WO2019148715A1 - Information processing method and apparatus, and computer device and storage medium - Google Patents

Abstract

An information processing method, comprising: acquiring multiple pieces of first position data of a current user within a first time period and multiple pieces of second position data of the current user within a second time period; acquiring consumption data corresponding to these pieces of first position data, and obtaining corresponding consumption features according to these pieces of consumption data; inputting the consumption features into a pre-trained credit model to obtain a target credit corresponding to the current user; comparing the second position data with pre-set position data to obtain an abnormal position comparison result; and obtaining a resource transfer policy corresponding to the current user according to the abnormal position comparison result and the target credit.

Description

信息处理方法、装置、计算机设备和存储介质Information processing method, device, computer device and storage medium

相关申请的交叉引用Cross-reference to related applications

本申请要求于2018年2月01日提交中国专利局，申请号为2018101023297，申请名称为“信息处理方法、装置、计算机设备和存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。This application claims the priority of the Chinese Patent Application entitled "Information Processing Method, Apparatus, Computer Equipment, and Storage Media" by the Chinese Patent Office, filed on February 1, 2018, the entire disclosure of which is incorporated by reference. In this application.

技术领域Technical field

本申请涉及一种信息处理方法、装置、计算机设备和存储介质。The present application relates to an information processing method, apparatus, computer device, and storage medium.

背景技术Background technique

随着互联网技术的发展，人们对互联网的使用越来越频繁，越来越多的用户通过网络进行资源数值转移。因此在很多场景下都存在确定资源转移策略的需求，例如在信贷领域中，往往需要根据用户信息例如年龄、收入等确定向用户发放贷款的额度，目前主要依靠用户自己填写的信息进行资源转移策略评估，然而用户提交的信息查证难度较大，不准确，导致得到资源数值转移策略的准确性低。With the development of Internet technology, people use the Internet more and more frequently, and more and more users use the network to transfer resource values. Therefore, in many scenarios, there is a need to determine a resource transfer strategy. For example, in the credit field, it is often necessary to determine the amount of loans to be issued to users according to user information such as age, income, etc. Currently, the information that the user fills in is mainly used for resource transfer strategies. Evaluation, however, the information submitted by the user is difficult to verify, which is inaccurate, resulting in low accuracy of the resource value transfer strategy.

发明内容Summary of the invention

根据本申请公开的各种实施例，提供一种信息处理方法、装置、计算机设备和存储介质。According to various embodiments disclosed herein, an information processing method, apparatus, computer device, and storage medium are provided.

一种信息处理方法包括：An information processing method includes:

获取当前用户在第一时间段的多个第一位置数据以及所述当前用户在第二时间段的多个第二位置数据；Obtaining a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period;

获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征；Obtaining consumption data corresponding to each of the first location data, and obtaining corresponding consumption features according to the respective consumption data;

将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户对应的目标信用度；Entering the consumption feature into a pre-trained credit model to obtain a target credit corresponding to the current user;

将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果；及Comparing the second location data with the preset location data to obtain an abnormal location comparison result; and

根据所述异常位置比对结果以及所述目标信用度得到所述当前用户对应的资源转移策略。And obtaining, according to the abnormal location comparison result and the target credit degree, a resource transfer policy corresponding to the current user.

一种信息处理装置包括：An information processing apparatus includes:

位置数据获取模块，用于获取当前用户在第一时间段的多个第一位置数据以及所述当前用户在第二时间段的多个第二位置数据；a location data obtaining module, configured to acquire a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period;

特征得到模块，用于获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征；a feature obtaining module, configured to acquire consumption data corresponding to each of the first location data, and obtain corresponding consumption features according to the respective consumption data;

信用度得到模块，用于将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户对应的目标信用度；a credit obtaining module, configured to input the consumption feature into a pre-trained credit model, to obtain a target credit corresponding to the current user;

对比模块，用于将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果；及a comparison module, configured to compare the second location data with the preset location data to obtain an abnormal location comparison result; and

策略得到模块，用于根据所述异常位置比对结果以及所述目标信用度得到所述当前用户对应的资源转移策略。And a policy obtaining module, configured to obtain, according to the abnormal location comparison result and the target credit, a resource transfer policy corresponding to the current user.

一种计算机设备，包括存储器和一个或多个处理器，存储器中存储有计算机可读指令，计算机可读指令被处理器执行时实现本申请任意一个实施例中提供的信息处理方法的步骤。A computer apparatus comprising a memory and one or more processors having stored therein computer readable instructions that, when executed by a processor, implement the steps of the information processing method provided in any one of the embodiments of the present application.

一个或多个存储有计算机可读指令的非易失性存储介质，计算机可读指令被一个或多个处理器执行时，使得一个或多个处理器实现本申请任意一个实施例中提供的信息处理方法的步骤。One or more non-volatile storage media storing computer readable instructions, when executed by one or more processors, cause one or more processors to implement the information provided in any one embodiment of the present application The steps of the processing method.

本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征和优点将从说明书、附图以及权利要求书变得明显。Details of one or more embodiments of the present application are set forth in the accompanying drawings and description below. Other features and advantages of the present invention will be apparent from the description, drawings and claims.

附图说明DRAWINGS

为了更清楚地说明本申请实施例中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其它的附图。In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings to be used in the embodiments will be briefly described below. Obviously, the drawings in the following description are only some embodiments of the present application, Those skilled in the art can also obtain other drawings based on these drawings without any creative work.

图1为根据一个或多个实施例中信息处理方法的应用场景图。FIG. 1 is an application scenario diagram of an information processing method according to one or more embodiments.

图2为根据一个或多个实施例中信息处理方法的流程示意图。2 is a flow diagram of an information processing method in accordance with one or more embodiments.

图3为根据一个或多个实施例中将消费特征输入预先训练得到的信用度模型中，得到当前用户对应的当前用户的信用度的步骤的流程示意图。FIG. 3 is a schematic flowchart of a step of obtaining a credit degree of a current user corresponding to a current user by inputting a consumption feature into a pre-trained credit model according to one or more embodiments.

图4为另一个实施例中信息处理方法的流程示意图。4 is a schematic flow chart of an information processing method in another embodiment.

图5为根据一个或多个实施例中获取各个第一位置数据对应的消费数据，根据各个消费数据得到对应的消费特征的步骤的流程示意图。FIG. 5 is a schematic flowchart of a step of obtaining consumption data corresponding to each first location data according to one or more embodiments, and obtaining corresponding consumption features according to respective consumption data.

图6为根据一个或多个实施例中根据异常位置比对结果以及目标信用度得到当前用户对应的资源转移策略的步骤的流程示意图。FIG. 6 is a schematic flowchart of steps of obtaining a resource transfer policy corresponding to a current user according to an abnormal position comparison result and a target credit degree according to one or more embodiments.

图7为根据一个或多个实施例中将第二位置数据与预设位置数据进行对比，得到异常位置比对结果的步骤的流程示意图。FIG. 7 is a flow diagram showing the steps of comparing the second location data with the preset location data to obtain an abnormal location comparison result according to one or more embodiments.

图8为根据一个或多个实施例中信息处理装置的框图。FIG. 8 is a block diagram of an information processing apparatus in accordance with one or more embodiments.

图9为根据一个或多个实施例中信用度得到模块的框图。9 is a block diagram of a credit get module in accordance with one or more embodiments.

图10为另一个实施例中信息处理装置的框图。Figure 10 is a block diagram of an information processing apparatus in another embodiment.

图11为根据一个或多个实施例中计算机设备的框图。11 is a block diagram of a computer device in accordance with one or more embodiments.

具体实施方式Detailed ways

为了使本申请的技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。In order to make the technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the application and are not intended to be limiting.

本申请提供的信息处理方法，可以应用于如图1所示的应用环境中。终端102通过网络与服务器104通过网络进行通信。终端102为当前用户登录的终端，例如，终端上安装有进行资源转移的应用，当前用户可以登录该应用，应用可以获取当前终端的位置信息作为当前用户的位置数据发送到服务器104中，终端102可以实时、随机或者在固定时间向服务器104发送位置数据，服务器104接收到终端102发送的位置数据后，对位置数据进行保存，当需要获取当前用户对应的资源转移策略时例如当接收到当前用户发送的贷款请求后，服务器104可以获取终端102在第一时间段的多个第一位置数据以及终端102在第二时间段的多个第二位置数据，以得到当前用户对应的资源转移策略。其中，终端102可以但不限于是各种个人计算机、笔记本电脑、智能手机、平板电脑和便携式可穿戴设备，服务器104可以用独立的服务器或者是多个服务器组成的服务器集群来实现。The information processing method provided by the present application can be applied to an application environment as shown in FIG. 1. Terminal 102 communicates with server 104 over a network over a network. The terminal 102 is a terminal that is logged in by the current user. For example, the application for resource transfer is installed on the terminal. The current user can log in to the application. The application can obtain the location information of the current terminal and send the location data of the current user to the server 104. The terminal 102 The location data may be sent to the server 104 in real time, randomly, or at a fixed time. After receiving the location data sent by the terminal 102, the server 104 saves the location data. When the resource transfer policy corresponding to the current user needs to be obtained, for example, when the current user is received. After the loan request is sent, the server 104 may obtain a plurality of first location data of the terminal 102 in the first time period and a plurality of second location data of the terminal 102 in the second time period to obtain a resource transfer policy corresponding to the current user. The terminal 102 can be, but is not limited to, various personal computers, notebook computers, smart phones, tablets, and portable wearable devices. The server 104 can be implemented by a separate server or a server cluster composed of multiple servers.

在一些实施例中，如图2所示，提供了一种信息处理方法，以该方法应用于图1中的服务器为例进行说明，包括以下步骤：In some embodiments, as shown in FIG. 2, an information processing method is provided, which is applied to the server in FIG. 1 as an example, and includes the following steps:

步骤S202，获取当前用户在第一时间段的多个第一位置数据以及当前用户在第二时间段的多个第二位置数据。Step S202: Acquire a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period.

具体地，第一位置数据以及第二位置数据的个数可以根据需要进行设置，第一位置数据以及第二位置数据的个数可以相同也可以不同，第一时间段和第二时间段可以相同也可以不同，位置数据可以是当前用户对应的当前终端发送到服务器中，也可以是服务器从其他存储当前用户的位置数据的设备中获取的，本申请实施例对服务器获取当前用户对应的位置数据的来源不做限制。例如，可以获取当前用户在过去一个月内的位置数据作为第一位置数据，获取当前用户在过去两周内的位置数据作为第二位置数据。或者可以获取当前用户在消费地点例如商场、娱乐地点以及餐饮等地点中的一种或多种位置数据作为第一位置数据，将用户在住宅以及办公地点中的一种或多种的位置数据作为第二位置数据。Specifically, the number of the first location data and the second location data may be set as needed, and the number of the first location data and the second location data may be the same or different, and the first time period and the second time period may be the same. Alternatively, the location data may be sent to the server by the current terminal corresponding to the current user, or may be obtained by the server from another device that stores the location data of the current user. The embodiment of the present application obtains the location data corresponding to the current user from the server. The source is not limited. For example, the location data of the current user in the past month may be acquired as the first location data, and the location data of the current user in the past two weeks is acquired as the second location data. Or, the one or more location data of the current user in a shopping location such as a shopping mall, an entertainment location, and a catering location may be obtained as the first location data, and the location data of the user in one or more of the residence and the office location is used as the location data. Second location data.

