CN116307227A - Service information processing method, device and computer equipment - Google Patents
Service information processing method, device and computer equipment Download PDFInfo
- Publication number
- CN116307227A CN116307227A CN202310316723.1A CN202310316723A CN116307227A CN 116307227 A CN116307227 A CN 116307227A CN 202310316723 A CN202310316723 A CN 202310316723A CN 116307227 A CN116307227 A CN 116307227A
- Authority
- CN
- China
- Prior art keywords
- service processing
- service
- time
- model
- historical
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Data Mining & Analysis (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present application relates to a service information processing method, apparatus, computer device, storage medium, and computer program product. The method comprises the following steps: acquiring a historical service information sequence; inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model; and inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point. The method can improve the service processing efficiency.
Description
Technical Field
The present invention relates to the field of computer technology and the field of financial technology, and in particular, to a business information processing method, apparatus, computer device, storage medium, and computer program product.
Background
With the development of computer technology, when a user handles related services in a service organization, service processing record data is often left in a service information recording system, for example, when the user handles services in the service organization, the user needs to take a number at a number taking device and wait for a service window to call a number.
At present, when a user calls a number of a waiting window in a service organization, the queuing time of each service window cannot be accurately predicted, so that the user often spends a great deal of time, and the service processing efficiency of the service organization is low.
Therefore, the conventional technology has a problem of low service processing efficiency.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a business information processing method, apparatus, computer device, computer readable storage medium, and computer program product that are capable of accurately determining a business processing waiting time.
A business information processing method, characterized in that the method comprises:
acquiring a historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing characteristic data comprises characteristic data generated in the service processing window service processing process;
Inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model;
and inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point.
In one embodiment, inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point, including:
inputting the historical service information sequence into a time sequence season model to obtain a first output result;
inputting the historical service information sequence into a time sequence trend model to obtain a second output result;
inputting the historical service information sequence into a time sequence holiday model to obtain a third output result;
And obtaining an initial prediction result aiming at the waiting time of the service processing window at the target time point according to the first output result, the second output result and the third output result.
In one embodiment, obtaining an initial prediction result for the waiting time of the service processing window at the target time point according to the first output result, the second output result and the third output result includes:
fusing the first output result, the second output result and the third output result to obtain a fused result;
and taking the fusion result as an initial prediction result.
In one embodiment, the method further comprises:
acquiring original historical service information; the original historical service information comprises original service processing characteristic data corresponding to a preset number of historical time points in a preset time period before a target time point;
and carrying out numerical coding on the original service processing characteristic data to obtain a historical service information sequence.
In one embodiment, the performing numeric encoding on the original service processing feature data to obtain a historical service information sequence includes:
performing numerical coding on the original service processing characteristic data to obtain numerical coding service processing characteristic data;
And performing feature de-duplication processing on the feature data of the numerical coding service processing to obtain a historical service information sequence.
In one embodiment, performing feature deduplication processing on the digitized encoded service processing feature data to obtain a historical service information sequence, including:
performing characteristic duplication elimination processing on the numerical coding service processing characteristic data to obtain duplicated numerical coding service processing characteristic data;
and carrying out normalization processing on the de-duplicated numerical coding service processing characteristic data to obtain a historical service information sequence.
In one embodiment, performing feature deduplication processing on the digitized encoded service processing feature data to obtain the digitized encoded service processing feature data after deduplication, where the feature deduplication processing includes:
according to the numerical coding service processing characteristic data, determining weight information corresponding to each service processing characteristic;
determining at least one target service processing characteristic according to the weight information corresponding to each service processing characteristic;
and taking the numerical coding service processing characteristic data corresponding to each target service processing characteristic as the numerical coding service processing characteristic data after de-duplication.
A service information processing apparatus, characterized in that the apparatus comprises:
The acquisition module is used for acquiring the historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing characteristic data comprises characteristic data generated in the service processing window service processing process;
the output module is used for inputting the historical service information sequence into the pre-trained first waiting time output model to obtain an initial prediction result of waiting time of the service processing window at a target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model;
the prediction module is used for inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point.
A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method described above when executing the computer program.
A computer readable storage medium having stored thereon a computer program, characterized in that the computer program when executed by a processor realizes the steps of the above-mentioned method.
A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method described above.
The service information processing method, the device, the computer equipment, the storage medium and the computer program product acquire a historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing characteristic data comprises characteristic data generated in the service processing window service processing process; inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model; inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at a target time point; therefore, the initial prediction result output by the first waiting time output model and the output result of the time sequence trend model and the historical service information sequence are input into the second waiting time output model, so that the time characteristics in the historical service information sequence are fully considered, the target prediction result output by the second waiting time output model is more accurate, the service processing waiting time can be accurately determined, and the service processing efficiency is improved.
