Data processing method and device

Abstract

The invention discloses a data processing method and device. The method comprises the following steps of: acquiring a first occurrence moment, a last occurrence moment, a frequency accumulation item and a time attenuation accumulation item as initial parameters, receiving a data acquisition request, determining candidate position points according to position information, determining time sequence characteristics of each candidate position point according to the initial parameters and time information, and acquiring the time sequence characteristics of each candidate position point. And obtaining push data according to the time sequence characteristics of the candidate position points and sending the push data to the user terminal. Therefore, the calculation overhead can be reduced, and the accuracy of the pushed data is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a data processing method and device. Background Art

[0002] With the development of Internet technology, geographic information systems and services are gradually applied to various life scenarios, such as recommendation of points of interest, boarding point recommendation, travel departure point and destination prediction, etc. Since the geographical location points required by users will show a certain degree of aggregation over time, the characteristics of the geographical location points have an important impact on the pushed data. In the prior art, a commonly used method is to perform sliding window counting on the time axis through sliding window counting, and express the aggregation of event occurrences through the changes in count values ​​in different time intervals. However, the sliding window counting scheme has a large computational overhead, and serious information loss leads to low accuracy. Summary of the invention

[0003] In view of this, an object of an embodiment of the present invention is to provide a data processing method and device, which can reduce computing overhead and improve the accuracy of pushed data.

[0004] In a first aspect, an embodiment of the present invention provides a data processing method, the method comprising:

[0005] receiving a data acquisition request, wherein the data acquisition request includes location information and time information;

[0006] Determine a candidate location point according to the location information;

[0007] Determine the time series characteristics of each of the candidate position points according to the pre-acquired initial parameters and the time information, wherein the initial parameters include the first occurrence time, the last occurrence time, the frequency accumulation item and the time attenuation accumulation item;

[0008] Acquiring push data according to the time series characteristics of the candidate location points; and

[0009] The push data is sent to a user terminal.

[0010] In some embodiments, the initial parameters are obtained by the following steps:

[0011] Obtaining a time series of each location point, wherein the time series is a time series of a target event occurring at the location point;

[0012] Initial parameters are determined according to the time series.

[0013] In some embodiments, the time series includes multiple time points, and the multiple time points are arranged in chronological order.

[0014] In some embodiments, the first occurrence time is the first time point in the time series, the last occurrence time is the last time point in the time series, the frequency accumulation term is the number of time points in the time series, and the calculation formula of the time decay accumulation term is as follows:

[0015]

[0016] Among them, A is the time decay accumulation term, t i is the i-th time point in the time series, t n is the nth time point in the time series, n is the number of time points in the time series, K is a predetermined unit time, and i=1, 2, ..., n.

[0017] In some embodiments, determining the time series characteristics of each of the candidate position points according to the pre-acquired initial parameters and the time information includes:

[0018] Determining a time scale according to the time information and a first time point in the time series;

[0019] Determining a long-term signal based on the frequency accumulation term and the time scale;

[0020] Determine a short-term signal according to the time information, the last time point in the time series and the time decay accumulation item; and

[0021] The timing characteristics of each of the candidate position points are determined according to the long-term signal and the short-term signal.

[0022] In some embodiments, the time scale is calculated as follows:

[0023]

[0024] Among them, d t is the time scale, t is the time information, t1 is the first time point in the time series, and K is the predetermined unit time.

[0025] In some embodiments, the calculation formula of the long-term signal is as follows:

[0026]

[0027] Among them, μ t is the long-term signal, μ is the frequency accumulation term, d t is the time scale.

[0028] In some embodiments, the calculation formula of the short-term signal is as follows:

[0029]

[0030] Among them, A t is the short-term signal, A is the time-attenuated cumulative term, t n is the nth time point in the time series, t is the time information, K is the predetermined unit time, and K is the predetermined unit time.

[0031] In some embodiments, the calculation formula of the timing characteristics is as follows:

[0032] λ t =μ t +α*A t

[0033] Among them, λ t is the timing characteristic, μ t is the long-term signal, A t is the short-term signal, and α is a preset value.

[0034] In some embodiments, acquiring the push data according to the time series characteristics of the candidate location point includes:

[0035] determining an auxiliary feature, wherein the auxiliary feature includes one or more of time information, location information, and user information; and

[0036] Obtaining the recommendation probability of each of the candidate location points through a pre-trained prediction model according to the auxiliary features and the time series features; and

[0037] Push data is acquired according to the recommendation probability of each candidate location point, the push data includes a target location point, the target location point is one or more of the candidate location points, and the target location point includes one or more of a recommended boarding point, an interest location point, a departure point, and a destination.