在一些实施例中，当前终端可以向服务器发送GPS(Global Positioning System，全球定位***)坐标信息以及当前用户的用户标识，因此，服务器接收到GPS信息后，将GPS信息作为当前用户对应的位置数据。在一些实施例中，当前终端可以向服务器发送当前用户的位置相关信息，例如可以是信号发射设备发送的信号强度信息以及信号发射设备的标识，因此服务器可以根据信号强度计算得到当前终端与信号发射设备的距离，然后再根据信号发射设备的位置以及当前终端与信号发射设备的距离得到当前用户的位置。信号发射设备例如可以为WiFi(Wireless Fidelity，无线保真)设备、蓝牙设备以及Zigbee设备等。In some embodiments, the current terminal may send GPS (Global Positioning System) coordinate information and a user identifier of the current user to the server. Therefore, after receiving the GPS information, the server uses the GPS information as the location data corresponding to the current user. . In some embodiments, the current terminal may send the location related information of the current user to the server, for example, may be signal strength information sent by the signal transmitting device and an identifier of the signal transmitting device, so the server may calculate the current terminal and signal transmission according to the signal strength. The distance of the device, and then the current user's location based on the location of the signal transmitting device and the distance between the current terminal and the signal transmitting device. The signal transmitting device may be, for example, a WiFi (Wireless Fidelity) device, a Bluetooth device, a Zigbee device, or the like.

在一些实施例中，若获取不到足够精确的位置，只能得到当前用户所在的位置区域时，则可以获取当前用户的多个位置区域信息，然后根据位置区域间的位置关系得到用户的精确位置。在一些实施例中，可以获取三个位置区域，然后计算三个位置区域的重合区域，根据重合区域得到对应的位置信息。例如，当3个位置区域有重合区域时，则可以获取重合区域分别与三个位置区域的边界之间的相交点，以3个相交点互相连线形成的三角形的内心坐标作为当前用户的位置。当3个位置区域两两相交时，则可以分别获取两两相交得到的重合区域与位置区域的边界的交点，将交点的连线的中点作为三角形的顶点，然后将三角形的内心 作为当前用户的位置。可以理解，当三个区域没有相交时则可以获取另外的3个位置区域。若连续三次获取的三个位置区域没有相交时，可以利用最后的三个区域的中心所形成的三角形的内心作为当前用户的位置。In some embodiments, if a location that is not accurate enough is obtained, and only the location area where the current user is located, the location information of the current user may be obtained, and then the accuracy of the user is obtained according to the location relationship between the location areas. position. In some embodiments, three location areas may be acquired, then the coincidence areas of the three location areas are calculated, and corresponding location information is obtained according to the coincidence area. For example, when there are overlapping regions in the three position regions, the intersection point between the coincident regions and the boundaries of the three position regions may be obtained, and the inner coordinates of the triangle formed by connecting the three intersecting points to each other are used as the current user's position. . When the three positional regions intersect at two or two, respectively, the intersections of the coincident regions obtained by the intersection of the two and the two and the boundary of the location region can be respectively obtained, the midpoint of the intersection of the intersection points is taken as the vertices of the triangle, and then the inner core of the triangle is used as the current user. s position. It can be understood that when the three regions do not intersect, another three location regions can be acquired. If the three position areas acquired three times in succession do not intersect, the inner center of the triangle formed by the center of the last three areas can be utilized as the position of the current user.

在一些实施例中，利用三个位置区域得到当前用户的位置数据的算法具体如下：获取这三个位置区域的中心点，假设为(X1，Y1)，(X2，Y2)，(X3，Y3),每个位置区域的半径为d1、d2以及d3，然后构造这3个位置区域对应的方程：(x-Xi)2+(y-Yi)2＝di，其中Xi指的是第i个位置区域的横坐标，Yi指的是第i个位置区域的纵坐标，di指第i个位置区域的半径，i可以取1、2、3。将这3个方程进行两两组合，形成三个方程组，求解三个方程组分别得到目标坐标点。若求解方程组得到的目标坐标点为小于等于2个，可以继续找下一组位置区域或者以这三个位置区域的中心形成的三角形的内心作为当前用户的位置。若目标坐标点大于等于3个，则可以将每3个目标坐标点形成的三角形对应的区域与3个位置区域进行比较，若存在均位于这三个位置区域的三角形，则该三角形的内心为当前用户的位置。若不存在均位于这三个位置区域的三角形，则获取每个方程组的两个目标坐标点的连线的中点，然后以这3个中点形成的三角形的内心作为当前用户对应的位置。In some embodiments, the algorithm for obtaining the location data of the current user by using three location areas is as follows: acquiring the center points of the three location areas, assuming (X1, Y1), (X2, Y2), (X3, Y3) ), the radius of each position area is d1, d2, and d3, and then the equation corresponding to the three position areas is constructed: (x-Xi)2+(y-Yi)2=di, where Xi refers to the ith The abscissa of the position area, Yi refers to the ordinate of the i-th position area, and di refers to the radius of the i-th position area, i can take 1, 2, and 3. The three equations are combined in pairs to form three equations, and the three equations are solved to obtain the target coordinate points. If the target coordinate points obtained by solving the equations are less than or equal to two, the next set of positional regions or the inner core of the triangle formed by the centers of the three positional regions may be used as the current user's position. If the target coordinate point is greater than or equal to three, the region corresponding to the triangle formed by each of the three target coordinate points can be compared with the three position regions. If there are triangles located in the three position regions, the inner center of the triangle is The location of the current user. If there are no triangles located in the three position regions, the midpoints of the lines connecting the two target coordinate points of each equation group are obtained, and then the inner center of the triangle formed by the three midpoints is taken as the position corresponding to the current user. .

在一些实施例中，三个位置区域可以是通过GPS信息获取的，也可以获取终端的IP地址，根据IP地址得到对应的基站位置，然后将基站所在的位置区域作为终端的位置区域。在一些实施例中，对于无法通过GPS进行精确定位得到终端所处精确位置的地点例如室内，可以通过室内例如商场上的Zigbee装置的位置以及当前终端接收的信号强度计算得到用户精确位置。In some embodiments, the three location areas may be obtained by using GPS information, or may acquire the IP address of the terminal, obtain the corresponding base station location according to the IP address, and then use the location area where the base station is located as the location area of the terminal. In some embodiments, for a location that is not accurately located by GPS to obtain a precise location of the terminal, such as a room, the user's precise location can be calculated by the location of the Zigbee device on the store, such as a store, and the signal strength received by the current terminal.

在一些实施例中，由于需要持续的获取用户位置，为了保护用户的隐私，当前终端可以通过位置混淆算法对得到的位置信息进行混淆，因此服务器接收到的位置数据为根据实际位置进行混淆后的位置数据，故服务器可以进行反混淆处理得到实际的位置数据。反混淆处理的方法具体可以根据位置混淆算法进行设置。例如当混淆算法为实际经度和纬度各加2度时，则反混淆算法为将接收到的经纬度各减2度得到实际位置。In some embodiments, since the user location needs to be continuously acquired, in order to protect the privacy of the user, the current terminal may confuse the obtained location information by using a location confusion algorithm, so the location data received by the server is confusing according to the actual location. Location data, so the server can perform anti-aliasing to get the actual location data. The method of anti-aliasing can be specifically set according to the location confusion algorithm. For example, when the confusion algorithm adds 2 degrees to the actual longitude and latitude, the anti-aliasing algorithm obtains the actual position by subtracting 2 degrees of the received latitude and longitude.

步骤S204，获取各个第一位置数据对应的消费数据，根据各个消费数据得到对应的消费特征。Step S204: Acquire consumption data corresponding to each first location data, and obtain corresponding consumption features according to each consumption data.

具体地，得到第一位置数据后，获取各个第一位置所在的地点的消费数据。所在的地点可以是商场、旅游、住宿以及娱乐等地点中的一种或多种，各个地点的消费数据可以是该地点的人均消费例如为800元，也可以是利用消费等级表示，例如消费数据可以是该地点的消费级别为高、中、低等。利用消费数据得到对应的消费特征的方法可以根据实际需要进行设置，在一些实施例中，可以根据每一个第一位置数据对应的消费数据得到一个消费特征，在一些实施例中,也可以根据多个消费数据得到一个消费特征。例如将每天在各个位置对应的消费数据的和得到一个消费特征。在一些实施例中还可以将各个第一位置所在的商户进行分类，例如分成生存类、发展类以及享受类。然后统计每一商户类型的消费水平如生成类商户的人均消费为60元，得到用户对应的消费特征。还可以统计当前用户进入各个商户类型以及相应消费水平的商户的次数，如平均每月进入生活类商户且消费水平为200～500元的商户的次数 为3次，得到对应一个的消费特征，平均每月进入生活类商户且消费水平为1000～1500元的商户的次数为1次，得到另一个消费特征。根据消费数据得到对应的消费特征的规则具体可以根据实际需要进行设置。例如，可以将得到的消费数据映射到特征向量空间，如对生活类商户且消费水平为200～500元的商户的次数为3次对应的消费特征可以为[10000000000]，进入生活类商户且消费水平为1000～1500元的商户的次数为1次的行为特征可以表示为[010000000000]，消费特征的维度具体可以根据实际需要进行设置。例如50维等。Specifically, after the first location data is obtained, the consumption data of the location where each first location is located is obtained. The location may be one or more of shopping malls, travel, accommodation, and entertainment. The consumption data of each location may be, for example, 800 yuan per person, or may be expressed by consumption level, such as consumption data. It can be that the consumption level of the place is high, medium and low. The method for obtaining the corresponding consumption feature by using the consumption data may be set according to actual needs. In some embodiments, a consumption feature may be obtained according to the consumption data corresponding to each first location data, and in some embodiments, according to more The consumption data gets a consumption characteristic. For example, a sum of consumer data corresponding to each location at each location is obtained for each day. In some embodiments, the merchants in which the respective first locations are located may also be classified, for example, into a survival class, a development class, and a enjoyment class. Then, the consumption level of each merchant type is counted as the per-capita consumption of the generated merchants is 60 yuan, and the consumption characteristics corresponding to the user are obtained. It is also possible to count the number of times the current user enters each merchant type and the corresponding consumption level. For example, the average number of merchants who enter the lifestyle category and consumes 200 to 500 yuan per month is 3 times, and the corresponding consumption characteristics are obtained. The number of merchants who enter the living category merchants every month and the consumption level is 1000-1500 yuan is 1 time, which gives another consumption characteristic. The rules for obtaining the corresponding consumption characteristics according to the consumption data may be specifically set according to actual needs. For example, the obtained consumption data may be mapped to the feature vector space. For example, the number of times of the merchants who have a consumption level of 200 to 500 yuan for the life class merchants is 3 times, and the consumption characteristic may be [10000000000], and enters the lifestyle merchants and consumes The behavioral feature of the number of merchants with a level of 1000 to 1500 yuan can be expressed as [010000000000], and the dimension of the consumption feature can be set according to actual needs. For example, 50 dimensions and so on.