Drawings
FIG. 1 is an application environment diagram of a business information processing method in one embodiment;
FIG. 2 is a flow chart of a method for processing service information in one embodiment;
FIG. 3 is a flowchart of a method for processing service information according to another embodiment;
FIG. 4 is a block diagram showing a construction of a service information processing apparatus in one embodiment;
fig. 5 is an internal structural diagram of a computer device in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
It should be noted that the present application discloses a business information processing method, a device, a computer apparatus, a computer readable storage medium and a computer program product, which can be applied to the technical field of financial science and technology.
It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the foregoing figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the disclosure described herein may be capable of operation in sequences other than those illustrated or described herein. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure.
The service information processing method provided by the embodiment of the application can be applied to an application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The data storage system may store data that the server 104 needs to process. The data storage system may be integrated on the server 104 or may be located on a cloud or other network server. The server 104 acquires a historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing characteristic data comprises characteristic data generated in the service processing window service processing process; the server 104 inputs the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model; the server 104 inputs the initial prediction result, the historical service information sequence and the output result of the time sequence trend model to a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point. The server 104 transmits the target prediction result to the terminal 102. The terminal 102 may be, but not limited to, various personal computers, notebook computers, smart phones, tablet computers, internet of things devices, and portable wearable devices, where the internet of things devices may be smart speakers, smart televisions, smart air conditioners, smart vehicle devices, and the like. The portable wearable device may be a smart watch, smart bracelet, headset, or the like. The server 104 may be implemented as a stand-alone server or as a server cluster of multiple servers.
In one embodiment, as shown in fig. 2, a service information processing method is provided, and the method is applied to the server 104 in fig. 1 for illustration, and includes the following steps:
step S202, acquiring a history service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing feature data includes feature data generated during the service processing window processing service.
The service processing characteristic data may be the number taking time, the number calling time, the service type required to be processed, the number of windows corresponding to a certain service currently processed, the service type provided by each service processing window currently, whether the number is exceeded when waiting for processing a certain service, the user attribute, whether the user processes a certain service, and the like of the user when the service processing window processes the service.
The target time point may be one time point when the service processing waiting time needs to be predicted, or may be a plurality of time points when the service processing waiting time needs to be predicted.
The preset time period may be a certain time period before the target time point.
Wherein the historical time point may be a time point before the target time point.
Wherein the service processing window processes the window of the service at the service processing mechanism. For example, the banking business processing window may be a business for handling personal business, public business, cashier's office, honored guest's office, financial management office, etc., where the personal business may be a business for depositing and withdrawing money, remittance, fund, financial management, debit card, etc., and the public business may be a public storage business. In practical application, the service processing window may be an integrated service processing window, where the integrated service processing window may process multiple types of services.
In a specific implementation, a server acquires service processing characteristic data sequences corresponding to a preset number of historical time points in a preset time period before a target time point.
For example, in a banking institute, the server obtains all clients at a certain website at 15 pm: 9 before 00: 00. 10: 00. 11: 00. 12: 00. 13: 00. 14:00, the number of windows for processing the service corresponding to the target service, the type of window service for processing the service, whether the number is exceeded when waiting for processing the service, client attribute information, whether the client processes the service, the number of clients waiting for the service at the time of fetching the number, and average waiting time of the clients for processing the service, wherein the server is as follows: 00. 10: 00. 11: 00. 12: 00. 13: 00. 14: and 00, sorting the corresponding service processing characteristic data according to the time sequence when the target service is handled, and obtaining a historical service information sequence.
Step S204, inputting a historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point; the initial prediction result is determined according to the output results of the time sequence seasonal model, the time sequence trend model and the time sequence holiday model in the first waiting time output model.
The first waiting time output model may refer to a model for determining a waiting time required for a user to wait for processing a service. In practice, the first waiting time output model may be a Prophet time series prediction model (a time series prediction model).
The waiting time may be a waiting time required for the user to wait for the service to be processed.
The initial prediction result may refer to an initial prediction value of the waiting time.
The time sequence seasonal model may refer to a model corresponding to a seasonal term factor in the first waiting time output model. Among them, the seasonal term factor may refer to a factor of a periodic influence of the time series.
The time series trend model may be a model corresponding to a trend term factor in the first waiting time output model. Among other things, trend term factors may refer to factors of the time series that are affected by long-term trends.
The time sequence holiday model may be a model corresponding to a holiday term factor in the first waiting time output model. The holiday term factor may refer to a factor of the time series that is affected by a particular event.