[0038] In some embodiments, the method further comprises:

[0039] The initial parameters of the location point are updated according to new data, wherein the new data is that a target event occurs at the location point and the occurrence time is after the last time point in the time series.

[0040] In some embodiments, updating the initial parameters of the location point according to the new data includes:

[0041] Adding the occurrence time of the new data to the time series to update the time series; and

[0042] Get the updated initial parameters based on the updated time series.

[0043] In a second aspect, an embodiment of the present invention provides a data processing device, the device comprising:

[0044] A request receiving unit, configured to receive a data acquisition request, wherein the data acquisition request includes location information and time information;

[0045] A candidate location point determination unit, configured to determine a candidate location point according to the location information;

[0046] A time series feature determination unit, configured to determine the time series features of each of the candidate position points according to pre-acquired initial parameters and the time information, wherein the initial parameters include a first occurrence time, a last occurrence time, a frequency accumulation item, and a time attenuation accumulation item;

[0047] A push data acquisition unit, configured to acquire push data according to the time series characteristics of the candidate location points; and

[0048] The push data sending unit is used to send the push data to the user terminal.

[0049] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method described in the first aspect.

[0050] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing computer program instructions, wherein the computer program instructions implement the method described in the first aspect when executed by a processor.

[0051] The technical solution of the embodiment of the present invention obtains the first occurrence time, the last occurrence time, the frequency accumulation item and the time decay accumulation item as initial parameters, receives a data acquisition request, determines a candidate location point according to the location information, determines the time series characteristics of each candidate location point according to the initial parameters and the time information, obtains push data according to the time series characteristics of the candidate location point and sends it to the user terminal. In this way, the calculation overhead can be reduced and the accuracy of the push data can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The above and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:

[0053] Figure 1 is a schematic diagram of a data processing system according to an embodiment of the present invention;

[0054] Figure 2 is a flowchart of obtaining initial data according to an embodiment of the present invention;

[0055] Figure 3is a flow chart of a data processing method according to an embodiment of the present invention;

[0056] Figure 4 is a flow chart of obtaining timing characteristics according to an embodiment of the present invention;

[0057] Figure 5 is a flowchart of acquiring push data according to an embodiment of the present invention;

[0058] Figure 6 is a schematic diagram of a data processing device according to an embodiment of the present invention;

[0059] Figure 7 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The present invention is described below based on embodiments, but the present invention is not limited to these embodiments. In the detailed description of the present invention below, some specific details are described in detail. It is possible for a person skilled in the art to fully understand the present invention without the description of these details. In order to avoid confusing the essence of the present invention, known methods, processes, flows, components and circuits are not described in detail.

[0061] In addition, persons of ordinary skill in the art will appreciate that the drawings provided herein are for illustration purposes and are not necessarily drawn to scale.

[0062] Unless the context clearly requires otherwise, the words "include", "comprising" and the like throughout this application should be interpreted as including rather than exclusive or exhaustive; that is, as meaning "including but not limited to".

[0063] In the description of the present invention, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0064] Figure 1 Schematic diagram of a data processing system according to an embodiment of the present invention. Figure 1 In the illustrated embodiment, the data processing system includes a user terminal 1 and a server 2. The user terminal 1 and the server 2 are connected in communication.

[0065] The user terminal 1 sends a data acquisition request to the server, and the server 2 acquires corresponding push data according to the data acquisition request and sends the push data to the user terminal 1 .

[0066] The user terminal 1 is a terminal device used by a user, and the user may be a driver, a passenger, etc. The user terminal 1 may be implemented by a mobile phone, a tablet computer, a laptop computer, a desktop computer, or other dedicated data processing terminals.

[0067] The server 2 has a data processing function, and may be a separate server or a server cluster consisting of multiple servers.

[0068] The server 2 obtains the initial data of each location point in advance, and after receiving the data processing request sent by the user terminal 1, obtains the corresponding push data according to the initial data and sends it to the user terminal 1.

[0069] Specifically, Figure 2 is a flowchart of obtaining initial data according to an embodiment of the present invention. Figure 2 In the embodiment shown, the server obtains the initial data including the following steps:

[0070] Step S110: Obtain the time series of each location point.

[0071] In this embodiment, the time series is the time series in which the target event occurs at the location point.