步骤S206，将消费特征输入预先训练得到的信用度模型中，得到当前用户对应的当前用户的目标信用度。In step S206, the consumption feature is input into the credit model obtained in advance, and the target credit of the current user corresponding to the current user is obtained.

具体地，信用度用于衡量用户信用高低的程度，信用度可以用具体的数值表示例如信用分进行表示，也可以通过级别进行表示，例如高、中、低等。将消费特征输入到预先训练的信用度模型中，可以得到对应的目标信用度。信用度模型可以为一个或多个。当有多个时，可以根据多个信用度模型得到的信用度得到目标信用度，例如可以将多个信用度模型得到的信用度的平均值、中位数、最高信用度或者最低信用度作为目标信用度，也可以获取各个信用度模型对应的权重，根据信用度模型的权重以及各个信用度模型得到的信用度得到目标信用度。信用度模型的权重可以是根据经验、需要设置的，也可以通过模型训练的效果得到。例如，当有三个信用度模型时，各个信用度模型对应的权重可以分别设置为0.5、0.3、0.2，若各个信用度模型对应的信用分数分别为600、500、800，则目标信用度＝0.5*600+0.3*500+0.2*800＝610分。Specifically, the credit is used to measure the degree of credit of the user, and the credit can be expressed by a specific numerical value such as a credit score, or can be expressed by a level, such as high, medium, and low. By inputting the consumption feature into the pre-trained credit model, the corresponding target credit can be obtained. The credit model can be one or more. When there are multiple, the target credit can be obtained according to the credit obtained by multiple credit models. For example, the average, median, highest credit or minimum credit of the credit obtained by multiple credit models can be used as the target credit, and each credit can also be obtained. The weight corresponding to the credit model is obtained according to the weight of the credit model and the credit obtained by each credit model. The weight of the credit model can be set according to experience, needs, or can be obtained through the effect of model training. For example, when there are three credit models, the weights corresponding to each credit model can be set to 0.5, 0.3, and 0.2 respectively. If the credit scores of each credit model are 600, 500, and 800, respectively, the target credit = 0.5 * 600 + 0.3. *500+0.2*800=610 points.

信用度模型是预先根据训练数据进行模型训练得到的。通过训练数据进行模型训练，能够确定每个消费特征对应的模型参数，从而根据训练得到的模型参数得到信用度模型。在进行模型训练时，可以采用有监督的模型训练方式，例如逻辑回归模型、贝叶斯模型、自适应算法以及SVM(Support Vector Machine,支持向量机)等等。例如，可以获取已知信用度以及对应的消费特征，然后将消费特征以及已知的信用度作为训练数据进行模型训练。以SVM为例，在训练过程中可以采用随机梯度下降算法进行模型训练，在梯度下降过程中需要使得代价函数J(θ)最小对应的模型参数，从而得到信用度模型。The credit model is obtained by training the model based on the training data in advance. Through the training of the training data, the model parameters corresponding to each consumption feature can be determined, and the credit model is obtained according to the model parameters obtained by the training. When performing model training, supervised model training methods such as logistic regression model, Bayesian model, adaptive algorithm, and SVM (Support Vector Machine) can be used. For example, the known credits and corresponding consumption characteristics can be obtained, and then the consumption characteristics and the known credits are modeled as training data. Taking SVM as an example, the stochastic gradient descent algorithm can be used to train the model during the training process. In the process of gradient descent, the cost function J(θ) needs to be minimized to obtain the credit model.

步骤S208，将第二位置数据与预设位置数据进行对比，得到异常位置比对结果。Step S208, comparing the second location data with the preset location data to obtain an abnormal location comparison result.

具体地，异常比对结果包括存在异常或者不存在异常。预设位置数据可以是当前用户的家庭住址、办公位置等，预设位置数据可以是当前用户填写的，也可以是根据当前用户日常的位置信息例如可以获取用户的历史轨迹，根据历史轨迹的规律得到用户的家庭地址、办公地址。将第二位置数据与预设位置数据进行对比，得到是否存在异常的判断标准可以根据需要进行设置，例如根据第二位置数据得到当前用户的行动轨迹例如在第二时间段内每一天的行动轨迹，然后将行动轨迹与用户填写的家庭地址、办公地址或者过去的日常行动轨迹进行比较，判断是是否不同即是否存在异常的行动轨迹。Specifically, the abnormal alignment result includes the presence or absence of an abnormality. The preset location data may be the current user's home address, office location, etc. The preset location data may be filled in by the current user, or may be obtained according to the current user's daily location information, for example, the user's historical trajectory may be acquired, according to the history trajectory. Get the user's home address and office address. Comparing the second location data with the preset location data to determine whether there is an abnormality determination criterion may be set as needed, for example, obtaining a current user's motion trajectory according to the second location data, for example, a motion trajectory of each day in the second time period. Then, the action track is compared with the home address, the office address, or the past daily action track filled out by the user, and it is judged whether it is different whether there is an abnormal action track.

步骤S210，根据异常位置比对结果以及目标信用度得到当前用户对应的资源转移策略。Step S210: Obtain a resource transfer policy corresponding to the current user according to the abnormal position comparison result and the target credit degree.

具体地，资源可以从一个账户转到另一个账户例如可以是贷款。资源转移策略包括资源转移数值，在一些实施例中，还可以包括资源转移条件，资源转移条件例如可以包括担保条 件，进行贷款的利息等等。需要综合考虑异常位置比对结果以及当前用户的目标信用度得到资源转移策略。在一些实施例中，可以根据目标信用度得到当前用户的资源转移数值，根据异常位置比对结果得到当前用户进行资源转移的条件例如贷款的利息或者需要提供的担保条件等，若异常位置的数量多或者存在异常时，则对应的条件要求高。在一些实施例中，可以根据目标信用度得到当前用户的资源转移数值，根据异常位置比对结果确定是否满足当前用户提出的临时提高资源转移数值的请求。Specifically, resources can be transferred from one account to another, such as a loan. The resource transfer policy includes resource transfer values. In some embodiments, resource transfer conditions may also be included. The resource transfer conditions may include, for example, guarantee conditions, interest on loans, and the like. It is necessary to comprehensively consider the abnormal position comparison result and the current user's target credit degree to obtain a resource transfer strategy. In some embodiments, the current user's resource transfer value may be obtained according to the target credit degree, and the current user's resource transfer condition, such as the loan interest or the guarantee condition to be provided, may be obtained according to the abnormal position comparison result, and the number of abnormal positions is large. Or when there is an abnormality, the corresponding condition is required to be high. In some embodiments, the resource transfer value of the current user may be obtained according to the target credit degree, and whether the request for temporarily raising the resource transfer value proposed by the current user is satisfied according to the abnormal location comparison result.

在一些实施例中，在时间轴上，第一时间段对应的起始时间早于第二时间段对应的起始时间，在一些实施例中，第一时间段的时间长度还可以大于第二时间段的时间长度，例如第二时间段可以为近期的数据例如过去一个月或者两个月的数据，第一时间段的可以为过去一年的数据。因此，利用长期的消费特征确定用户的信用度得到用户的资源转移数值的准确性高，而短期的第二时间段内是否存在异常位置比对结果则可以用于说明当前用户近期是否存在异常，故可以利用异常位置比对结果对资源转移策略进行调整，以提高资源转移策略的准确性。例如，若通过异常位置比对结果确定用户连续四周没有到达办公区域，则可以初步判断当前用户失业，当资源为贷款时，可以降低根据目标信用度得到的贷款额度，以降低风险，或者提高担保条件。In some embodiments, on the time axis, the start time corresponding to the first time period is earlier than the start time corresponding to the second time period. In some embodiments, the time length of the first time period may also be greater than the second time. The length of time of the time period, for example, the second time period may be recent data such as data of the past month or two months, and the data of the first time period may be data of the past year. Therefore, using the long-term consumption feature to determine the user's credit degree, the accuracy of the user's resource transfer value is high, and whether the abnormal position comparison result exists in the short-term second time period can be used to indicate whether the current user has an abnormality in the near future. The resource transfer strategy can be adjusted by using the abnormal position comparison result to improve the accuracy of the resource transfer strategy. For example, if it is determined by the abnormal position comparison result that the user has not reached the office area for four consecutive weeks, the current user may be initially determined to be unemployed. When the resource is a loan, the loan amount obtained according to the target credit degree may be reduced to reduce the risk or improve the guarantee condition. .

在一些实施例中，还可以将得到的资源转移策略推送到当前用户的当前终端上。可以是在接收到当前终端发送的资源转移请求例如贷款请求时发送的，也可以是服务器主动推送的，在此不做限制。In some embodiments, the obtained resource transfer policy can also be pushed to the current terminal of the current user. It may be sent when receiving a resource transfer request, such as a loan request, sent by the current terminal, or may be actively pushed by the server, and is not limited herein.

上述信息处理方法中，通过获取当前用户在第一时间段的多个第一位置数据以及当前用户在第二时间段的多个第二位置数据，获取各个第一位置数据对应的消费数据，根据各个消费数据得到对应的消费特征，将消费特征输入预先训练得到的模型中，得到当前用户的目标信用度，将第二位置数据与预设位置数据进行对比，得到异常位置比对结果，根据异常位置比对结果以及目标信用度得到当前用户对应的资源转移策略。由于可以利用用户位置得到用户的信用度，且进一步结合用户位置得到的异常位置比对结果得到的资源转移策略，因此得到资源转移策略的准确性高，也可以减少无效的资源转移策略的推送，提高计算机网络资源的利用率。In the above information processing method, by acquiring a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period, the consumption data corresponding to each first location data is obtained, according to Each consumption data obtains a corresponding consumption characteristic, and the consumption feature is input into a pre-trained model to obtain a target credit degree of the current user, and the second position data is compared with the preset position data to obtain an abnormal position comparison result according to the abnormal position. The comparison result and the target credit degree are obtained by the resource transfer strategy corresponding to the current user. Since the user's location can be used to obtain the user's credit, and the resource transfer strategy obtained by the abnormal location comparison result obtained by the user location is further combined, the accuracy of the resource transfer strategy is high, and the push of the invalid resource transfer strategy can be reduced and improved. Utilization of computer network resources.