In the specific implementation, the server inputs the historical service information sequence to a pre-trained first waiting time output model to obtain waiting time required when a certain service is handled at a target time point.
For example, in a banking institution, the server will at 9: 00. 10: 00. 11: 00. 12: 00. 13: 00. 14: the corresponding service processing characteristic data are sequenced according to the time sequence when the 00 handles the target service to obtain a historical service information sequence, and the service inputs the historical service information sequence into a pre-trained Prophet time sequence prediction model to obtain a data sequence of 15: 00. 16: 00. 17:00 and 18:00, waiting time required for handling the target service.
Step S206, inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point.
The second waiting time output model may refer to a model for predicting a waiting time required to wait for processing a certain service. In practical applications, the second waiting time output model may be an LSTM time series prediction model (a time series prediction model based on a long-term memory artificial neural network).
In the specific implementation, the server inputs the initial prediction result, the historical service information sequence and the trend item information of the output of the time sequence trend model to a pre-trained second waiting time output model to obtain a waiting time prediction result required by waiting for processing the target service.
In practical application, when the service processing waiting time in the banking service processing mechanism is estimated, the waiting time can be predicted by adopting a pre-trained first waiting time output model and a pre-trained second waiting time output model. The training method of the first waiting time output model and the second waiting time output model comprises the following steps:
step 1: the method comprises the steps of obtaining the business data of the customers of 50 banks for one year, wherein the fields comprise serial numbers, number taking time, number calling time, business types, window numbers, window service types, whether numbers are passed, customer attributes and whether the customers transact the business.
Step 2: and rejecting the business data records lacking the fields. If the field data of the client is missing, all the data of the client are removed.
Step 3: the number of customers waiting for each window is calculated. The average handling time in minutes, e.g. how many customers, types of traffic, are waiting in line within a window in a certain minute.
Step 4: and numbering the business type and the window service type. The bank window is generally divided into personal business, public business, cashier's special cabinet, honored guest room, financial room, personal business including deposit and withdrawal, remittance, fund, financial management, debit card, etc., some sites may have comprehensive business window, the window type can be adjusted according to self business, each window is coded according to the type and number one-to-one correspondence, for example, deposit and withdraw is 1, remittance is 2, fund is 3, financial management is 4, debit card is 5, comprehensive business is 6, etc., specific coding can be adjusted according to window naming of each site.
Step 5: after numerical coding, the reserved characteristics include a service type (numerical type), a window number, a window service type (numerical type), whether the number is exceeded (1 or 0), the number of clients waiting for each service type (number of the number minus the number of the service being processed), average waiting time of the clients for the sub-service, and whether the clients have transacted the service (1 or 0) with the client attributes (attribute 1, attribute 2, … and attribute n).
Step 6: and (3) carrying out normalization processing on each index in the step (1).
Step 7: and (3) performing deduplication on the characteristic attributes of the data after normalization processing in the step (6) by adopting a Relief algorithm (Feature weighting algorithms, a characteristic weight algorithm). And (3) screening out the final required characteristics based on the characteristics reserved in the step (5) by using a Relief algorithm, wherein the characteristics are the service type, the window number, the window service type, the number of waiting clients of each service type, the average waiting time of sub-service clients and the client attribute, and whether the clients transact the service or not.
Specifically, in the process of determining the attribute measurement value of each data characteristic by the Relief algorithm, randomly selecting one sample and one characteristic from a normalized data set (serving as a training set), recording the value of the sample under the first characteristic, finding out the value of the similar sample of the sample under the characteristic, calculating the difference value under the characteristic, and then finding out the value of the heterogeneous sample of the sample under the characteristic. Taking waiting time as an example, randomly selecting one sample and one feature, recording the value of the sample under the first feature, finding out the value of the similar sample of the sample under the feature, calculating the difference value under the feature, and then finding out the value of the heterogeneous sample of the sample under the feature.
Step 8: the data set with the characteristics de-duplicated is sampled by using a cluster-based approach (a sampling method based on a clustering technology), and 2/3 samples are used for model training and 1/3 samples are used for verification.
Among them, the cluster-based approach is capable of picking out the most class samples without the representative disadvantage. In practical application, the raw data is divided into K classes by adopting a clustering algorithm, for example, the K classes are clustered into 168 classes by using a DBSCAN algorithm (a clustering method), then 2/3 samples are used for model training and 1/3 samples are used for verification according to random proportional extraction in each cluster.
Step 9: and (3) inputting the data obtained in the step (8) into a Prophet time sequence prediction model, and inputting the prediction result, trend item information and the data obtained in the step (8) output by the Prophet time sequence prediction model into an LSTM time sequence prediction model.
