activity rules included in the query, the latency acceptable to the user, and the like.
If the number of target objects which is matched with the expected number cannot be obtained through multiple updates, on one hand, the system consumes a great amount of computing power, on the other hand, the expected number of target objects is possibly incorrect, and at the moment, the updating and inquiring process can be stopped, and wrong results can be fed back to the user.
In an embodiment, in practical use, the method of this embodiment may also be configured to perform the calculation in a specific time (e.g. 8 hours at 23-7 o' clock night when the user does not substantially use the system) so as to reduce the occupation of calculation power during the peak use of the system. It is understood that if the expected target object number range is not reached in one calculation cycle, the calculation may be stopped and continued until a second calculation cycle is reached.
The method for querying the target object data from the object data to be queried according to an embodiment of the present application may be executed by the electronic device 1 shown in fig. 1, and may be applied to a person finding scene based on big data, so as to identify a target object meeting a query condition in a target area based on a query instruction. The method comprises the following steps:
1. and butting each service scene data, and performing structured storage on the service scene data by using a big data technology so as to facilitate the analysis and application of the following big data.
2. And establishing an atomicity activity rule, such as various atomicity activity rules of a presence situation, a perception presence/departure, multi-region loitering, a peer and the like, wherein the activity rule can be refined according to the common situation summarized by the user and is not limited to the above.
In case of emergence, for example, in a certain specific time period, at least N days appear in a certain place, and M times and more appear every day;
presence/absence sensing: present/absent at the designated location for more than N consecutive days.
Multi-zone loitering: a segment occurs in at least N suspect regions;
in the same way: more than M times with at least N known suspicious persons.
3. Aiming at a business scene, the atomic activity rules are arranged and combined to form a complex query condition, and corresponding thresholds in the activity rules are configured to perform target personnel query.
4. And setting the expected number of matched persons expected to meet the set query condition.
5. And 3, calculating the service scene data according to the query conditions configured in the step 3 by the system to obtain the actual number of matched personnel.
6. And comparing the actual number of matched personnel calculated in the last step with the expected number of matched personnel, and automatically adjusting the parameters by the system. And when the actual number of matched persons is lower than the expected number of matched persons, the threshold value of the activity rule is adjusted to be low, and when the actual number of matched persons is higher than the expected number of matched persons, the threshold value of the activity rule is adjusted to be high.
7. And the system calculates again to obtain a new actual matched personnel number based on the automatically adjusted activity rule threshold value.
8. And (5) continuously repeating the steps 6 and 7, and finally enabling the number of the personnel matched with the query condition to reach the number expected by the user, thereby completing calculation and displaying the actually matched personnel to the user through a visual interface.
The method for querying the target object data from the object data to be queried according to another embodiment of the present application may be executed by the electronic device 1 shown in fig. 1, and may be applied to a person-finding scene based on big data, so as to identify a target object meeting a query condition in a target area based on a query instruction. The method comprises the following steps:
1. and butting each service scene data, and performing structured storage on the service scene data by using a big data technology so as to facilitate the analysis and application of the following big data.
2. And establishing an atomicity activity rule.
3. Aiming at a business scene, the atomic activity rules are arranged and combined to form a complex query condition, and corresponding thresholds in the activity rules are configured to perform target personnel query.
For example, when the target person is a stolen person, the query condition is that activity rule 1 and activity rule 2 are satisfied:
activity rule 1: at least N days appear in more than 100 cells belonging to the Haihai district between 11 o 'clock and 5 o' clock of the next day, and M times and more appear every day;
activity rule 2: which occurred in at least K areas in which burglary had occurred in the last month.
For example, when the target person is a ticket seller, the query condition is that the activity rule 3 and the activity rule 4 are satisfied:
activity rule 3: the method comprises the following steps of (1) continuously appearing in places known to sell tickets for more than H days;
activity rule 4: more than Q times concurrent with at least P known vendors.
4. And setting the expected number of matched persons expected to meet the inquiry conditions of the thieves and the ticket sellers, such as 5-20 thieves and 1-10 ticket sellers.
5. The system calculates 0 stolen persons and 500 ticket sellers according to the rules configured in the step 3.