在一些实施例中，如图3所示，步骤S206即将消费特征输入预先训练得到的信用度模型中，得到当前用户对应的当前用户的信用度的步骤包括：In some embodiments, as shown in FIG. 3, in step S206, the step of inputting the consumption feature into the pre-trained credit model, and obtaining the credit degree of the current user corresponding to the current user includes:

步骤S302，将消费特征分别输入到多个预先训练的信用度模型中，得到各个信用度模型输出的当前用户的初始信用度。Step S302, the consumption features are respectively input into a plurality of pre-trained credit models, and the initial credits of the current users output by the respective credit models are obtained.

具体地，信用度模型的数量可以根据实际需要进行设置，例如可以为3个。得到消费特征后，将消费特征分别输入到预先训练得到的信用度模型中，得到各个信用度模型输出的当前用户的初始信用度。Specifically, the number of credit models may be set according to actual needs, for example, may be three. After the consumption feature is obtained, the consumption features are respectively input into the pre-trained credit model to obtain the initial credit of the current user output by each credit model.

步骤S304，根据初始信用度以及对应的信用度模型的权重计算得到当前用户对应的目标信用度。Step S304, calculating, according to the initial credit degree and the weight of the corresponding credit model, the target credit degree corresponding to the current user.

具体地，各个信用度模型对应的权重可以是根据需要自定义设置的，也可以根据进行模 型训练时模型的准确率确定各个信用度模型对应的权重。得到初始信用度后，根据各个初始信用度以及对应的信用度模型对应的权重得到当前用户对应的目标信用度。例如，当有四个信用度模型时，各个信用度模型对应的权重分别为0.4、0.3、0.2、0.1，各个信用度模型输出的初始信用等级分别为高、高、低、中，由于信用等级为高的模型的权重为0.7，大于其他等级的权重，则目标信用度为高。Specifically, the weights corresponding to the respective credit models may be customized according to requirements, and the weights corresponding to the respective credit models may be determined according to the accuracy of the model when performing the model training. After the initial credit is obtained, the target credit corresponding to the current user is obtained according to each initial credit and the weight corresponding to the corresponding credit model. For example, when there are four credit models, the weights corresponding to each credit model are 0.4, 0.3, 0.2, and 0.1, respectively. The initial credit ratings output by each credit model are high, high, low, and medium, respectively, because the credit rating is high. The weight of the model is 0.7, which is greater than the weight of other levels, and the target credit is high.

如图4所示，在一些实施例中，信息处理方法还包括:As shown in FIG. 4, in some embodiments, the information processing method further includes:

步骤S402，获取进行模型训练的样本集，样本集包括多个样本，样本包括多个训练消费特征以及对应的样本信用度。Step S402: Acquire a sample set for performing model training. The sample set includes a plurality of samples, and the sample includes a plurality of training consumption features and corresponding sample credits.

具体地，样本集中的样本个数可以根据需要设置或者随机选取，例如可以为10万个，每个样本中的训练消费特征可以是根据多个训练用户的位置的消费数据得到的，例如根据多个训练用户所在的位置的商户的消费数据得到训练用户的消费特征。样本信用度可以是人工标注，也可以通过其他渠道例如银行根据用户的征信记录得到的。样本用于对模型进行训练，以训练得到信用度模型。Specifically, the number of samples in the sample set may be set as needed or randomly selected, for example, may be 100,000, and the training consumption feature in each sample may be obtained according to consumption data of positions of multiple training users, for example, according to multiple The consumption data of the merchants at the location where the training user is located is obtained by the training user. The sample credit can be manually labeled or obtained through other channels such as banks based on the user's credit history. The sample is used to train the model to train the credit model.

步骤S404，根据样本集以及多种不同的模型训练方法进行模型训练，得到各个不同的模型训练方法训练得到的多个信用度模型。Step S404, performing model training according to the sample set and a plurality of different model training methods, and obtaining a plurality of credit models trained by the different model training methods.

具体地，不同的模型训练方法可以是指采用的模型不同或者训练的过程不同等。例如，分别采用SVM、神经网络为不同的模型训练方法。采用SVM时，如果采用的核函数不同，也是不同的模型训练方法。得到样本集后，利用样本集以及多种不同的模型训练方法进行模型训练得到多种模型。在进行模型训练的过程中，由于样本信用度是已知的且为有监督的模型，因此可以通过不断调整模型参数使得根据输入的样本的消费特征以及模型参数得到的信用度是符合实际或者接近已知的样本信用度的，从而可以根据得到的模型参数得到信用度模型。模型训练的模型可以为SVM(Support Vector Machine，支持向量机)分类器模型，神经网络(Artificial Neural Network，ANN)分类器模型，逻辑回归算法(logistic Regression，LR)分类器模型和隐马尔可夫模型(Hidden Markov Model，HMM)等各种可以进行机器学习的模型。Specifically, different model training methods may refer to different models used or different training processes. For example, SVM and neural networks are respectively used as different model training methods. When using SVM, if the kernel function is different, it is also a different model training method. After obtaining the sample set, the model is trained using the sample set and a variety of different model training methods to obtain a variety of models. In the process of model training, since the sample credit is known and is a supervised model, the credits obtained according to the consumption characteristics of the input samples and the model parameters can be adjusted according to the actual or near known by continuously adjusting the model parameters. The sample credit is so that the credit model can be derived from the obtained model parameters. Model training models can be SVM (Support Vector Machine) classifier model, Neural Network (ANN) classifier model, logistic regression (LR) classifier model and hidden Markov Models (Hidden Markov Model, HMM) and other models that can be machine-learned.

步骤S406，将样本对应的消费特征输入到信用度模型中，得到样本对应的模型信用度。Step S406, input the consumption feature corresponding to the sample into the credit model, and obtain the model credit corresponding to the sample.

具体地，当训练得到各个信用度模型时，分别将样本集中样本的消费特征输入到训练好的信用度模型中，得到各个样本输入训练好的信用度模型后输出的模型信用度。Specifically, when each credit model is obtained by training, the consumption characteristics of the sample set samples are respectively input into the trained credit model, and the model credits output by each sample after inputting the trained credit model are obtained.

步骤S408，根据各个模型中样本对应的模型信用度与样本信用度的差距得到各个信用度模型对应的权重。Step S408, obtaining weights corresponding to the respective credit models according to the difference between the model credits and the sample credits corresponding to the samples in the respective models.

具体地，得到各个模型中样本的模型信用度后，计算模型信用度与样本信用度的差距，以根据差距得到各个信用度模型对应的权重。例如，假设a样本的模型信用度为80分，a样本进行模型训练时的样本信用度为90分，则模型信用度与样本信用度的差距为10分。又例如，a样本在B信用度模型输出的信用等级为高的概率为0.8，而样本中的样本信用度为高，即概率为1。则a样本在b模型的偏差值为0.2。根据模型中样本对应的模型信用度与样本信用度的差距得到各个信用度模型对应的权重具体可以根据实际需要进行设置。在一些实施例中，可以计算得到每个信用度模型中各个样本的模型信用度与样本信用度的差距的和，然后 根据差距的和得到信用度模型对应的权重。差距的和与权重呈负相关关系。例如，信用度模型的权重与为该信用度模型对应的差距的和的倒数。在一些实施例中，得到每个信用度模型对应的差距的和后，可以对各个模型对应的差距的和进行归一化，将归一化后的值作为对应的权重。Specifically, after obtaining the model credits of the samples in each model, the difference between the model credits and the sample credits is calculated, so as to obtain the weights corresponding to the respective credit models according to the gaps. For example, if the model credit of a sample is 80 points and the sample credit of a sample training model is 90 points, the difference between model credit and sample credit is 10 points. For another example, the probability that the a sample has a high credit rating in the B credit model is 0.8, and the sample credit in the sample is high, that is, the probability is 1. Then the deviation of the a sample in the b model is 0.2. According to the difference between the model credit degree and the sample credit degree corresponding to the sample in the model, the weight corresponding to each credit model can be set according to actual needs. In some embodiments, the sum of the differences between the model credits and the sample credits for each sample in each credit model can be calculated, and then the weights corresponding to the credit models are obtained based on the sum of the gaps. The sum of the gaps is inversely related to the weight. For example, the weight of the credit model is the reciprocal of the sum of the gaps corresponding to the credit model. In some embodiments, after obtaining the sum of the gaps corresponding to each credit model, the sum of the gaps corresponding to the respective models may be normalized, and the normalized values are taken as corresponding weights.

在一些实施例中，根据各个模型中样本对应的模型信用度与样本信用度的差距得到各个模型对应的权重的步骤可以包括：计算各个信用度模型中样本对应的模型信用度与样本信用度的偏差，对各个信用度模型对应的偏差进行求和计算，得到各个信用度模型对应的总偏差，根据各个信用度模型对应的总偏差以及预设的权重算法得到各个信用度模型对应的权重，其中，权重算法中总偏差与权重为负相关关系。In some embodiments, the step of obtaining the weight corresponding to each model according to the difference between the model credit and the sample credit corresponding to the sample in each model may include: calculating a deviation between the model credit and the sample credit corresponding to the sample in each credit model, for each credit The deviation corresponding to the model is summed and calculated, and the total deviation corresponding to each credit model is obtained. The weights corresponding to each credit model are obtained according to the total deviation corresponding to each credit model and the preset weight algorithm. The total deviation and weight in the weight algorithm are Negative correlation.

具体地，负相关关系指若总偏差大，则权重小，若总偏差小，则权重大。例如，若第一模型的总偏差为90，第二模型的总偏差为100。则根据权重算法得到的第一模型的权重比第二模型的权重大。权重算法可以根据实际需要进行设置。例如可以为一次函数，也可以为指数函数。Specifically, the negative correlation means that if the total deviation is large, the weight is small, and if the total deviation is small, the weight is significant. For example, if the total deviation of the first model is 90, the total deviation of the second model is 100. Then, the weight of the first model obtained according to the weighting algorithm is greater than the weight of the second model. The weight algorithm can be set according to actual needs. For example, it can be a one-time function or an exponential function.

例如，假设样本集有三个样本：A样本、B样本、C样本。且根据三个样本训练得到两个信用度模型：第一模型以及第二模型，将A样本、B样本、C样本分别输入到预先训练的第一模型以及第二模型中，得到A样本、B样本、C样本在第一模型输出的模型信用度a1、b1以及c1，A样本、B样本、C样本在第二模型输出的模型信用度分别为a2、b2以及c2。得到模型输出的模型信用度后，计算A样本信用度与a1、A样本信用度与a2、B样本信用度与b1、B样本信用度与b2、C样本信用度与c1、C样本信用度与c2的偏差，假设为a11、a21、b11、b12、c11以及c12。则第一模型对应的模型偏差值进行求和计算得到第一模型的总偏差为a11+b11+c11，第二模型对应的总偏差值a21+b21+c21。然后将总偏差值的倒数进行归一化后作为模型对应的权重。For example, suppose the sample set has three samples: A sample, B sample, and C sample. And according to the three sample training, two credit models are obtained: the first model and the second model, and the A sample, the B sample, and the C sample are respectively input into the pre-trained first model and the second model to obtain the A sample and the B sample. The model credits a1, b1, and c1 of the C sample output in the first model, and the model credits of the A sample, the B sample, and the C sample in the second model are a2, b2, and c2, respectively. After obtaining the model credit of the model output, calculate the deviation between the A sample credit and the a1, A sample credit and a2, B sample credit and b1, B sample credit and b2, C sample credit and c1, C sample credit and c2, assuming a11 , a21, b11, b12, c11, and c12. Then, the model deviation value corresponding to the first model is summed to obtain a total deviation of the first model as a11+b11+c11, and a total deviation value corresponding to the second model a21+b21+c21. Then, the reciprocal of the total deviation value is normalized and used as the weight corresponding to the model.