Wherein, the ACO algorithm (ant colony optimization, ant colony algorithm, a probability type algorithm for searching the optimized path in the graph) is adopted to optimize and select the model parameters.
The Gaussian calculation in the Prophet time sequence prediction model adopts a three-time smoothing method. The method can effectively learn multidimensional features, avoid continuous local optimization of the model, and save calculation time of the model. The Prophet algorithm comprises a season term, a trend term, a holiday term and a residual term, wherein the following formula is as follows:
y(t)＝s(t)+g(t)+h(t)+ε(t)
Wherein, season items are: s (t), mainly describing periodic characteristics such as periodic characteristics of seasons, weeks, months and the like; the trend term is: g (t) represents a trend of change in the time series over a long-term trend; holiday terms are: h (t), holiday items represent conditions in which the trend in holidays is significantly different from that in daily life, but is normal; the remaining items are: ε (t), typically the error value.
In this way, training of the first latency output model and the second latency output model may be achieved. In real-time monitoring, the first waiting time output model and the second waiting time output model can be adopted to predict the waiting time, and the waiting time is used as a numerical basis to determine a business window service adjustment scheme.
The service information processing method, the device, the computer equipment, the storage medium and the computer program product acquire a historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing characteristic data comprises characteristic data generated in the service processing window service processing process; inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model; inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at a target time point; therefore, the initial prediction result output by the first waiting time output model and the output result of the time sequence trend model and the historical service information sequence are input into the second waiting time output model, so that the time characteristics in the historical service information sequence are fully considered, the target prediction result output by the second waiting time output model is more accurate, the service processing waiting time can be accurately determined, and the service processing efficiency is improved.
In another embodiment, inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of a service processing window at a target time point, including: inputting the historical service information sequence into a time sequence season model to obtain a first output result; inputting the historical service information sequence into a time sequence trend model to obtain a second output result; inputting the historical service information sequence into a time sequence holiday model to obtain a third output result; and obtaining an initial prediction result aiming at the waiting time of the service processing window at the target time point according to the first output result, the second output result and the third output result.
The first output result may be a value corresponding to a seasonal item impact factor corresponding to the waiting time at the target time point.
The second output result may be a value corresponding to a trend item influence factor corresponding to the waiting time at the target time point.
The third output result may be a value corresponding to a holiday term influence factor corresponding to the waiting time at the target time point.
In the specific implementation, a server inputs a historical service information sequence to a time sequence seasonal model to obtain a numerical value corresponding to a seasonal item influence factor corresponding to the waiting time at a target time point, the server inputs the historical service information sequence to a time sequence trend model to obtain a numerical value corresponding to a trend item influence factor corresponding to the waiting time at the target time point, the server inputs the historical service information sequence to a time sequence holiday model to obtain a numerical value corresponding to a holiday item influence factor corresponding to the waiting time at the target time point, and the server determines an initial prediction result of the waiting time at the target time point according to the numerical value corresponding to the seasonal item influence factor corresponding to the waiting time at the target time point, the numerical value corresponding to the trend item influence factor corresponding to the waiting time at the target time point and the numerical value corresponding to the holiday item influence factor corresponding to the waiting time at the target time point.
According to the technical scheme, a first output result is obtained by inputting a historical service information sequence into a time sequence season model; inputting the historical service information sequence into a time sequence trend model to obtain a second output result; inputting the historical service information sequence into a time sequence holiday model to obtain a third output result; obtaining an initial prediction result aiming at the waiting time of the service processing window at a target time point according to the first output result, the second output result and the third output result; thus, the periodicity factor, the long-term trend factor and the special event influence factor in the time sequence can be integrated to obtain a more accurate waiting time initial prediction result.
In another embodiment, obtaining an initial prediction result for the waiting time of the service processing window at the target time point according to the first output result, the second output result and the third output result includes: fusing the first output result, the second output result and the third output result to obtain a fused result; and taking the fusion result as an initial prediction result.
The fusion result may be a result obtained by adding a value corresponding to the first output result, a value corresponding to the second output result, and a value corresponding to the third output result.
In specific implementation, the server adds the value corresponding to the first output result, the value corresponding to the second output result and the value corresponding to the third output result to obtain an addition result, and the server takes the addition result as an initial prediction result.
According to the technical scheme, a first output result, a second output result and a third output result are fused to obtain a fusion result; taking the fusion result as an initial prediction result; therefore, the values corresponding to the periodic factors, the long-term trend factors and the special influence factors in the time sequence can be added to obtain the values corresponding to the prediction results, and the waiting time can be accurately predicted.