6. Based on the comparison between the result data calculated in the last step and the expected number of matched people, the system automatically adjusts the parameters. If the activity rule thresholds N, M, K in the query condition of the thief are all adjusted to be N-1, M-1 and K-1, the activity rule thresholds H, P, Q in the query condition of the ticket vendor are all adjusted to be H +1, P +1 and Q + 1.
7. And (4) calculating again to obtain 0 stolen persons and 170 ticket sellers based on the threshold value automatically adjusted by the system.
8. And (5) continuously repeating the steps 6 and 7, and finally enabling the number of the personnel matched with the query condition to reach the number expected by the user, thereby completing the calculation and displaying the calculation result to the user through a visual interface.
Please refer to fig. 3, which is a target object data query apparatus 600 according to an embodiment of the present application, which is applied to the electronic device 1 shown in fig. 1 and can be applied to a person-finding scene based on big data to identify a target object meeting a determination rule in a target area based on a query instruction. The device includes: the receiving module 601, the obtaining module 602, the determining module 603 and the querying module 604, the principle relationship of each module is as follows:
a receiving module 601, configured to receive a query instruction for a target object, where the query instruction includes: the number of expected target objects and the rule content of the at least one activity rule; an obtaining module 602, configured to obtain an initial rule parameter; the initial rule parameters are contained in a query instruction, or the initial rule parameters are given by a system running the query method; a determining module 603, configured to use the initial rule parameter as a current rule parameter; a query module 604 for querying the steps of: constructing a current query condition based on the rule content and the current rule parameter; processing object data based on the current query condition to obtain current target object data; and obtaining final target object data according with the expected target object quantity based on the current target object data.
In one embodiment, the query module 604 is configured to: and taking the object data with the activity rule meeting the current query condition as the current target object data.
In one embodiment, the query module 604 is configured to: if the current target object quantity corresponding to the current target object data meets the expected target object quantity, taking the current target object data as the final target object data; if the number of the current target objects does not accord with the expected number of the target objects, the current rule parameter of at least one active rule is updated, and the query step is executed again.
In one embodiment, the query module 604 is configured to: sorting the object data according to the matching degree of the current query condition to obtain current target object data; obtaining final target object data according with the expected target object quantity based on the current target object data, wherein the final target object data comprises the following steps: and taking the current target object data of the expected target object number as final target object data.
In an embodiment, the sorting the object data according to the matching degree to the current query condition includes: according to the rule content, counting the object data to obtain actual parameters of each object in the object data; calculating the matching degree of each object and the current query condition according to the actual parameters and the current query parameters; and sequencing the objects according to the matching degree.
In one embodiment, the query instruction further includes: a rule range of the activity rule; the query module 604 is configured to: and constructing a current query condition based on the rule content, the current rule parameter and the rule range.
In one embodiment, the query instruction further includes: an object appearance time range and/or an object appearance place range; the query module 604 is configured to: and processing the object data of the objects with the appearance time within the object appearance time range and/or the appearance place within the object appearance place range based on the current query condition.
In one embodiment, the object data is pre-statistical snapshot data; the apparatus further comprises a pre-statistics module 605 for pre-statistics of the snapshot data, comprising: grouping the snapshot data of the same object into a same object image set; the snapshot data correspond to the snapshot time and the snapshot place of the snapshot data; and counting the appearance positions of the objects in the object image set according to the snapshot positions.
For a detailed description of the target object data query device 600, please refer to the description of the related method steps in the above embodiments.
An embodiment of the present invention further provides a non-transitory electronic device readable storage medium, including: a program that, when run on an electronic device, causes the electronic device to perform all or part of the procedures of the methods in the above-described embodiments. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a Flash Memory (Flash Memory), a Hard Disk (Hard Disk Drive, abbreviated as HDD), a Solid State Drive (SSD), or the like. The storage medium may also comprise a combination of memories of the kind described above.
Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, it is to be understood that various changes and modifications may be effected therein by one skilled in the art without departing from the spirit and scope of the invention, and that such changes and modifications are to be within the scope of the appended claims.