如图5所示，在一些实施例中，获取各个第一位置数据对应的消费数据，根据各个消费数据得到对应的消费特征的步骤包括：As shown in FIG. 5, in some embodiments, the acquiring the consumption data corresponding to each of the first location data, and obtaining the corresponding consumption feature according to the respective consumption data includes:

步骤S502，获取各个第一位置数据对应的商户信息，商户信息包括商户的消费数据以及商户的商户类型。Step S502: Obtain the merchant information corresponding to each first location data, where the merchant information includes the merchant's consumption data and the merchant's merchant type.

具体地，得到第一位置数据后，获取第一位置上的商户作为第一位置数据对应的商户信息，商户可以是餐饮商户、娱乐商户或者酒店商户等。消费数据可以是商户的人均消费例如为800元，也可以是利用该商户的消费等级表示，例如消费数据可以是该商户的消费级别为高、中、低。商户类型可以根据需要进行分类得到。例如可以是生活类、发展类以及享受类等。生活类可以是超市，发展类例如可以是各类学习培训学校、享受类可以是旅游景点、娱乐场景例如KTV等。Specifically, after the first location data is obtained, the merchant in the first location is obtained as the merchant information corresponding to the first location data, and the merchant may be a restaurant merchant, an entertainment merchant, or a hotel merchant. The consumption data may be that the per capita consumption of the merchant is, for example, 800 yuan, or may be expressed by the consumption level of the merchant. For example, the consumption data may be that the merchant's consumption level is high, medium, or low. Merchant types can be categorized as needed. For example, it can be a life class, a development class, and a enjoy class. The living class may be a supermarket, and the development class may be, for example, various types of learning training schools, and enjoyment classes may be tourist attractions, entertainment scenes such as KTV, and the like.

步骤S504，根据各个商户的消费数据以及商户的消费类型得到对应的消费特征。Step S504, obtaining corresponding consumption characteristics according to the consumption data of each merchant and the consumption type of the merchant.

具体地，消费特征的个数可以根据需要设置，例如可以是50个。得到消费数据以及商户的消费类型后，可以统计每一个消费类型的商户对应的消费数据，根据消费类型对应的消费数据得到消费特征，例如统计每个消费类型的商户的平均消费水平或者在各个消费水平的 商户的比例，得到对应的消费特征，例如，假设生活类消费类型的商户有三个，消费水平分别为800、900以及1000元，则可以计算得到生活类消费类型的商户的平均消费水平为(800+900+1000)/3＝800元，假设发展类消费类型的商户有5个，消费水平分别为500、900、1500、5500以及8000元，则可以计算到发展类消费类型的商户在消费水平为0～1000元的比例为2/5*100％＝40％，消费水平为1001元～3000元的比例为1/5*100％＝20％，消费水平为3001元以上的比例为2/5*100％＝40％。在一些实施例中，还可以统计当前用户进入各个商户类型以及相应消费水平的商户的次数，如平均每月进入生活类商户且消费水平为200～500元的商户的次数为3次，得到对应的消费特征。Specifically, the number of consumption features can be set as needed, for example, 50. After obtaining the consumption data and the consumption type of the merchant, the consumption data corresponding to the merchants of each consumption type can be counted, and the consumption characteristics are obtained according to the consumption data corresponding to the consumption type, for example, the average consumption level of the merchants of each consumption type is calculated or consumed at each consumption. The proportion of horizontal merchants is corresponding to the consumption characteristics. For example, if there are three merchants in the consumption category of life, the consumption levels are 800, 900 and 1000 yuan respectively, then the average consumption level of the merchants that can calculate the consumption type of life is (800+900+1000)/3=800 yuan. Assuming that there are 5 merchants in the development type of consumption, the consumption levels are 500, 900, 1500, 5500 and 8000 yuan respectively, then the merchants who can calculate the development type of consumption are The consumption level is 0 to 1000 yuan, the ratio is 2/5*100%=40%, the consumption level is 1001 yuan to 3,000 yuan, the ratio is 1/5*100%=20%, and the consumption level is more than 3001 yuan. 2/5*100%=40%. In some embodiments, the number of times that the current user enters each merchant type and the corresponding consumption level of the merchant may also be counted, for example, the number of merchants who enter the life-like merchants monthly and the consumption level is 200-500 yuan is 3 times, and the corresponding number is obtained. Consumption characteristics.

在一些实施例中，如图6所示，步骤S210即根据异常位置比对结果以及目标信用度得到当前用户对应的资源转移策略的步骤包括：In some embodiments, as shown in FIG. 6, the step S210 is to obtain a resource transfer policy corresponding to the current user according to the abnormal position comparison result and the target credit degree, including:

步骤S602，根据目标信用度得到当前用户对应的初始资源转移数值。Step S602, obtaining an initial resource transfer value corresponding to the current user according to the target credit degree.

具体地，设置了信用度与资源转移数值的对应关系，如信用度为90分对应的资源转移数值为9000元。也可以结合其他因素例如用户的收入、目前的债务情况等确定资安院转移数值，其中，收入与资源转移数值可以为正相关关系，债务数额与资源转移数值可以为负相关关系。Specifically, the correspondence between the credit and the resource transfer value is set, for example, the resource transfer value corresponding to the credit score of 90 is 9000 yuan. The transfer value of the Qiaoyuan may also be determined in combination with other factors such as the user's income and the current debt situation. The income and resource transfer values may be positively correlated, and the debt amount and the resource transfer value may be negatively correlated.

步骤S604，根据异常位置比对结果确定当前用户的行为模式。Step S604, determining a behavior mode of the current user according to the abnormal position comparison result.

具体地，行为模式可以包括正常、失业、分居、离婚以及超额消费中的一种或多种等。根据异常位置比对结果确定当前用户的行为模式的方法可以根据需要设置，例如，当当前用户的异常位置比对结果为每天起点相同，终点不同时，则确定用户失业。或者当异常位置比对结果为位置所在的商户的消费水平高于用户的承受能力即为超额消费。或者当用户每天早上出发的起点位置不同，但是终点位置相同，则行为模式可以为分居。Specifically, the behavior pattern may include one or more of normal, unemployment, separation, divorce, and excess consumption. The method for determining the current user's behavior pattern according to the abnormal position comparison result may be set as needed. For example, when the current user's abnormal position comparison result is the same as the daily starting point and the end point is different, the user is determined to be unemployed. Or when the abnormal position comparison result is that the consumption level of the merchant where the location is higher than the user's affordability is excessive consumption. Or when the user starts at a different starting point every morning, but the end position is the same, the behavior mode can be separated.

步骤S606，根据行为模式以及初始资源转移数值得到当前用户对应的目标资源转移数值。Step S606, obtaining a target resource transfer value corresponding to the current user according to the behavior mode and the initial resource transfer value.

具体地，可以设置每个行为模式对应的影响数值，然后将初始资源转移数值与影响数值进行相加，得到目标资源转移数值，例如，失业对应的影响数值为-9000，分居对应的影响数值为-6000，离婚对应的影响数值为-10000。正常行为模式对应的影响数值为0。也可以设置每种行为模式对应的异常分数，如失业80分，分居50分，超额消费40分，得到行为模式后，对异常分数进行统计，得到异常总分，然后根据异常总分得到对应的总影响数值。根据总影响数值以及初始资源转移数值得到当前用户对应的目标资源转移数值。Specifically, the impact value corresponding to each behavior mode may be set, and then the initial resource transfer value is added to the impact value to obtain the target resource transfer value. For example, the impact value corresponding to the unemployment is -9000, and the impact value corresponding to the separation is -6000, the impact value of divorce corresponds to -10000. The normal behavior mode corresponds to an impact value of zero. You can also set the abnormal score corresponding to each behavior mode, such as 80 points for unemployment, 50 points for separation, and 40 points for excess consumption. After getting the behavior pattern, the abnormal scores are counted, and the total score is obtained. Then, according to the abnormal total score, the corresponding score is obtained. Total impact value. The target resource transfer value corresponding to the current user is obtained according to the total impact value and the initial resource transfer value.

在一些实施例中，预设位置数据包括当前用户在第三时间段的历史行动轨迹，第二位置数据包括当前用户在第二时间段的目标行动轨迹。如图7所示，步骤S208即将第二位置数据与预设位置数据进行对比，得到异常位置比对结果的步骤包括：In some embodiments, the preset location data includes a historical action track of the current user for a third time period, and the second location data includes a target action track of the current user for the second time period. As shown in FIG. 7, the step S208 compares the second location data with the preset location data, and the step of obtaining the abnormal location comparison result includes:

步骤S702，获取当前用户在第三时间段的历史行动轨迹，第三时间段所在的时间早于第二时间段所在的时间。Step S702: Acquire a historical action track of the current user in a third time period, where the time of the third time period is earlier than the time of the second time period.

具体地，第三时间段早于第二时间段所在的时间是指第三时间段所包括的时间在总体上比第二时间段早。第三时间段在一些实施例中可以包括全部或者部分第二时间段，第三时间段与第二时间段也可以是相接的时间段，例如，第三时间段是2017年2月到八月，第二时间 段是2017年9月至今。例如，第三时间段可以是过去一年内的数据，第二时间段可以是过去一个月内的数据。行动轨迹指经过的路线，第三时间段的历史行动轨迹可以有多种，例如可以是第三时间段中每天的行动轨迹、每周的行动轨迹、工作日的行动轨迹或者节假日的行动轨迹等周期轨迹之中的一个或者多个。Specifically, the time that the third time period is earlier than the second time period means that the time included in the third time period is generally earlier than the second time period. The third time period may include all or part of the second time period in some embodiments, and the third time period and the second time period may also be the connected time period, for example, the third time period is February to August 2017. Month, the second period is September 2017 to date. For example, the third time period may be data in the past year, and the second time period may be data in the past month. The action track refers to the route that passes. The historical action track of the third time period can be various, for example, it can be the daily action track in the third time period, the weekly action track, the action track of the work day, or the action track of the holiday. One or more of the periodic trajectories.

步骤S704，将当前用户在第二时间段的目标行动轨迹与当前用户在第三时间段的历史行动轨迹进行对比，得到异常位置比对结果。Step S704, comparing the target action track of the current user in the second time period with the historical action track of the current user in the third time period, to obtain an abnormal position comparison result.