In another embodiment, the method further comprises: acquiring original historical service information; the original historical service information comprises original service processing characteristic data corresponding to a preset number of historical time points in a preset time period before a target time point; and carrying out numerical coding on the original service processing characteristic data to obtain a historical service information sequence.
The original historical service information may refer to historical service data that is not subjected to the digital encoding process.
In the specific implementation, a server acquires original historical service information to obtain original service processing characteristic data corresponding to a preset number of historical time points in a preset time period before a target time point, and the server performs numerical coding on the original service processing characteristic data to obtain a historical service information sequence.
For example, in a banking institution, the server encodes the deposit and withdrawal service in the service type as a value of 1, the remittance service in the service type as a value of 2, the fund service in the service type as a value of 4, the debit card service in the service type as a value of 5, and the aggregate service as a value of 6; the server encodes the pass number as a value of 1 and encodes the non-pass number as a value of 0; the server encodes a client attribute of 1 into 1 and encodes a client attribute of 2 into 2; the server encodes the codes that have transacted the target business as a value of 1 and encodes the codes that have not transacted the target business as a value of 0. The server expresses the history service information in the order of 'service type-whether number is exceeded-client attribute-whether client transacts target service', wherein 9:00 corresponds to a data item (1, 0, 1) represented at 9: and (00) handling the deposit and withdrawal service by the client, wherein the client does not pass the number, the client attribute is 1, and the client handles the deposit and withdrawal service. The server will 9: 00. 10: 00. 11: 00. 12: 00. 13: 00. 14: the data items corresponding to 00 are sequenced according to time sequence and are used as prediction 15:00 customer waits for the corresponding historical service information sequence when processing the target service.
According to the technical scheme, original historical service information is obtained, and then the original service processing characteristic data is subjected to numerical coding to obtain a historical service information sequence; therefore, the original service processing characteristic data can be converted into the numerical coding data, which is favorable for extracting the characteristics of the data, thereby being favorable for obtaining more accurate waiting time prediction results.
In another embodiment, the step of digitally encoding the original service processing feature data to obtain a historical service information sequence includes: performing numerical coding on the original service processing characteristic data to obtain numerical coding service processing characteristic data; and performing feature de-duplication processing on the feature data of the numerical coding service processing to obtain a historical service information sequence.
The numerically encoded service processing feature data may refer to data obtained by numerically encoding the service processing feature data. For example, after the service type data is encoded, the deposit and withdrawal service is encoded to a value of 1, and the remittance service in the service type is encoded to a value of 2.
In the specific implementation, the server performs the numerical coding processing on the original service processing characteristic data to obtain the numerical coding service processing characteristic data, and the server performs characteristic de-duplication on the numerical coding service processing characteristic data to obtain the historical service information sequence.
In practical application, the server performs the numerical coding feature processing on the original service processing feature data to obtain the numerical coding service processing feature data, where the original service processing feature data may refer to data shown in the following table 1:
TABLE 1
The digitalized encoded service processing characteristic data may refer to data shown in the following table 2:
TABLE 2
According to the technical scheme, the original service processing characteristic data is subjected to numerical coding to obtain numerical coding service processing characteristic data; performing characteristic de-duplication processing on the characteristic data of the numerical coding service processing to obtain a historical service information sequence; therefore, redundant service processing characteristic data can be removed, important service processing characteristic data is reserved, and accurate waiting time prediction results are obtained.
In another embodiment, performing feature deduplication processing on the digitized encoded service processing feature data to obtain a historical service information sequence, including: performing characteristic duplication elimination processing on the numerical coding service processing characteristic data to obtain duplicated numerical coding service processing characteristic data; and carrying out normalization processing on the de-duplicated numerical coding service processing characteristic data to obtain a historical service information sequence.
In the specific implementation, the server performs characteristic duplication removal processing on the digital coding service characteristic data, waits for the duplication removed digital coding service processing characteristic data, and performs normalization processing on the duplication removed digital coding service processing characteristic data to obtain a historical service information sequence.
According to the technical scheme, feature de-duplication processing is carried out on the feature data of the digital coding service processing, so that the feature data of the digital coding service processing after de-duplication is obtained; normalizing the de-duplicated numerical coding service processing characteristic data to obtain a historical service information sequence; therefore, the data can be standardized, so that the characteristic data of each numerical coding service processing is in the same order of magnitude, the influence of singular sample data is eliminated, and more accurate waiting time prediction results are obtained.