[0072] Specifically, the target event can be any event. For example, in the online car-hailing scenario, the target event can be getting on the bus at a location, getting off the bus at a location, a predicted departure point, a predicted destination, etc. The location can be determined based on the location in the historical records, and all the location points where the target event occurred in the historical records are obtained. For example, in the scenario of predicting the location of the user's boarding point, all the boarding point locations in the historical records are obtained.

[0073] The server collects a data set that needs to describe time-varying information from a database, where the data set includes an identifier of a location point and a historical occurrence time of a target event at the location point.

[0074] In some embodiments, for a position point P a The data set can be shown as follows:

[0075] “Anonymous user 1, August 13, 2023 at 7:00, location P a ,arrive

[0076] Anonymous user 2, August 15, 2023 at 8:00, location P a ,arrive

[0077] Anonymous user 1, August 16, 2023 at 11:00, location P a ,arrive"

[0078] The data set includes one or more historical records, each of which includes a user ID, time, location information, and reachability, wherein the reachability includes two states: reached and not reached.

[0079] The location information includes longitude and latitude information (lng, lat), where lng is longitude and lat is latitude.

[0080] In some embodiments, the server obtains a data set within a predetermined time period, and the predetermined time period can be set according to actual conditions, such as one month, three months, or six months. At the same time, the historical records of reachability "arrival" within the predetermined time period are obtained, and a time series is generated according to the order of the time when the target event occurs at the location point.

[0081] For example, suppose that for location point P, in the data set within the predetermined time period, there are n historical records with reachability of "arrival", and the occurrence times of the n historical records are arranged in chronological order as t1, t2, ..., t n The time sequence is from morning to night. The time sequence T of the position point P is [t1, t2, ..., t n ]. Based on a similar method, the time series of each location point is obtained.

[0082] It should be noted that when the historical records in the data set of a certain location point are lower than a predetermined threshold, the location point may not be processed. The predetermined threshold may be any value greater than or equal to 2.

[0083] Step S120: determining initial parameters according to the time series.

[0084] In this embodiment, after the server obtains the time series of each location point, it generates the initial parameters corresponding to each location point according to the time series of each location point, wherein the initial parameters include the first occurrence time, the last occurrence time, the frequency accumulation item and the time decay accumulation item.

[0085] The first occurrence time is the first time point in the time series.

[0086] The last occurrence time is the last time point in the time series.

[0087] The frequency accumulation item is the number of time points in the time series.

[0088] The calculation formula of the time decay accumulation term is as follows:

[0089]

[0090] Among them, A is the time decay accumulation term, t iis the i-th time point in the time series, t n is the nth time point in the time series, n is the number of time points in the time series, K is a predetermined unit time, and i=1, 2, ..., n.

[0091] in, Indicates that x is rounded down. K is a predetermined time unit, for example, K may be 6*3600, in seconds, and K represents a time length of 6 hours.

[0092] Assume that the time series T of the position point P is [t1, t2, ..., t n ], the first occurrence time is t1, and the last occurrence time is t n , the frequency accumulation term μ=n, and the time attenuation accumulation term A are calculated and obtained by the above formula.

[0093] In a specific example, assuming n=3, the time series T=[7:00 on August 13, 2023, 8:00 on August 15, 2023, 11:00 on August 16, 2023]. The time in the time series is converted into numerical values, from 00:00:00 on January 1, 1970 to the number of seconds at the time point. After conversion, the time series T=[1691881200, 1692057600, 1692154800].

[0094] Therefore, the calculation formula of the time decay accumulation term is as follows:

[0095]

[0096] Therefore, the initial parameters of each position point can be obtained through the above method, and then data processing is performed according to the initial parameters.

[0097] In some embodiments, the server obtaining the initial data further comprises the following steps:

[0098] Step S130: Update the initial parameters of the location point according to the new data.

[0099] In this embodiment, the new data is that the target event occurs at the location point and the occurrence time is after the last time point in the time series.

[0100] Assume that the initial parameters of a certain location point are obtained according to the time series T = [7:00 on August 13, 2023, 8:00 on August 15, 2023, 11:00 on August 16, 2023], and the last time point in the time series is 11:00 on August 16, 2023. If the target event occurs at the location point again after this time, the data corresponding to this target time is new data.

[0101] Furthermore, updating the initial parameters includes:

[0102] Step S131: Add the occurrence time of the new data to the time series to update the time series.

[0103] In this embodiment, the occurrence time of the new data is added to the time series to update the time series.