具体地，得到历史行动轨迹后，将目标行动轨迹与历史行动轨迹进行对比，以确认轨迹是否异常。在进行轨迹对比时，若第三时间段的历史行动轨迹的类型有多种，则可以对每种目标行动轨迹以及历史行动轨迹进行对比，得到每一种类型的轨迹的比对结果，确认当前用户的轨迹是否存在异常。例如，将第二时间段内的节假日行动轨迹与第三时间段内的节假日行动轨迹进行对比。将第二时间段内的工作日行动轨迹与第三时间段内的工作日行动轨迹进行对比。判断轨迹是否异常的标准可以根据实际需要进行确定。在一些实施例中，若轨迹的起点不同为异常，或者是轨迹所在的商户的消费数据高于或低于当前用户的消费能力为异常。在一些实施例中，可以结合所要确定的行为模式确定是否为异常位置比对结果。当要确定的行为模式为失业时，可以是第二时间段内连续3周的工作日行动轨迹与第三时间段内的工作日行动轨迹不同即为异常，也可以是起点不同，但是如果终点为同一属性例如同为办公楼则为正常。对于分居或离婚行为模式，可以是第二时间段内的工作日行动轨迹如果起点不同则为异常，也可以进一步判断当前用户的配偶的工作日的轨迹的起点是否与当前用户的起点相同，若不同，则位置比对结果为异常。Specifically, after obtaining the historical action track, the target action track is compared with the historical action track to confirm whether the track is abnormal. When performing trajectory comparison, if there are multiple types of historical action trajectories in the third time period, each target action trajectory and historical action trajectory can be compared to obtain the comparison result of each type of trajectory, and the current confirmation is confirmed. Whether the user's trajectory is abnormal. For example, the holiday action trajectory in the second time period is compared with the holiday action trajectory in the third time period. The working day trajectory in the second time period is compared with the working day trajectory in the third time period. The criteria for judging whether the trajectory is abnormal can be determined according to actual needs. In some embodiments, if the starting point of the trajectory is abnormal, or the consumption data of the merchant where the trajectory is located is higher or lower than the current user's spending power is abnormal. In some embodiments, whether or not the abnormal position alignment result is determined may be determined in conjunction with the behavior pattern to be determined. When the behavior mode to be determined is unemployed, it may be that the action trajectory of the working day for three consecutive weeks in the second time period is different from the action trajectory of the working day in the third time period, or may be an origin, but if the end point is different, For the same property, for example, the same as the office building is normal. For the separation or divorce behavior mode, it may be that the working day action trajectory in the second time period is abnormal if the starting point is different, and may further determine whether the starting point of the current user's spouse's working day trajectory is the same as the current user's starting point. Different, the position comparison result is abnormal.

应该理解的是，虽然上述流程图中的各个步骤按照箭头的指示依次显示，但是这些步骤并不是必然按照箭头指示的顺序依次执行。除非本文中有明确的说明，这些步骤的执行并没有严格的顺序限制，这些步骤可以以其它的顺序执行。而且，上述中的至少一部分步骤可以包括多个子步骤或者多个阶段，这些子步骤或者阶段并不必然是在同一时刻执行完成，而是可以在不同的时刻执行，这些子步骤或者阶段的执行顺序也不必然是依次进行，而是可以与其它步骤或者其它步骤的子步骤或者阶段的至少一部分轮流或者交替地执行。It should be understood that although the various steps in the above-described flowcharts are sequentially displayed as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Except as explicitly stated herein, the execution of these steps is not strictly limited, and the steps may be performed in other orders. Moreover, at least some of the steps may include a plurality of sub-steps or stages, which are not necessarily performed at the same time, but may be executed at different times, the execution order of the sub-steps or stages Nor is it necessarily performed sequentially, but may be performed alternately or alternately with at least a portion of other steps or sub-steps or stages of other steps.

在一些实施例中，如图8所示，提供了一种信息处理装置，包括：位置数据获取模块802、特征得到模块804、信用度得到模块806、对比模块808和策略得到模块810，其中：In some embodiments, as shown in FIG. 8, an information processing apparatus is provided, including: a location data obtaining module 802, a feature obtaining module 804, a credit obtaining module 806, a comparing module 808, and a policy obtaining module 810, where:

位置数据获取模块802，用于获取当前用户在第一时间段的多个第一位置数据以及当前用户在第二时间段的多个第二位置数据。The location data obtaining module 802 is configured to acquire a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period.

特征得到模块804，用于获取各个第一位置数据对应的消费数据，根据各个消费数据得到对应的消费特征。The feature obtaining module 804 is configured to obtain consumption data corresponding to each first location data, and obtain corresponding consumption features according to the respective consumption data.

信用度得到模块806，用于将消费特征输入预先训练得到的信用度模型中，得到当前用户对应的目标信用度。The credit obtaining module 806 is configured to input the consumption feature into the pre-trained credit model to obtain the target credit corresponding to the current user.

对比模块808，用于将第二位置数据与预设位置数据进行对比，得到异常位置比对结果。The comparison module 808 is configured to compare the second location data with the preset location data to obtain an abnormal location comparison result.

策略得到模块810，用于根据异常位置比对结果以及目标信用度得到当前用户对应的资源转移策略。The policy obtaining module 810 is configured to obtain a resource transfer policy corresponding to the current user according to the abnormal location comparison result and the target credit degree.

在一些实施例中，如图9所示，信用度得到模块806包括：In some embodiments, as shown in FIG. 9, the credit get module 806 includes:

初始信用度得到单元806A，用于将消费特征分别输入到多个预先训练得到的信用度模型中，得到各个信用度模型输出的当前用户的初始信用度。The initial credit obtaining unit 806A is configured to input the consumption features into the plurality of pre-trained credit models, respectively, to obtain the initial credits of the current users output by the respective credit models.

目标信用度得到单元806B，用于根据各个初始信用度以及对应的信用度模型的权重计算得到当前用户对应的目标信用度。The target credits obtaining unit 806B is configured to calculate the target credits corresponding to the current user according to the weights of the respective initial credits and the corresponding credit models.

在一些实施例中，如图10所示，信息处理装置还包括：In some embodiments, as shown in FIG. 10, the information processing apparatus further includes:

样本集获取模块1002，用于获取进行模型训练的样本集，样本集包括多个样本，样本包括多个训练消费特征以及对应的样本信用度。The sample set obtaining module 1002 is configured to obtain a sample set for performing model training, where the sample set includes a plurality of samples, and the sample includes a plurality of training consumption features and corresponding sample credits.

训练模块1004，用于根据样本集以及多种不同的模型训练方法进行模型训练，得到各个不同的模型训练方法训练得到的多个信用度模型。The training module 1004 is configured to perform model training according to a sample set and a plurality of different model training methods, and obtain a plurality of credit models trained by different model training methods.

模型信用度模块1006，用于将样本对应的消费特征输入到信用度模型中，得到样本对应的模型信用度。The model credit module 1006 is configured to input the consumption feature corresponding to the sample into the credit model to obtain a model credit corresponding to the sample.

权重得到模块1008，用于根据各个信用度模型中样本对应的模型信用度与样本信用度的差距得到各个信用度模型对应的权重。The weight obtaining module 1008 is configured to obtain weights corresponding to the respective credit models according to the difference between the model credits and the sample credits corresponding to the samples in the respective credit models.

在一些实施例中，权重得到模块1008用于：计算各个信用度模型中样本对应的模型信用度与样本信用度的偏差。对各个信用度模型对应的偏差进行求和计算，得到各个信用度模型对应的总偏差。根据各个信用度模型对应的总偏差以及预设的权重算法得到各个信用度模型对应的权重，其中，权重算法中总偏差与权重为负相关关系。In some embodiments, the weight obtaining module 1008 is configured to: calculate a deviation of a model credit corresponding to a sample in a respective credit model from a sample credit. The deviations corresponding to the respective credit models are summed to obtain the total deviation corresponding to each credit model. The weights corresponding to the respective credit models are obtained according to the total deviation corresponding to each credit model and the preset weight algorithm, wherein the total deviation and the weight in the weight algorithm are negatively correlated.

在一些实施例中，特征得到模块804包括：商户信息获取单元，用于获取各个第一位置数据对应的商户信息，商户信息包括商户的消费数据以及商户的商户类型。特征得到单元，用于根据各个商户的消费数据以及商户的消费类型得到对应的消费特征。In some embodiments, the feature obtaining module 804 includes: a merchant information acquiring unit, configured to acquire merchant information corresponding to each first location data, where the merchant information includes the merchant's consumption data and the merchant's merchant type. The feature obtaining unit is configured to obtain a corresponding consumption feature according to the consumption data of each merchant and the consumption type of the merchant.

在一些实施例中，预设位置数据包括当前用户在第三时间段的历史行动轨迹，第二位置数据包括当前用户在第二时间段的目标行动轨迹，对比模块808包括：历史轨迹获取单元，用于获取当前用户在第三时间段的历史行动轨迹，第三时间段所在的时间早于第二时间段所在的时间。比对单元，用于将当前用户在第二时间段的目标行动轨迹与当前用户在第三时间段的历史行动轨迹进行对比，得到异常位置比对结果。In some embodiments, the preset location data includes a historical action track of the current user in a third time period, the second location data includes a target action track of the current user in the second time period, and the comparison module 808 includes: a historical track acquiring unit, It is used to obtain a historical action track of the current user in the third time period, and the time of the third time period is earlier than the time of the second time period. The comparison unit is configured to compare the target action track of the current user in the second time period with the historical action track of the current user in the third time period to obtain an abnormal position comparison result.

在一些实施例中，策略得到模块810包括：初始数值得到单元，用于根据目标信用度得到当前用户对应的初始资源转移数值。行为模式确定单元，用于根据异常位置比对结果确定当前用户的行为模式。目标数值得到单元，用于根据行为模式以及初始资源转移数值得到当前用户对应的目标资源转移数值。In some embodiments, the policy obtaining module 810 includes: an initial value obtaining unit, configured to obtain an initial resource transfer value corresponding to the current user according to the target credit degree. The behavior mode determining unit is configured to determine a behavior mode of the current user according to the abnormal position comparison result. The target value obtaining unit is configured to obtain a target resource transfer value corresponding to the current user according to the behavior mode and the initial resource transfer value.

关于信息处理装置的具体限定可以参见上文中对于信息处理方法的限定，在此不再赘述。上述信息处理装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中，也可以以软件形式存储于计算机设备中的存储器中，以便于处理器调用执行以上各个模块对应的操作。For the specific definition of the information processing apparatus, reference may be made to the definition of the information processing method in the above, and details are not described herein again. The various modules in the above information processing apparatus may be implemented in whole or in part by software, hardware, and combinations thereof. Each of the above modules may be embedded in or independent of the processor in the computer device, or may be stored in a memory in the computer device in a software form, so that the processor invokes the operations corresponding to the above modules.