In another embodiment, the feature deduplication processing is performed on the digitized encoded service processing feature data to obtain the digitized encoded service processing feature data after deduplication, including: according to the numerical coding service processing characteristic data, determining weight information corresponding to each service processing characteristic; determining at least one target service processing characteristic according to the weight information corresponding to each service processing characteristic; and taking the numerical coding service processing characteristic data corresponding to each target service processing characteristic as the numerical coding service processing characteristic data after de-duplication.
The service processing feature may refer to feature data generated in a service processing process.
The weight information may be information characterizing the extent to which the service processing features affect the latency.
The target service processing feature may refer to information that the weight size is greater than a preset threshold.
In the specific implementation, the server determines weight information corresponding to each service processing feature according to the numerical coding service processing feature data, removes the service processing feature with the weight smaller than a preset threshold value, reserves the service processing feature with the weight larger than the preset threshold value, takes the service processing feature with the weight larger than the preset threshold value as a target service processing feature, and takes the numerical coding service processing feature data corresponding to each target service processing feature as the numerical coding service feature data after the weight is removed.
In practical application, the server processes the feature data according to the numeric coding service in the above table 2, adopts the Relief algorithm to determine the relevance between the feature 1 to the feature 5 and the waiting time in the above table 2, and gives different weights to the feature 1 to the feature 5, wherein the weight of the feature 1 is 0.1, the weight of the feature 2 is 0.2, the weight of the feature 3 is 0.2, the weight of the feature 4 is 0.2, the weight of the feature 5 is 0.3, the server removes the feature with the weight less than 0.2, and the feature with the weight greater than or equal to 0.2 is reserved, so that the server reserves the features 2, 3, 4 and 5, and removes the feature 1, namely the server reserves the data shown in the following table 3:
TABLE 3 Table 3
According to the technical scheme of the embodiment, weight information corresponding to each service processing characteristic is determined according to the numerical coding service processing characteristic data; determining at least one target service processing characteristic according to the weight information corresponding to each service processing characteristic; taking the numerical coding service processing characteristic data corresponding to each target service processing characteristic as the numerical coding service processing characteristic data after de-duplication; therefore, the method can realize the screening of the service processing characteristics according to the weight of different numeric coding service processing characteristic data, brings the effective service processing characteristics into a waiting time prediction process, eliminates the service processing characteristics with smaller influence degree on the waiting time correlation, and is beneficial to accurately predicting the waiting time.
In another embodiment, as shown in fig. 3, a service information processing method is provided, and the method is applied to the server 104 in fig. 1 for illustration, and includes the following steps:
step S302, acquiring a history service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing characteristic data comprises characteristic data generated in the service processing window service processing process;
Step S304, the historical business information sequence is input into a time sequence season model, and a first output result is obtained.
Step S306, the historical business information sequence is input into a time sequence trend model, and a second output result is obtained.
Step S308, the historical service information sequence is input into a time sequence holiday model, and a third output result is obtained.
Step S310, obtaining an initial prediction result of the waiting time of the business processing window at the target time point according to the first output result, the second output result and the third output result.
Step S312, the output results of the initial prediction result, the historical service information sequence and the time sequence trend model are input into a pre-trained second waiting time output model, and a target prediction result corresponding to the waiting time of the service processing window at the target time point is obtained.
It should be noted that, the specific limitation of the above steps may be referred to the specific limitation of a service information processing method.
It should be understood that, although the steps in the flowcharts related to the embodiments described above are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.
Based on the same inventive concept, the embodiment of the application also provides a service information processing device for implementing the service information processing method. The implementation of the solution provided by the device is similar to the implementation described in the above method, so the specific limitation in the embodiments of one or more service information processing devices provided below may refer to the limitation of the service information processing method hereinabove, and will not be repeated herein.
In one embodiment, as shown in fig. 4, there is provided a service information processing apparatus including:
an acquisition module 402, configured to acquire a historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before the target time point; the service processing characteristic data comprises characteristic data generated in the service processing window service processing process;
the output module 404 is configured to input the historical service information sequence to a pre-trained first waiting time output model, and obtain an initial prediction result of waiting time of the service processing window at a target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model;
The prediction module 406 is configured to input the initial prediction result, the historical service information sequence, and the output result of the time sequence trend model to a pre-trained second waiting time output model, so as to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point.
In one embodiment, the output module 404 is specifically configured to input the historical service information sequence to the time sequence season model, so as to obtain a first output result; inputting the historical service information sequence into a time sequence trend model to obtain a second output result; inputting the historical service information sequence into a time sequence holiday model to obtain a third output result; and obtaining an initial prediction result aiming at the waiting time of the service processing window at the target time point according to the first output result, the second output result and the third output result.
In one embodiment, the output module 404 is specifically configured to fuse the first output result, the second output result, and the third output result to obtain a fused result; and taking the fusion result as an initial prediction result.