[0104] Assume that the initial parameters of a certain location point are obtained according to the time series T = [7:00 on August 13, 2023, 8:00 on August 15, 2023, 11:00 on August 16, 2023]. n+1 = The target event occurs at 06:00 on August 17, 2023, then the occurrence time corresponding to the new data is 06:00 on August 17, 2023. 06:00 on August 17, 2023 is added to the time series T to form a new time series T'=[07:00 on August 13, 2023, 08:00 on August 15, 2023, 11:00 on August 16, 2023, 06:00 on August 17, 2023].

[0105] Step S132: Obtain updated initial parameters according to the updated time series.

[0106] In this embodiment, updated initial parameters are obtained according to the updated time series.

[0107] Assume that the updated time series T' = [7:00 on August 13, 2023, 8:00 on August 15, 2023, 11:00 on August 16, 2023, 06:00 on August 17, 2023]. Then the updated initial parameters include:

[0108] The time of first occurrence remains unchanged, and is still the first time point in the time series t1 = 7:00 on August 13, 2023.

[0109] The last occurrence time becomes t n = 06:00 on August 17, 2023. At this time, n=4.

[0110] The frequency accumulation item is the number of time points in the time series, which is updated to 4.

[0111] The time decay accumulation term A=1.0507.

[0112] The specific calculation method can refer to the above formula, and the embodiment of the present invention will not be repeated here.

[0113] The technology of efficiently modeling the changes of an event over time is an important part of the geographic information system. For example, different exits of the same shopping mall may be used to depart at different times. In the bus travel scenario on the map, the geographical location points required by the user will show a certain degree of clustering over time. For example, at the "day" granularity, there are visits to the office building during the week and to the shopping mall on weekends, and at the "hour" granularity, there are departures from the east gate in the morning and the west gate in the afternoon. The present invention characterizes the clustering of events by modeling the dependencies between events, and performs lossless information updates in a recursive form, and can ultimately respond quickly and at low cost to changes in user needs over time.

[0114] The existing technology mainly uses sliding window counting to count the time axis, and expresses the clustering of events through the change of count values ​​in different time intervals. In terms of calculation logic, the sliding window counting scheme generally directly discards the "out-of-window data" or accumulates it after multiplying it by a fixed discount coefficient, which will cause uneven information updates. When updating data, there are two schemes: "streaming real-time update" and "batch scheduled update". Among them, the "streaming real-time update" method has the problem of high storage overhead, while the "batch scheduled update" method has shortcomings in real-time performance.

[0115] Through the method of the embodiment of the present invention, time series information can be progressively characterized to achieve low-cost and smooth information updates. The problems of high overhead, information loss and non-smoothness in the current solution are solved. The Hawkes (a mathematical model for modeling self-excitation processes) process is used to characterize the characteristics of the gradual aggregation of target events in the time domain, and a completely equivalent low-cost real-time update solution is obtained through mathematical rewriting. Progressive, low-cost and lossless characterization of time series information meets the low-cost and high-efficiency requirements of large-scale geographic information systems and services when responding to time-varying information of geographic locations.

[0116] Figure 3 is a flow chart of a data processing method according to an embodiment of the present invention. Figure 3 As shown, the data processing method of the embodiment of the present invention includes the following steps:

[0117] Step S200: receiving a data acquisition request.

[0118] In this embodiment, the server receives a data acquisition request sent by a user terminal, where the data acquisition request includes location information and time information.

[0119] Specifically, when a user has a need to obtain data, a data acquisition request is sent to a server through a user terminal.

[0120] For example, in the online car-hailing scenario, the data acquisition request can be a user's online car-hailing order request. For an online car-hailing order request, the location information includes the departure place and the destination, and the time information includes the departure time and the arrival time, where the departure place and the destination are entered by the user. The departure time is the current time, or the departure time is the predicted time for the user to get on the car based on the driver's location and the departure place after the server dispatches the order, or when the order is a reservation order, the departure time is the reservation time entered by the user. The arrival time is the predicted time to arrive at the destination.

[0121] Step S300: Determine a candidate location point according to the location information.

[0122] In this embodiment, the server determines candidate locations based on the location information, and the candidate locations are locations where the target event may occur. For example, the server needs to predict the user's boarding point as an example, the location information is the user's departure point, and the candidate locations are the boarding points within a predetermined range around the user's departure point. Specifically, assuming that the departure point in the data acquisition request is S a , when obtaining candidate location points, all boarding points within a certain range around the departure place Sa are obtained according to historical data, and these boarding points are candidate location points.

[0123] Step S400: determining the time series characteristics of each of the candidate position points according to the pre-acquired initial parameters and the time information.