在一些实施例中，提供了一种计算机设备，该计算机设备可以是服务器，其内部结构图可以如图11所示。该计算机设备包括通过***总线连接的处理器、存储器和网络接口。其中， 该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质和内存储器。该非易失性存储介质存储有操作***和计算机可读指令。该内存储器为非易失性存储介质中的操作***和计算机可读指令的运行提供环境。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机可读指令被处理器执行时以实现一种信息处理方法。In some embodiments, a computer device is provided, which may be a server, the internal structure of which may be as shown in FIG. The computer device includes a processor, memory, and network interface coupled by a system bus. Wherein, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores operating systems and computer readable instructions. The internal memory provides an environment for operation of an operating system and computer readable instructions in a non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. The computer readable instructions are executed by a processor to implement an information processing method.

本领域技术人员可以理解，图11中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备的限定，具体的计算机设备可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。It will be understood by those skilled in the art that the structure shown in FIG. 11 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation of the computer device to which the solution of the present application is applied. The specific computer device may It includes more or fewer components than those shown in the figures, or some components are combined, or have different component arrangements.

一种计算机设备，包括存储器和一个或多个处理器，存储器中存储有计算机可读指令，计算机可读指令被处理器执行时实现本申请任意一个实施例中提供的信息处理方法的步骤。A computer apparatus comprising a memory and one or more processors having stored therein computer readable instructions that, when executed by a processor, implement the steps of the information processing method provided in any one of the embodiments of the present application.

一个或多个存储有计算机可读指令的非易失性存储介质，计算机可读指令被一个或多个处理器执行时，使得一个或多个处理器实现本申请任意一个实施例中提供的信息处理方法的步骤。One or more non-volatile storage media storing computer readable instructions, when executed by one or more processors, cause one or more processors to implement the information provided in any one embodiment of the present application The steps of the processing method.

本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机可读指令来指令相关的硬件来完成，前述的计算机可读指令可存储于一非易失性计算机可读取存储介质中，该计算机可读指令在执行时，可包括如上述各方法的实施例的流程。其中，本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用，均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限，RAM以多种形式可得，诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchl ink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。A person skilled in the art can understand that all or part of the process of implementing the above embodiments can be completed by computer-readable instructions for instructing related hardware, and the aforementioned computer readable instructions can be stored in a non-volatile computer. In reading a storage medium, the computer readable instructions, when executed, may include the flow of an embodiment of the methods described above. Any reference to a memory, storage, database or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. Non-volatile memory can include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of formats, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronization chain. Synchl ink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

以上实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。The technical features of the above embodiments may be arbitrarily combined. For the sake of brevity of description, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, It is considered to be the range described in this specification.

以上实施例仅表达了本申请的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本申请构思的前提下，还可以做出若干变形和改进，这些都属于本申请的保护范围。因此，本申请专利的保护范围应以所附权利要求为准。The above embodiments are merely illustrative of several embodiments of the present application, and the description thereof is more specific and detailed, but is not to be construed as limiting the scope of the invention. It should be noted that a number of variations and modifications may be made by those skilled in the art without departing from the spirit and scope of the present application. Therefore, the scope of the invention should be determined by the appended claims.

Claims (20)

1. 一种信息处理方法，包括：An information processing method includes:

   获取当前用户在第一时间段的多个第一位置数据以及所述当前用户在第二时间段的多个第二位置数据；Obtaining a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period;

   获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征；Obtaining consumption data corresponding to each of the first location data, and obtaining corresponding consumption features according to the respective consumption data;

   将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户对应的目标信用度；Entering the consumption feature into a pre-trained credit model to obtain a target credit corresponding to the current user;

   将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果；及Comparing the second location data with the preset location data to obtain an abnormal location comparison result; and

   根据所述异常位置比对结果以及所述目标信用度得到所述当前用户对应的资源转移策略。And obtaining, according to the abnormal location comparison result and the target credit degree, a resource transfer policy corresponding to the current user.
2. 根据权利要求1所述的方法，其特征在于，所述将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户的目标信用度，包括：The method according to claim 1, wherein the inputting the consumption feature into a pre-trained credit model to obtain the target credit of the current user comprises:

   将所述消费特征分别输入到多个预先训练得到的所述信用度模型中，得到所述各个信用度模型输出的所述当前用户的初始信用度；及Entering the consumption features into the plurality of pre-trained credit models to obtain initial credits of the current user output by the respective credit models; and

   根据所述各个初始信用度以及对应的信用度模型的权重计算得到所述当前用户对应的目标信用度。Calculating the target credits corresponding to the current user according to the respective initial credits and the weights of the corresponding credit models.
3. 根据权利要求2所述的方法，其特征在于，还包括：The method of claim 2, further comprising:

   获取进行模型训练的样本集，所述样本集包括多个样本，所述样本包括多个训练消费特征以及对应的样本信用度；Obtaining a sample set for performing model training, the sample set including a plurality of samples, the sample including a plurality of training consumption features and corresponding sample credits;

   根据所述样本集以及多种不同的模型训练方法进行模型训练，得到所述各个不同的模型训练方法训练得到的多个信用度模型；Performing model training according to the sample set and a plurality of different model training methods, and obtaining a plurality of credit models trained by the different model training methods;

   将所述样本对应的消费特征输入到所述信用度模型中，得到所述样本对应的模型信用度；及Entering a consumption characteristic corresponding to the sample into the credit model to obtain a model credit corresponding to the sample; and

   根据所述各个信用度模型中样本对应的模型信用度与所述样本信用度的差距得到所述各个信用度模型对应的权重。The weights corresponding to the respective credit models are obtained according to the difference between the model credits corresponding to the samples in the respective credit models and the sample credits.
4. 根据权利要求3所述的方法，其特征在于，所述根据所述各个模型中样本对应的模型信用度与所述样本信用度的差距得到所述各个模型对应的权重，包括：The method according to claim 3, wherein the weights corresponding to the models are obtained according to the difference between the model credits corresponding to the samples in the respective models and the sample credits, including:

   计算所述各个信用度模型中样本对应的模型信用度与所述样本信用度的偏差；Calculating deviations between model credits corresponding to samples in the respective credit models and the sample credits;

   对所述各个信用度模型对应的偏差进行求和计算，得到所述各个信用度模型对应的总偏差；及Performing a summation calculation on the deviations corresponding to the respective credit models to obtain a total deviation corresponding to the respective credit models; and

   根据所述各个信用度模型对应的总偏差以及预设的权重算法得到所述各个信用度模型对应的权重，其中，所述权重算法中总偏差与权重为负相关关系。The weights corresponding to the respective credit models are obtained according to the total deviation corresponding to the respective credit models and the preset weight algorithm, wherein the total deviation and the weight in the weight algorithm are negatively correlated.
5. 根据权利要求1所述的方法，其特征在于，所述获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征，包括：The method according to claim 1, wherein the obtaining the consumption data corresponding to the respective first location data, and obtaining the corresponding consumption features according to the respective consumption data, comprises:

   获取所述各个第一位置数据对应的商户信息，所述商户信息包括所述商户的消费数据以 及所述商户的商户类型；及Acquiring the merchant information corresponding to each of the first location data, where the merchant information includes the merchant's consumption data and the merchant's merchant type;

   根据所述各个商户的消费数据以及所述商户的消费类型得到对应的消费特征。Corresponding consumption characteristics are obtained according to the consumption data of the respective merchants and the consumption type of the merchant.
6. 根据权利要求1所述的方法，其特征在于，所述预设位置数据包括所述当前用户在第三时间段的历史行动轨迹，所述第二位置数据包括所述当前用户在第二时间段的目标行动轨迹，所述将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果，包括：The method according to claim 1, wherein the preset location data comprises a historical action track of the current user in a third time period, and the second location data comprises the current user in a second time period The target action track, the comparing the second position data with the preset position data to obtain an abnormal position comparison result, including:

   获取所述当前用户在所述第三时间段的历史行动轨迹，所述第三时间段所在的时间早于所述第二时间段所在的时间；及Acquiring a historical action track of the current user in the third time period, where the third time period is earlier than the time of the second time period; and

   将所述当前用户在所述第二时间段的目标行动轨迹与所述当前用户在所述第三时间段的历史行动轨迹进行对比，得到异常位置比对结果。Comparing the target action trajectory of the current user in the second time period with the historical action trajectory of the current user in the third time period, and obtaining an abnormal position comparison result.
7. 根据权利要求1所述的方法，其特征在于，所述根据所述异常位置比对结果以及所述目标信用度得到所述当前用户对应的资源转移策略，包括：The method according to claim 1, wherein the resource transfer policy corresponding to the current user is obtained according to the abnormal position comparison result and the target credit degree, including:

   根据所述目标信用度得到所述当前用户对应的初始资源转移数值；Obtaining an initial resource transfer value corresponding to the current user according to the target credit degree;

   根据所述异常位置比对结果确定所述当前用户的行为模式；及Determining a behavior mode of the current user according to the abnormal position comparison result; and

   根据所述行为模式以及所述初始资源转移数值得到所述当前用户对应的目标资源转移数值。And obtaining, according to the behavior mode and the initial resource transfer value, a target resource transfer value corresponding to the current user.
8. 一种信息处理装置，包括：An information processing apparatus comprising:

   位置数据获取模块，用于获取当前用户在第一时间段的多个第一位置数据以及所述当前用户在第二时间段的多个第二位置数据；a location data obtaining module, configured to acquire a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period;

   特征得到模块，用于获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征；a feature obtaining module, configured to acquire consumption data corresponding to each of the first location data, and obtain corresponding consumption features according to the respective consumption data;

   信用度得到模块，用于将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户对应的目标信用度；a credit obtaining module, configured to input the consumption feature into a pre-trained credit model, to obtain a target credit corresponding to the current user;

   对比模块，用于将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果；及a comparison module, configured to compare the second location data with the preset location data to obtain an abnormal location comparison result; and

   策略得到模块，用于根据所述异常位置比对结果以及所述目标信用度得到所述当前用户对应的资源转移策略。And a policy obtaining module, configured to obtain, according to the abnormal location comparison result and the target credit, a resource transfer policy corresponding to the current user.
9. 一种计算机设备，包括存储器及一个或多个处理器，所述存储器中储存有计算机可读指令，所述计算机可读指令被所述一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：A computer device comprising a memory and one or more processors having stored therein computer readable instructions, the computer readable instructions being executed by the one or more processors to cause the one or more The processors perform the following steps:

   获取当前用户在第一时间段的多个第一位置数据以及所述当前用户在第二时间段的多个第二位置数据；Obtaining a plurality of first location data of the current user in the first time period and a plurality of second location data of the current user in the second time period;

   获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征；Obtaining consumption data corresponding to each of the first location data, and obtaining corresponding consumption features according to the respective consumption data;

   将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户对应的目标信用度；Entering the consumption feature into a pre-trained credit model to obtain a target credit corresponding to the current user;

   将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果；及Comparing the second location data with the preset location data to obtain an abnormal location comparison result; and

   根据所述异常位置比对结果以及所述目标信用度得到所述当前用户对应的资源转移策略。And obtaining, according to the abnormal location comparison result and the target credit degree, a resource transfer policy corresponding to the current user.
10. 根据权利要求9所述的计算机设备，其特征在于，所述处理器所执行的所述将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户的目标信用度，包括：The computer device according to claim 9, wherein the performing, by the processor, the input of the consumption feature into a pre-trained credit model, and obtaining the target credit of the current user, comprising:

    将所述消费特征分别输入到多个预先训练得到的所述信用度模型中，得到所述各个信用度模型输出的所述当前用户的初始信用度；及Entering the consumption features into the plurality of pre-trained credit models to obtain initial credits of the current user output by the respective credit models; and

    根据所述各个初始信用度以及对应的信用度模型的权重计算得到所述当前用户对应的目标信用度。Calculating the target credits corresponding to the current user according to the respective initial credits and the weights of the corresponding credit models.
11. 根据权利要求10所述的计算机设备，其特征在于，所述计算机可读指令被所述处理器执行时还执行以下步骤：The computer apparatus according to claim 10, wherein said computer readable instructions are further executed by said processor to perform the following steps:

    获取进行模型训练的样本集，所述样本集包括多个样本，所述样本包括多个训练消费特征以及对应的样本信用度；Obtaining a sample set for performing model training, the sample set including a plurality of samples, the sample including a plurality of training consumption features and corresponding sample credits;

    根据所述样本集以及多种不同的模型训练方法进行模型训练，得到所述各个不同的模型训练方法训练得到的多个信用度模型；Performing model training according to the sample set and a plurality of different model training methods, and obtaining a plurality of credit models trained by the different model training methods;

    将所述样本对应的消费特征输入到所述信用度模型中，得到所述样本对应的模型信用度；及Entering a consumption characteristic corresponding to the sample into the credit model to obtain a model credit corresponding to the sample; and

    根据所述各个信用度模型中样本对应的模型信用度与所述样本信用度的差距得到所述各个信用度模型对应的权重。The weights corresponding to the respective credit models are obtained according to the difference between the model credits corresponding to the samples in the respective credit models and the sample credits.
12. 根据权利要求11所述的计算机设备，其特征在于，所述处理器所执行的所述根据所述各个模型中样本对应的模型信用度与所述样本信用度的差距得到所述各个模型对应的权重，包括：The computer device according to claim 11, wherein the difference between the model credits corresponding to the samples in the respective models and the sample credits performed by the processor obtains weights corresponding to the respective models, include:

    计算所述各个信用度模型中样本对应的模型信用度与所述样本信用度的偏差；Calculating deviations between model credits corresponding to samples in the respective credit models and the sample credits;

    对所述各个信用度模型对应的偏差进行求和计算，得到所述各个信用度模型对应的总偏差；及Performing a summation calculation on the deviations corresponding to the respective credit models to obtain a total deviation corresponding to the respective credit models; and

    根据所述各个信用度模型对应的总偏差以及预设的权重算法得到所述各个信用度模型对应的权重，其中，所述权重算法中总偏差与权重为负相关关系。The weights corresponding to the respective credit models are obtained according to the total deviation corresponding to the respective credit models and the preset weight algorithm, wherein the total deviation and the weight in the weight algorithm are negatively correlated.
13. 根据权利要求9所述的计算机设备，其特征在于，所述处理器所执行的所述获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征，包括：The computer device according to claim 9, wherein the acquiring, by the processor, the consumption data corresponding to the respective first location data, and obtaining corresponding consumption features according to the respective consumption data, includes:

    获取所述各个第一位置数据对应的商户信息，所述商户信息包括所述商户的消费数据以及所述商户的商户类型；及Acquiring the merchant information corresponding to each of the first location data, where the merchant information includes the merchant's consumption data and the merchant's merchant type;

    根据所述各个商户的消费数据以及所述商户的消费类型得到对应的消费特征。Corresponding consumption characteristics are obtained according to the consumption data of the respective merchants and the consumption type of the merchant.
14. 根据权利要求9所述的计算机设备，其特征在于，所述预设位置数据包括所述当前用户在第三时间段的历史行动轨迹，所述第二位置数据包括所述当前用户在第二时间段的目标行动轨迹，所述处理器所执行的所述将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果，包括：The computer device according to claim 9, wherein the preset location data comprises a historical action track of the current user during a third time period, and the second location data comprises the current user at a second time The target action track of the segment, the performing, by the processor, comparing the second location data with the preset location data to obtain an abnormal location comparison result, including:

    获取所述当前用户在所述第三时间段的历史行动轨迹，所述第三时间段所在的时间早于 所述第二时间段所在的时间；及Acquiring a historical action track of the current user in the third time period, where the third time period is earlier than the time of the second time period; and

    将所述当前用户在所述第二时间段的目标行动轨迹与所述当前用户在所述第三时间段的历史行动轨迹进行对比，得到异常位置比对结果。Comparing the target action trajectory of the current user in the second time period with the historical action trajectory of the current user in the third time period, and obtaining an abnormal position comparison result.
15. 一个或多个存储有计算机可读指令的非易失性计算机可读存储介质，所述计算机可读指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：One or more non-transitory computer readable storage mediums storing computer readable instructions, when executed by one or more processors, cause the one or more processors to perform the following steps:

    获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征；Obtaining consumption data corresponding to each of the first location data, and obtaining corresponding consumption features according to the respective consumption data;

    将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户对应的目标信用度；Entering the consumption feature into a pre-trained credit model to obtain a target credit corresponding to the current user;

    将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果；及Comparing the second location data with the preset location data to obtain an abnormal location comparison result; and

    根据所述异常位置比对结果以及所述目标信用度得到所述当前用户对应的资源转移策略。And obtaining, according to the abnormal location comparison result and the target credit degree, a resource transfer policy corresponding to the current user.
16. 根据权利要求15所述的存储介质，其特征在于，所述处理器所执行的所述将所述消费特征输入预先训练得到的信用度模型中，得到所述当前用户的目标信用度，包括：The storage medium according to claim 15, wherein the performing, by the processor, the input of the consumption feature into a pre-trained credit model, and obtaining the target credit of the current user, comprising:

    将所述消费特征分别输入到多个预先训练得到的所述信用度模型中，得到所述各个信用度模型输出的所述当前用户的初始信用度；及Entering the consumption features into the plurality of pre-trained credit models to obtain initial credits of the current user output by the respective credit models; and

    根据所述各个初始信用度以及对应的信用度模型的权重计算得到所述当前用户对应的目标信用度。Calculating the target credits corresponding to the current user according to the respective initial credits and the weights of the corresponding credit models.
17. 根据权利要求16所述的存储介质，其特征在于，所述计算机可读指令被所述处理器执行时还执行以下步骤：The storage medium of claim 16 wherein said computer readable instructions are further executed by said processor to perform the following steps:

    获取进行模型训练的样本集，所述样本集包括多个样本，所述样本包括多个训练消费特征以及对应的样本信用度；Obtaining a sample set for performing model training, the sample set including a plurality of samples, the sample including a plurality of training consumption features and corresponding sample credits;

    根据所述样本集以及多种不同的模型训练方法进行模型训练，得到所述各个不同的模型训练方法训练得到的多个信用度模型；Performing model training according to the sample set and a plurality of different model training methods, and obtaining a plurality of credit models trained by the different model training methods;

    将所述样本对应的消费特征输入到所述信用度模型中，得到所述样本对应的模型信用度；及Entering a consumption characteristic corresponding to the sample into the credit model to obtain a model credit corresponding to the sample; and

    根据所述各个信用度模型中样本对应的模型信用度与所述样本信用度的差距得到所述各个信用度模型对应的权重。The weights corresponding to the respective credit models are obtained according to the difference between the model credits corresponding to the samples in the respective credit models and the sample credits.
18. 根据权利要求17所述的存储介质，其特征在于，所述处理器所执行的所述根据所述各个模型中样本对应的模型信用度与所述样本信用度的差距得到所述各个模型对应的权重，包括：The storage medium according to claim 17, wherein the difference between the model credits corresponding to the samples in the respective models and the sample credits performed by the processor obtains weights corresponding to the respective models, include:

    计算所述各个信用度模型中样本对应的模型信用度与所述样本信用度的偏差；Calculating deviations between model credits corresponding to samples in the respective credit models and the sample credits;

    对所述各个信用度模型对应的偏差进行求和计算，得到所述各个信用度模型对应的总偏差；及Performing a summation calculation on the deviations corresponding to the respective credit models to obtain a total deviation corresponding to the respective credit models; and

    根据所述各个信用度模型对应的总偏差以及预设的权重算法得到所述各个信用度模型对应的权重，其中，所述权重算法中总偏差与权重为负相关关系。The weights corresponding to the respective credit models are obtained according to the total deviation corresponding to the respective credit models and the preset weight algorithm, wherein the total deviation and the weight in the weight algorithm are negatively correlated.
19. 根据权利要求15所述的存储介质，其特征在于，所述处理器所执行的所述获取所述各个第一位置数据对应的消费数据，根据所述各个消费数据得到对应的消费特征，包括：The storage medium according to claim 15, wherein the acquiring, by the processor, the consumption data corresponding to the respective first location data, and obtaining corresponding consumption features according to the respective consumption data, includes:

    获取所述各个第一位置数据对应的商户信息，所述商户信息包括所述商户的消费数据以及所述商户的商户类型；及Acquiring the merchant information corresponding to each of the first location data, where the merchant information includes the merchant's consumption data and the merchant's merchant type;

    根据所述各个商户的消费数据以及所述商户的消费类型得到对应的消费特征。Corresponding consumption characteristics are obtained according to the consumption data of the respective merchants and the consumption type of the merchant.
20. 根据权利要求15所述的存储介质，其特征在于，所述预设位置数据包括所述当前用户在第三时间段的历史行动轨迹，所述第二位置数据包括所述当前用户在第二时间段的目标行动轨迹，所述处理器所执行的所述将所述第二位置数据与预设位置数据进行对比，得到异常位置比对结果，包括：The storage medium according to claim 15, wherein the preset location data comprises a historical action track of the current user during a third time period, and the second location data comprises the current user at a second time The target action track of the segment, the performing, by the processor, comparing the second location data with the preset location data to obtain an abnormal location comparison result, including:

    获取所述当前用户在所述第三时间段的历史行动轨迹，所述第三时间段所在的时间早于所述第二时间段所在的时间；及Acquiring a historical action track of the current user in the third time period, where the third time period is earlier than the time of the second time period; and

    将所述当前用户在所述第二时间段的目标行动轨迹与所述当前用户在所述第三时间段的历史行动轨迹进行对比，得到异常位置比对结果。Comparing the target action trajectory of the current user in the second time period with the historical action trajectory of the current user in the third time period, and obtaining an abnormal position comparison result.

Images