In one embodiment, the apparatus further comprises: the acquisition module is specifically used for acquiring original historical service information; the original historical service information comprises original service processing characteristic data corresponding to a preset number of historical time points in a preset time period before a target time point; and carrying out numerical coding on the original service processing characteristic data to obtain a historical service information sequence.
In one embodiment, the acquiring module is specifically configured to numerically encode the original service processing feature data to obtain numerically encoded service processing feature data; and performing feature de-duplication processing on the feature data of the numerical coding service processing to obtain a historical service information sequence.
In one embodiment, the obtaining module is specifically configured to perform feature deduplication processing on the feature data of the digitized encoded service processing, so as to obtain the feature data of the digitized encoded service processing after deduplication; and carrying out normalization processing on the de-duplicated numerical coding service processing characteristic data to obtain a historical service information sequence.
In one embodiment, the acquiring module is specifically configured to determine weight information corresponding to each service processing feature according to the digitalized encoded service processing feature data; determining at least one target service processing characteristic according to the weight information corresponding to each service processing characteristic; and taking the numerical coding service processing characteristic data corresponding to each target service processing characteristic as the numerical coding service processing characteristic data after de-duplication.
The respective modules in the above-described service information processing apparatus may be implemented in whole or in part by software, hardware, and a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured 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 an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used for storing business information processing data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by a processor, implements a business information processing method.
It will be appreciated by those skilled in the art that the structure shown in fig. 5 is merely a block diagram of some of the structures associated with the present application and is not limiting of the computer device to which the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided that includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of a business information processing method as described above. The steps of a service information processing method herein may be the steps of a service information processing method of the above-described respective embodiments.
In one embodiment, a computer readable storage medium is provided, storing a computer program which, when executed by a processor, causes the processor to perform the steps of a business information processing method as described above. The steps of a service information processing method herein may be the steps of a service information processing method of the above-described respective embodiments.
In one embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, causes the processor to perform the steps of a business information processing method as described above. The steps of a service information processing method herein may be the steps of a service information processing method of the above-described respective embodiments.
It should be noted that, user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, presented data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the various embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magnetic random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (Phase Change Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like. The databases referred to in the various embodiments provided herein may include at least one of relational databases and non-relational databases. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic units, quantum computing-based data processing logic units, etc., without being limited thereto.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples only represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the present application. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application shall be subject to the appended claims.
Claims (11)
1. A business information processing method, characterized in that the method comprises:
acquiring a historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before a target time point; the service processing characteristic data comprises characteristic data generated in the service processing process of the service processing window;
Inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of waiting time of the service processing window at the target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model;
and inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point.
2. The method of claim 1, wherein inputting the historical traffic information sequence into a pre-trained first latency output model to obtain an initial prediction of latency for the traffic processing window at the target point in time comprises:
inputting the historical service information sequence into the time sequence season model to obtain a first output result;
inputting the historical service information sequence into the time sequence trend model to obtain a second output result;
Inputting the historical service information sequence into the time sequence holiday model to obtain a third output result;
and obtaining an initial prediction result of the waiting time of the business processing window at the target time point according to the first output result, the second output result and the third output result.
3. The method of claim 2, wherein obtaining an initial prediction result for the latency of the traffic processing window at the target point in time based on the first output result, the second output result, and the third output result comprises:
fusing the first output result, the second output result and the third output result to obtain a fused result;
and taking the fusion result as the initial prediction result.
4. The method according to claim 1, wherein the method further comprises:
acquiring original historical service information; the original historical service information comprises original service processing characteristic data corresponding to the preset number of historical time points in the preset time period before the target time point;
and carrying out numerical coding on the original service processing characteristic data to obtain the historical service information sequence.
5. The method of claim 4, wherein said digitally encoding said raw business process feature data to obtain said historical business information sequence comprises:
performing numerical coding on the original service processing characteristic data to obtain numerical coding service processing characteristic data;
and performing characteristic de-duplication processing on the numerical coding service processing characteristic data to obtain the historical service information sequence.
6. The method of claim 5, wherein performing feature deduplication processing on the digitally encoded service processing feature data to obtain the historical service information sequence comprises:
performing characteristic de-duplication processing on the numerical coding service processing characteristic data to obtain de-duplicated numerical coding service processing characteristic data;
and carrying out normalization processing on the de-duplicated numerical coding service processing characteristic data to obtain the historical service information sequence.