[0124] In this embodiment, after the server obtains the candidate location points, it obtains the initial parameters corresponding to each location point. The method for obtaining the initial parameters is as described above, and the embodiment of the present invention will not be repeated here. The time information is the time when the target event occurs. Taking the online car-hailing scenario as an example, when predicting the user's boarding point, the time information is the time of the data acquisition request, which can be the sending time of the data acquisition request, the receiving time of the data acquisition request, or the appointment time carried in the data acquisition request. The embodiment of the present invention is explained by taking the time information as the current time as an example. The initial parameters include the first occurrence time, the last occurrence time, the frequency accumulation item, and the time attenuation accumulation item.

[0125] Figure 4 FIG. 1 is a flow chart of obtaining timing characteristics according to an embodiment of the present invention. Figure 4 In the illustrated embodiment, determining the time series features of each of the candidate position points according to the pre-acquired initial parameters and the time information includes the following steps:

[0126] Step S410: determine a time scale according to the time information and the first time point in the time series.

[0127] The time scale is the ratio of the difference between the time information and the first time point in the time series to a predetermined unit time.

[0128] The time scale is calculated as follows:

[0129]

[0130] Among them, d t is the time scale, t is the time information, t1 is the first time point in the time series, and K is the predetermined unit time.

[0131] Assume that the time series of a candidate location point is T = [7:00 on August 13, 2023, 8:00 on August 15, 2023, 11:00 on August 16, 2023]. Convert the time in the time series to the number of seconds from 00:00:00 on January 1, 1970 to the time point. After conversion, the time series T = [1691881200, 1692057600, 1692154800], that is, the first time point in the time series t1 = 1691881200. The time information is 19:00 on August 16, 2023. After conversion, the time information t = 1692183600. The time scale d is calculated by the above formula t =14.

[0132] Step S420: Determine a long-term signal according to the frequency accumulation item and the time scale.

[0133] The long-term signal is the ratio of the frequency accumulation item to the time scale.

[0134] Furthermore, the calculation formula of the long-term signal is as follows:

[0135]

[0136] Among them, μ t is the long-term signal, μ is the frequency accumulation term, d t is the time scale.

[0137] The example in step S410 is also used for illustration. As mentioned above, the time scale d t =14, at the same time, the frequency accumulation term μ=3, and the long-term signal μ is obtained through calculation t ≈0.2142.

[0138] Step S430: determine a short-term signal according to the time information, the last time point in the time series and the time attenuation accumulation item.

[0139] The calculation formula of the short-term signal is as follows:

[0140]

[0141] Among them, A t is the short-term signal, A is the time-attenuated cumulative term, t n is the nth time point in the time series, t is the time information, K is the predetermined unit time, and K is the predetermined unit time.

[0142] The example in step S420 is also used for illustration. As mentioned above, the time information t=1692183600, that is, the last time point t in the time series n =1692154800, time decay accumulation term A = 1.0183, K = 3600*6. The short-term signal A is obtained by the above formula. t =0.3746.

[0143] Step S440: Determine the timing characteristics of each of the candidate position points according to the long-term signal and the short-term signal.

[0144] The time series feature is the product of the short-term signal and the dependency value and the sum of the long-term signal.

[0145] Furthermore, the calculation formula of the time series feature is as follows:

[0146] λ t =μ t +α*A t

[0147] Among them, λ t is the timing characteristic, μ t is the long-term signal, A t is the short-term signal, and α is a preset value.

[0148] Furthermore, α is a preset dependency value, which represents the dependency of the location point with the short-term signal.

[0149] Taking the example in step S430 as an example, as mentioned above, the long-term signal μ t ≈0.2142, short-term signal A t =0.3746, assuming α = 1, then the time series feature λ t =0.5888.

[0150] In this way, the time series characteristics of each candidate position point can be obtained.

[0151] Step S500: acquiring push data according to the time series characteristics of the candidate location points.

[0152] In this embodiment, the server obtains push data according to the acquired time series characteristics of each candidate location point.

[0153] Specifically, Figure 5 FIG. 1 is a flowchart of acquiring push data according to an embodiment of the present invention. Figure 5 As shown, the server obtains push data according to the time series characteristics of the candidate location point, including the following steps:

[0154] Step S510: Determine auxiliary features.

[0155] In this embodiment, the auxiliary features include one or more of time information, location information, and user information.

[0156] Furthermore, the server collects other features related to the target event, such as time information, location information, user information, road information, road congestion, etc., and generates corresponding feature vectors, which are auxiliary features y.

[0157] Step S520: obtaining the recommendation probability of each of the candidate location points through a pre-trained prediction model according to the auxiliary features and the time series features.

[0158] In this embodiment, a pre-trained prediction model is obtained, and the auxiliary features and the time series features are input into the prediction model, and the prediction model outputs the recommendation probability of each position point according to the auxiliary features and the time series features.

[0159] Taking the prediction of boarding points as an example, the prediction model outputs the probability of boarding at each candidate location.

[0160] In some embodiments, the prediction model is a multi-layer perceptron (formally represented as f), and the recommendation probability f(λ) of each candidate position point can be obtained. t , y).

[0161] Among them, Multilayer Perceptron (MLP) is a deep learning model based on Feedforward Neural Network, which consists of multiple neuron layers, each of which is fully connected to the previous layer. Multilayer Perceptron can be used to solve various machine learning problems such as classification, regression and clustering. Each neuron layer of the multilayer perceptron is composed of many neurons, where the input layer receives input features, the output layer gives the final prediction results, and the hidden layer in the middle is used to extract features and perform nonlinear transformations. Each neuron receives the output of the previous layer, performs weighted sum and activation function operations, and obtains the output of the current layer. Through continuous iterative training, the multilayer perceptron can automatically learn the complex relationship between input features and predict new data.

[0162] Step S530: acquiring push data according to the recommendation probability of each candidate location point.

[0163] In this embodiment, the push data includes a target location point, which is one or more of the candidate location points, and the target location point includes one or more of a recommended boarding point, an interest location point, a departure point, and a destination. Among them, the recommended boarding point is the boarding point location recommended to the user terminal in the online car-hailing scenario. The interest location point can be a variety of locations. For example, when a user needs to search for an interest location point near a certain location (such as entertainment, shopping malls, food, parking lots, accommodation, subway stations / bus stations, gas stations, etc.), the interest location point can be obtained in the above manner.

[0164] After obtaining the recommendation probability of each candidate location point, the server selects a target recommendation point from the candidate location points according to the recommendation probability, and generates push data according to the target recommendation point.

[0165] In some embodiments, the push data also includes a ranking of each target recommendation point, and the ranking of the target recommendation points can be sorted according to the recommendation probability.

[0166] The target recommended point may be all the candidate position points, or the target recommended point may be a part of the candidate position points.

[0167] Step S600: Send the push data to the user terminal.

[0168] In this embodiment, the server sends the acquired push data to the user terminal, and the user terminal displays the push data.

[0169] In a specific scenario, after the user places an order, the server determines the recommended boarding point based on the candidate locations around the departure point and the recommended alighting point based on the candidate locations around the destination according to the data acquisition request, and the data push is completed.

[0170] The embodiment of the present invention obtains the first occurrence time, the last occurrence time, the frequency accumulation item and the time decay accumulation item as initial parameters, receives a data acquisition request, determines a candidate location point according to the location information, determines the time series characteristics of each candidate location point according to the initial parameters and the time information, obtains push data according to the time series characteristics of the candidate location point and sends it to the user terminal. In this way, the calculation overhead can be reduced and the accuracy of the push data can be improved.

[0171] Figure 6 Schematic diagram of a data processing device according to an embodiment of the present invention. Figure 6In the illustrated embodiment, the data processing device includes a request receiving unit 61, a candidate location point determination unit 62, a timing feature determination unit 63, a push data acquisition unit 64, and a push data sending unit 65. The request receiving unit 61 is used to receive a data acquisition request, and the data acquisition request includes location information and time information. The candidate location point determination unit 62 is used to determine the candidate location point according to the location information. The timing feature determination unit 63 is used to determine the timing features of each of the candidate location points according to the pre-acquired initial parameters and the time information, and the initial parameters include the first occurrence time, the last occurrence time, the frequency accumulation item, and the time attenuation accumulation item. The push data acquisition unit 64 is used to acquire push data according to the timing features of the candidate location points. The push data sending unit 65 is used to send the push data to the user terminal.

[0172] In some embodiments, the apparatus further comprises:

[0173] A time series acquisition unit, used to acquire a time series of each location point, wherein the time series is a time series of a target event occurring at the location point;

[0174] An initial parameter determination unit is used to determine initial parameters according to the time series.

[0175] In some embodiments, the time series includes multiple time points, and the multiple time points are arranged in chronological order.

[0176] In some embodiments, the first occurrence time is the first time point in the time series, the last occurrence time is the last time point in the time series, the frequency accumulation term is the number of time points in the time series, and the calculation formula of the time decay accumulation term is as follows:

[0177]

[0178] Among them, A is the time decay accumulation term, t i is the i-th time point in the time series, t n is the nth time point in the time series, n is the number of time points in the time series, K is a predetermined unit time, and i=1, 2, ..., n.

[0179] In some embodiments, the timing feature determination unit includes:

[0180] A time scale determination subunit, configured to determine a time scale according to the time information and a first time point in the time series;

[0181] A long-term signal determination subunit, used to determine a long-term signal according to the frequency accumulation item and the time scale;

[0182] a short-term signal determination subunit, configured to determine a short-term signal according to the time information, the last time point in the time series, and the time attenuation accumulation item; and

[0183] The time series feature acquisition subunit is used to determine the time series feature of each of the candidate position points according to the long-term signal and the short-term signal.

[0184] In some embodiments, the time scale determination subunit obtains the time scale by the following formula:

[0185]

[0186] Among them, d t is the time scale, t is the time information, t1 is the first time point in the time series, and K is the predetermined unit time.

[0187] In some embodiments, the long-term signal determination subunit obtains the long-term signal by the following formula:

[0188]

[0189] Among them, μ t is the long-term signal, μ is the frequency accumulation term, d t is the time scale.

[0190] In some embodiments, the short-term signal determination subunit obtains the short-term signal by the following formula:

[0191]

[0192] Among them, A t is the short-term signal, A is the time-attenuated cumulative term, t n is the nth time point in the time series, t is the time information, K is the predetermined unit time, and K is the predetermined unit time.

[0193] In some embodiments, the timing feature acquisition subunit acquires the timing feature by the following formula:

[0194] λ t =μ t +α*A t

[0195] Among them, λ t is the timing characteristic, μ t is the long-term signal, A t is the short-term signal, and α is a preset value.

[0196] In some embodiments, the push data acquisition unit includes:

[0197] an auxiliary feature acquisition subunit, used to determine auxiliary features, wherein the auxiliary features include one or more of time information, location information, and user information;

[0198] a recommendation probability acquisition subunit, configured to acquire the recommendation probability of each of the candidate location points through a pre-trained prediction model according to the auxiliary features and the time series features; and

[0199] The push data acquisition subunit is used to acquire push data according to the recommendation probability of each candidate location point, and the push data includes a target location point, which is one or more of the candidate location points, and the target location point includes one or more of a recommended boarding point, an interest location point, a departure point, and a destination.

[0200] In some embodiments, the apparatus further comprises:

[0201] The initial parameter updating unit is used to update the initial parameters of the location point according to new data, wherein the new data is that the target event occurs at the location point and the occurrence time is after the last time point in the time series.

[0202] In some embodiments, the initial parameter updating unit includes:

[0203] A first updating subunit, configured to add the occurrence time of the new data to the time series to update the time series; and

[0204] The second updating subunit is used to obtain updated initial parameters according to the updated time series.

[0205] The embodiment of the present invention obtains the first occurrence time, the last occurrence time, the frequency accumulation item and the time decay accumulation item as initial parameters, receives a data acquisition request, determines a candidate location point according to the location information, determines the time series characteristics of each candidate location point according to the initial parameters and the time information, obtains push data according to the time series characteristics of the candidate location point and sends it to the user terminal. In this way, the calculation overhead can be reduced and the accuracy of the push data can be improved.

[0206] Figure 7 is a schematic diagram of an electronic device according to an embodiment of the present invention. Figure 7The electronic device shown is a general data processing device, which includes a general computer hardware structure, which at least includes a processor 71 and a memory 72. The processor 71 and the memory 72 are connected via a bus 73. The memory 72 is suitable for storing instructions or programs executable by the processor 71. The processor 71 can be an independent microprocessor or a collection of one or more microprocessors. Thus, the processor 71 executes the instructions stored in the memory 72, thereby executing the method flow of the embodiment of the present invention as described above to realize the processing of data and the control of other devices. The bus 73 connects the above-mentioned multiple components together, and at the same time connects the above-mentioned components to the display controller 74 and the display device and the input / output (I / O) device 75. The input / output (I / O) device 75 can be a mouse, a keyboard, a modem, a network interface, a touch input device, a somatosensory input device, a printer, and other devices known in the art. Typically, the input / output device 75 is connected to the system via an input / output (I / O) controller 76.

[0207] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, devices (equipment) or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0208] The present invention is described with reference to flowcharts of methods, apparatuses (devices) and computer program products according to embodiments of the present application. It should be understood that each process in the flowchart can be implemented by computer program instructions.

[0209] These computer program instructions may be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device that implements the process Figure 1 A function specified in a process or multiple processes.

[0210] These computer program instructions may also be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce the instructions for implementing the process Figure 1 A device that specifies functions in a process or multiple processes.

[0211] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data processing method, characterized in that: The method comprises: receiving a data acquisition request, wherein the data acquisition request includes location information and time information; Determine a candidate location point according to the location information; Determine the time series characteristics of each of the candidate position points according to the pre-acquired initial parameters and the time information, wherein the initial parameters include the first occurrence time, the last occurrence time, the frequency accumulation item and the time attenuation accumulation item; Acquiring push data according to the time series characteristics of the candidate location points; and The push data is sent to a user terminal.

2. The method according to claim 1, characterized in that The initial parameters are obtained by the following steps: Obtaining a time series of each location point, wherein the time series is a time series of a target event occurring at the location point; Initial parameters are determined according to the time series.

3. The method according to claim 2, characterized in that The time series includes multiple time points, and the multiple time points are arranged in chronological order.

4. The method according to claim 3, characterized in that The first occurrence time is the first time point in the time series, the last occurrence time is the last time point in the time series, the frequency accumulation term is the number of time points in the time series, and the calculation formula of the time decay accumulation term is as follows: Among them, A is the time decay accumulation term, t i is the i-th time point in the time series, t n is the nth time point in the time series, n is the number of time points in the time series, K is a predetermined unit time, and i=1, 2, ..., n.

5. The method according to claim 4, characterized in that The determining of the time series characteristics of each of the candidate position points according to the pre-acquired initial parameters and the time information includes: Determining a time scale according to the time information and a first time point in the time series; Determining a long-term signal based on the frequency accumulation term and the time scale; Determine a short-term signal according to the time information, the last time point in the time series and the time decay accumulation item; and The timing characteristics of each of the candidate position points are determined according to the long-term signal and the short-term signal.

6. The method according to claim 5, characterized in that The time scale is calculated as follows: Among them, d t is the time scale, t is the time information, t1 is the first time point in the time series, and K is the predetermined unit time.

7. The method according to claim 5, characterized in that The calculation formula of the long-term signal is as follows: Among them, μ t is the long-term signal, μ is the frequency accumulation term, d t is the time scale.

8. The method according to claim 5, characterized in that The calculation formula of the short-term signal is as follows: Among them, A t is the short-term signal, A is the time-attenuated cumulative term, t n is the nth time point in the time series, t is the time information, K is the predetermined unit time, and K is the predetermined unit time.

9. The method according to claim 5, characterized in that The calculation formula of the timing characteristics is as follows: l t =μ t +a*A t Among them, λ t is the timing characteristic, μ t is the long-term signal, A t is the short-term signal, and α is a preset value.

10. The method according to claim 1, characterized in that The acquiring of push data according to the time series characteristics of the candidate location point comprises: Determine auxiliary features, where the auxiliary features include one or more of time information, location information, and user information; Obtaining the recommendation probability of each of the candidate location points through a pre-trained prediction model according to the auxiliary features and the time series features; and Push data is acquired according to the recommendation probability of each candidate location point, the push data includes a target location point, the target location point is one or more of the candidate location points, and the target location point includes one or more of a recommended boarding point, an interest location point, a departure point, and a destination.

11. The method according to claim 2, characterized in that The method further comprises: The initial parameters of the location point are updated according to new data, wherein the new data is that a target event occurs at the location point and the occurrence time is after the last time point in the time series.

12. The method according to claim 11, characterized in that The updating of the initial parameters of the location points according to the new data comprises: Adding the occurrence time of the new data to the time series to update the time series; and Get the updated initial parameters based on the updated time series.

13. A data processing device, characterized in that: The device comprises: A request receiving unit, configured to receive a data acquisition request, wherein the data acquisition request includes location information and time information; A candidate location point determination unit, configured to determine a candidate location point according to the location information; A time series feature determination unit, configured to determine the time series features of each of the candidate position points according to pre-acquired initial parameters and the time information, wherein the initial parameters include a first occurrence time, a last occurrence time, a frequency accumulation item, and a time attenuation accumulation item; a push data acquisition unit, configured to acquire push data according to the time series characteristics of the candidate location points; and The push data sending unit is used to send the push data to the user terminal.

14. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1 to 12.

15. A computer-readable storage medium storing computer program instructions, characterized in that: The computer program instructions implement the method according to any one of claims 1 to 12 when executed by a processor.

Images