Claims (11)

1. A method for querying target object data from object data to be queried is characterized by comprising the following steps:
receiving a query instruction for a target object, the query instruction comprising: the number of expected target objects and the rule content of the at least one activity rule;
acquiring initial rule parameters; the initial rule parameters are contained in the query instruction, or the initial rule parameters are given by a system running the method;
taking the initial rule parameter as a current rule parameter;
performing a query step, the query step comprising:
constructing a current query condition based on the rule content and the current rule parameter;
processing the object data to be queried based on the current query condition to obtain current target object data;
and obtaining final target object data according with the expected target object quantity based on the current target object data.
2. The method according to claim 1, wherein the processing the object data to be queried based on the current query condition to obtain current target object data comprises:
and taking the object data to be queried with the activity rule matched with the current query condition as the current target object data.
3. The method according to claim 1 or 2, wherein the deriving final target object data that corresponds to the expected number of target objects based on the current target object data comprises:
if the current target object quantity corresponding to the current target object data meets the expected target object quantity, taking the current target object data as the final target object data;
if the number of the current target objects does not accord with the expected number of the target objects, updating the current rule parameter of at least one activity rule, and re-executing the query step.
4. The method according to claim 1, wherein the processing the object data to be queried based on the current query condition to obtain current target object data comprises:
sorting the object data to be queried according to the matching degree of the current query condition to obtain current target object data;
obtaining final target object data according with the expected target object quantity based on the current target object data, wherein the final target object data comprises the following steps:
and taking the current target object data of the expected target object number as final target object data.
5. The method of claim 4, wherein said ranking said object data according to a degree of match to said current query condition comprises:
according to the rule content, counting the object data to obtain actual parameters of each object in the object data;
calculating the matching degree of each object and the current query condition according to the actual parameters and the current query parameters;
and sequencing the objects according to the matching degree.
6. The method of any of claims 1-5, wherein the query instruction further comprises: a rule range of the activity rule;
the constructing of the current query condition based on the rule content and the current rule parameter includes:
and constructing a current query condition based on the rule content, the current rule parameter and the rule range.
7. The method of any of claims 1-6, wherein the query instruction further comprises: an object appearance time range and/or an object appearance place range;
taking the object with the appearance time within the object appearance time range and/or the appearance place within the object appearance place range as the object to be inquired;
or taking the object data of which the shooting time is within the object appearance time range and/or the shooting place is within the object appearance place range as the object data to be inquired.
8. The method according to any one of claims 1-7, wherein the object data is pre-statistical snapshot data;
the step of pre-counting the snapshot data comprises the following steps:
grouping the snapshot data of the same object into a same object image set; the snapshot data correspond to the snapshot time and the snapshot place of the snapshot data;
according to the snapshot places, counting the time/times of the object in the object image set appearing in each snapshot place;
and taking the data obtained by statistics as object data.
9. A target object data query apparatus, comprising:
a receiving module, configured to receive a query instruction for a target object, where the query instruction includes: the number of expected target objects and the rule content of the at least one activity rule;
the acquisition module is used for acquiring initial rule parameters; the initial rule parameters are contained in the query instruction, or the initial rule parameters are given by a system operating the query method;
the determining module is used for taking the initial rule parameter as a current rule parameter;
a query module for querying:
constructing a current query condition based on the rule content and the current rule parameter;
processing object data based on the current query condition to obtain current target object data;
and obtaining final target object data according with the expected target object quantity based on the current target object data.
10. An electronic device, comprising:
a memory to store a computer program;
a processor configured to execute the method according to any one of claims 1 to 8 to query target object data that conforms to the query instruction.
11. A non-transitory electronic device readable storage medium, comprising: program which, when run by an electronic device, causes the electronic device to perform the method of any one of claims 1 to 8.
CN202011553754.1A 2020-12-24 2020-12-24 Target object data query method, device, equipment and storage medium Pending CN112632411A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011553754.1A CN112632411A (en) 2020-12-24 2020-12-24 Target object data query method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011553754.1A CN112632411A (en) 2020-12-24 2020-12-24 Target object data query method, device, equipment and storage medium

Publications (1)

Publication Number Publication Date
CN112632411A true CN112632411A (en) 2021-04-09

Family

ID=75324836

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011553754.1A Pending CN112632411A (en) 2020-12-24 2020-12-24 Target object data query method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN112632411A (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130091119A1 (en) * 2010-06-21 2013-04-11 Telefonaktiebolaget L M Ericsson (Publ) Method and Server for Handling Database Queries
CN109828991A (en) * 2018-12-03 2019-05-31 深圳市北斗智能科技有限公司 Sort method, device, equipment and storage medium are inquired under the conditions of a kind of multi-space
CN110334670A (en) * 2019-07-10 2019-10-15 北京迈格威科技有限公司 Object monitor method and device, electronic equipment, storage medium
CN110347698A (en) * 2019-07-16 2019-10-18 中国工商银行股份有限公司 Method for processing report data and device
CN110633296A (en) * 2018-05-31 2019-12-31 北京京东尚科信息技术有限公司 Data query method, device, medium and electronic equipment
CN110704491A (en) * 2019-09-30 2020-01-17 京东城市(北京)数字科技有限公司 Data query method and device
CN111814045A (en) * 2020-06-30 2020-10-23 平安科技（深圳）有限公司 Data query method and device, electronic equipment and storage medium

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130091119A1 (en) * 2010-06-21 2013-04-11 Telefonaktiebolaget L M Ericsson (Publ) Method and Server for Handling Database Queries
CN110633296A (en) * 2018-05-31 2019-12-31 北京京东尚科信息技术有限公司 Data query method, device, medium and electronic equipment
CN109828991A (en) * 2018-12-03 2019-05-31 深圳市北斗智能科技有限公司 Sort method, device, equipment and storage medium are inquired under the conditions of a kind of multi-space
CN110334670A (en) * 2019-07-10 2019-10-15 北京迈格威科技有限公司 Object monitor method and device, electronic equipment, storage medium
CN110347698A (en) * 2019-07-16 2019-10-18 中国工商银行股份有限公司 Method for processing report data and device
CN110704491A (en) * 2019-09-30 2020-01-17 京东城市(北京)数字科技有限公司 Data query method and device
CN111814045A (en) * 2020-06-30 2020-10-23 平安科技（深圳）有限公司 Data query method and device, electronic equipment and storage medium

Similar Documents

Publication Publication Date Title
EP1320041A2 (en) Searching profile information
CN111260129A (en) Multi-yard vehicle path planning method and device, computer equipment and storage medium
CN110555164B (en) Method, device, computer equipment and storage medium for generating group interest labels
CN109949154A (en) Customer information classification method, device, computer equipment and storage medium
CN104517052B (en) Invasion detection method and device
CN110807699A (en) Overdue event payment collection method and device and computer readable storage medium
CN112651321A (en) File processing method and device and server
WO2017095413A1 (en) Incremental automatic update of ranked neighbor lists based on k-th nearest neighbors
CN111475746A (en) Method and device for mining point of interest, computer equipment and storage medium
JP6570978B2 (en) Cluster selection device
CN112052251B (en) Target data updating method and related device, equipment and storage medium
CN113947800A (en) Face confidence method, system, equipment and medium based on space-time collision
WO2021212760A1 (en) Method and apparatus for determining identity type of person, and electronic system
CN112416590A (en) Server system resource adjusting method and device, computer equipment and storage medium
CN110162535B (en) Search method, apparatus, device and storage medium for performing personalization
CN112632411A (en) Target object data query method, device, equipment and storage medium
CN113920353B (en) Unsupervised face image secondary clustering method, unsupervised face image secondary clustering device and unsupervised face image secondary clustering medium
CN115292475A (en) Cloud computing service information processing method and system based on smart city
CN111382628B (en) Method and device for judging peer
CN117235297B (en) Image selection method and computer equipment
CN112532692A (en) Information pushing method and device and storage medium
CN112686312A (en) Data classification method, device and system
TWM580230U (en) Financial service application review system
CN112861034B (en) Method, device, equipment and storage medium for detecting information
CN110827924B (en) Clustering method and device for gene expression data, computer equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20210409

RJ01 Rejection of invention patent application after publication