7. The method of claim 6, wherein performing feature deduplication on the digitized encoded service processing feature data to obtain deduplicated digitized encoded service processing feature data comprises:
Determining weight information corresponding to each service processing characteristic according to the numerical coding service processing characteristic data;
determining at least one target service processing characteristic according to the weight information corresponding to each service processing characteristic;
and taking the numerical coding service processing characteristic data corresponding to each target service processing characteristic as the numerical coding service processing characteristic data after the duplication removal.
8. A service information processing apparatus, characterized in that the apparatus comprises:
the acquisition module is used for acquiring the historical service information sequence; the historical service information sequence comprises service processing characteristic data corresponding to a preset number of historical time points in a preset time period before a target time point; the service processing characteristic data comprises characteristic data generated in the service processing process of the service processing window;
the output module is used for inputting the historical service information sequence into a pre-trained first waiting time output model to obtain an initial prediction result of the waiting time of the service processing window at the target time point; the initial prediction result is obtained by determining the output results of a time sequence seasonal model, a time sequence trend model and a time sequence holiday model in the first waiting time output model;
And the prediction module is used for inputting the initial prediction result, the historical service information sequence and the output result of the time sequence trend model into a pre-trained second waiting time output model to obtain a target prediction result corresponding to the waiting time of the service processing window at the target time point.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 7 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.
11. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310316723.1A CN116307227A (en) | 2023-03-29 | 2023-03-29 | Service information processing method, device and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310316723.1A CN116307227A (en) | 2023-03-29 | 2023-03-29 | Service information processing method, device and computer equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116307227A true CN116307227A (en) | 2023-06-23 |
Family
ID=86792467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310316723.1A Pending CN116307227A (en) | 2023-03-29 | 2023-03-29 | Service information processing method, device and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116307227A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116542418A (en) * | 2023-07-06 | 2023-08-04 | 武汉星际互动智能技术有限公司 | Deep learning-based business handling method and system for office hall |
| CN117350526A (en) * | 2023-10-07 | 2024-01-05 | 交通银行股份有限公司广东省分行 | Intelligent sorting method, device, equipment and storage medium for self-service items |
-
2023
- 2023-03-29 CN CN202310316723.1A patent/CN116307227A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116542418A (en) * | 2023-07-06 | 2023-08-04 | 武汉星际互动智能技术有限公司 | Deep learning-based business handling method and system for office hall |
| CN116542418B (en) * | 2023-07-06 | 2023-09-15 | 武汉星际互动智能技术有限公司 | Deep learning-based business handling method and system for office hall |
| CN117350526A (en) * | 2023-10-07 | 2024-01-05 | 交通银行股份有限公司广东省分行 | Intelligent sorting method, device, equipment and storage medium for self-service items |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116307227A (en) | Service information processing method, device and computer equipment | |
| EP3602317A1 (en) | Time-based ensemble machine learning model | |
| Wang et al. | A DT-SVM strategy for stock futures prediction with big data | |
| RU2723448C1 (en) | Method of calculating client credit rating | |
| CN114693409A (en) | Product matching method, device, computer equipment, storage medium and program product | |
| CN116738009B (en) | Method for archiving and backtracking data | |
| CN116483822B (en) | Service data early warning method, device, computer equipment and storage medium | |
| CN116800831B (en) | Service data pushing method, device, storage medium and processor | |
| CN114495137B (en) | Bill abnormity detection model generation method and bill abnormity detection method | |
| CN115204984A (en) | Business product pushing method and device, computer equipment and storage medium | |
| CN114662570A (en) | Business risk prediction method and device, computer equipment and storage medium | |
| CN114881761A (en) | Determination method of similar sample and determination method of credit limit | |
| CN115186890A (en) | Early warning method, early warning device, computer equipment and storage medium | |
| CN114722941A (en) | Credit default identification method, apparatus, device and medium | |
| CN114648406A (en) | User credit integral prediction method and device based on random forest | |
| CN114170000A (en) | Credit card user risk category identification method, device, computer equipment and medium | |
| Bhatnagar | Data mining-based big data analytics: parameters and layered framework | |
| CN111581068A (en) | Terminal workload calculation method and device, storage medium, terminal and cloud service system | |
| CN113627653B (en) | Method and device for determining activity prediction strategy of mobile banking user | |
| CN117745423B (en) | Abnormal account identification method | |
| Zhang et al. | The constrained GAN with hybrid encoding in predicting financial behavior | |
| CN115587856A (en) | Customer classification model determination method, customer classification model determination device, computer equipment and storage medium | |
| CN117113131A (en) | Classification model determining method, device, computer equipment and storage medium | |
| Purushu et al. | Predictive Analysis of Transaction Fraud leveraging Big Data for Deep Learning | |
| CN117459576A (en) | Data pushing method and device based on edge calculation and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |