CN112561108A - Data processing method, device, equipment and medium - Google Patents
Data processing method, device, equipment and medium Download PDFInfo
- Publication number
- CN112561108A CN112561108A CN202011547888.2A CN202011547888A CN112561108A CN 112561108 A CN112561108 A CN 112561108A CN 202011547888 A CN202011547888 A CN 202011547888A CN 112561108 A CN112561108 A CN 112561108A
- Authority
- CN
- China
- Prior art keywords
- features
- training
- sub
- model
- driver
- 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
- 238000003672 processing method Methods 0.000 title claims description 5
- 230000004927 fusion Effects 0.000 claims abstract description 36
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 21
- 238000012549 training Methods 0.000 claims description 76
- 230000006399 behavior Effects 0.000 claims description 62
- 238000004590 computer program Methods 0.000 claims description 11
- 238000010586 diagram Methods 0.000 description 24
- 230000006870 function Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 238000007500 overflow downdraw method Methods 0.000 description 2
- 230000001902 propagating effect Effects 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 238000013481 data capture Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
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
- G06Q10/02—Reservations, e.g. for tickets, services or events
- G06Q10/025—Coordination of plural reservations, e.g. plural trip segments, transportation combined with accommodation
-
- 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
- G06N3/045—Combinations of networks
-
- 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
-
- 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"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06311—Scheduling, planning or task assignment for a person or group
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- General Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Primary Health Care (AREA)
- Traffic Control Systems (AREA)
Abstract
Embodiments of the present disclosure relate to methods, apparatuses, devices, and media for data processing. The method comprises the following steps: obtaining a plurality of characteristics associated with a detour behavior of a target driver; and applying the plurality of features to a fusion model to generate a final result indicating whether the target driver is responsible for the detour behavior, the fusion model including a plurality of sub-models corresponding to the plurality of features and a feature intersection model, the plurality of sub-models respectively generating a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features, the feature intersection model generating the final result based on the plurality of intermediate results. In this way, the accuracy of determining whether the driver is responsible for the detour behaviour can be improved.
Description
Technical Field
Embodiments of the present disclosure relate generally to data processing, and more particularly, relate to a data processing method, apparatus, electronic device, and computer storage medium.
Background
With the development of information technology, a travel mode using a network is more and more popular. This type of travel makes it difficult to avoid various disputes between the passengers and the drivers while facilitating the travel of the passengers. Detour complaints are a common dispute. A detour complaint is one in which the actual route traveled by the vehicle creates an unnecessary trip compared to the planned route, resulting in a cost or time that exceeds the expectations of the passenger, thereby causing the passenger to initiate the complaint. In such a case, there is a need to blame for the detour complaint, i.e., to determine whether the driver is responsible for the detour behavior. Therefore, whether punishment is carried out on the driver or not can be determined so as to maintain good travel ecological environment. However, conventional disclaimers for detour complaints are inefficient and less accurate.
Disclosure of Invention
According to an embodiment of the present disclosure, a data processing scheme is provided.
In a first aspect of the disclosure, a data processing method is provided. The method comprises the following steps: obtaining a plurality of characteristics associated with a detour behavior of a target driver; and applying the plurality of features to a fusion model to generate a final result indicating whether the target driver is responsible for the detour behavior, the fusion model including a plurality of sub-models corresponding to the plurality of features and a feature intersection model, the plurality of sub-models respectively generating a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features, the feature intersection model generating the final result based on the plurality of intermediate results.
In a second aspect of the disclosure, an apparatus for data processing is provided. The device includes: an acquisition module configured to acquire a plurality of characteristics associated with a detour behavior of a target driver; and a generation module configured to apply the plurality of features to a fusion model to generate a final result indicating whether the target driver is responsible for the detour behavior, the fusion model including a plurality of sub-models corresponding to the plurality of features and a feature intersection model, the plurality of sub-models respectively generating a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features, the feature intersection model generating the final result based on the plurality of intermediate results.
In a third aspect of the disclosure, an electronic device is provided. The electronic device includes: one or more processors; and memory for storing one or more programs that, when executed by the one or more processors, cause the electronic device to implement the method according to the first aspect of the disclosure.
In a fourth aspect of the present disclosure, a computer-readable medium is provided, on which a computer program is stored which, when executed by a processor, implements a method according to the first aspect of the present disclosure.
It should be understood that the statements herein reciting aspects are not intended to limit the critical or essential features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.
Drawings
The above and other features, advantages and aspects of various embodiments of the present disclosure will become more apparent by referring to the following detailed description when taken in conjunction with the accompanying drawings. In the drawings, like or similar reference characters designate like or similar elements, and wherein:
FIG. 1 illustrates a schematic diagram of an exemplary environment in which embodiments of the present disclosure can be implemented;
FIG. 2 illustrates a flow diagram of a method for data processing according to some embodiments of the present disclosure;
FIG. 3 illustrates a schematic diagram of an example of a fusion model, in accordance with some embodiments of the present disclosure;
FIG. 4 illustrates a flow diagram of a method for training a sub-model in accordance with some embodiments of the present disclosure;
FIG. 5 illustrates a flow diagram of a method for joint training in accordance with some embodiments of the present disclosure;
FIG. 6 illustrates a block diagram of an apparatus for data processing, in accordance with some embodiments of the present disclosure; and
FIG. 7 illustrates a block diagram of an electronic device capable of implementing embodiments of the present disclosure.
Detailed Description
Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but rather are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the disclosure are for illustration purposes only and are not intended to limit the scope of the disclosure.
In describing embodiments of the present disclosure, the terms "include" and its derivatives should be interpreted as being inclusive, i.e., "including but not limited to. The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first," "second," and the like may refer to different or the same object. Other explicit and implicit definitions are also possible below.
In determining whether a driver is responsible for a detour, various features may need to be considered to well characterize the scene in the trip where the detour occurred. However, many scenarios cannot be restored, subject to the feature compatibility requirements of the traditional model. For example, passenger fingering is an important reason for detour. However, conventionally, the characteristics characterizing passenger directions have not been considered. Even if the recording in the journey depicting the passenger's direction is converted into a text feature, the text feature cannot be directly applied to the traditional model for accountability. As another example, passengers adding waypoints or modifying destinations are also important reasons for detours. The trajectory feature may characterize whether the passenger adds waypoints or modifies the destination. However, the trajectory feature cannot be directly applied to the conventional model for accountability.
In the case where the features considered are insufficient, or the model structure is too simple, the model may be under-fitted and less accurate. Because the conventional model has low accountability accuracy, a large amount of human resources are required to be invested for accountability, so that the accountability efficiency is reduced, and the accountability cost is increased.
Conventionally, a result fusion method and an embedding (embedding) fusion method can be used to solve the above-mentioned problem that various features cannot be considered in combination. In a result fusion mode, different features are used for respectively training a plurality of sub-models to obtain model scores, and a responsibility judgment result is obtained by using a stacking method. The result fusion mode has no strict limitation on each submodel, so that the submodel can use a depth model and a tree model. However, since the score is used to represent the entire feature, a large amount of feature information is lost, which is likely to cause under-fitting.
In the embedding fusion mode, different features are used for respectively training a plurality of depth sub-models, and a hidden layer is respectively extracted from the plurality of sub-models and spliced to serve as a new feature so as to train a new model to judge responsibility. However, since the new model depends on the sub-model, joint training of the sub-model and the new model cannot be performed, resulting in a loss of a certain amount of feature information.
To this end, embodiments of the present disclosure provide a scheme for data processing. In the scheme, a plurality of characteristics related to the detour behavior of a target driver are acquired; and applying the plurality of features to a fusion model to generate a final result indicating whether the target driver is responsible for the detour behavior, the fusion model including a plurality of sub-models corresponding to the plurality of features and a feature intersection model, the plurality of sub-models respectively generating a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features, the feature intersection model generating the final result based on the plurality of intermediate results.
In this way, in the scheme, whether the driver is responsible for the detour behavior can be simply, quickly and accurately judged through the fusion model. Embodiments of the present disclosure will be described below in detail with reference to the accompanying drawings.
FIG. 1 illustrates a schematic diagram of an exemplary environment 100 in which embodiments of the present disclosure can be implemented. The environment 100 includes a computing device 110. Computing device 110 may contain at least a processor, memory, and other components typically found in a general purpose computer to implement computing, storage, communication, control, and the like functions. For example, the computing device 110 may be a smartphone, tablet computer, personal computer, desktop computer, notebook computer, server, mainframe, distributed computing system, and the like.
In the environment 100, the computing device 110 is configured to obtain a plurality of features associated with the detour behavior of the target driver, such as features 130-1 to 130-N (hereinafter collectively referred to as "features 130", where N is an integer greater than 1). In certain embodiments, the target driver's driving vehicle 120 may be equipped with various data capture devices, such as cameras, voice recorders, tachographs, navigation devices, and the like. In this case, the computing device 110 may obtain these features 130 from the vehicle 120. In some embodiments, the target driver may use a terminal device, such as a smart phone, in driving the vehicle 120. Such a terminal device may also be provided with various data capturing means. In this case, the computing device 110 may also obtain these features 130 from the terminal device.
The computing device 110 then applies the plurality of features 130 to the fusion model 140 to generate a final result 150 indicating whether the target driver is responsible for the detour behavior. The fusion model 140 may be deployed within the computing device 110 or external to the computing device 110, as the invention is not limited in this respect. The operation of the computing device 110 will be described in detail below in conjunction with fig. 2-5.
Fig. 2 illustrates a flow diagram of a method 200 for data processing according to some embodiments of the present disclosure. The method 200 may be implemented by the computing device 110 as shown in FIG. 1. Alternatively, method 200 may be implemented by subjects other than computing device 110. It should be understood that method 200 may also include additional steps not shown and/or may omit steps shown, as the scope of the present disclosure is not limited in this respect.
At 210, the computing device 110 obtains a plurality of features 130 associated with the detour behavior of the target driver. These features 130 may have different modalities in order to restore as much as possible the scene in the journey where the detour behaviour occurs. For example, these features 130 may include quantity features, category features, text features, track features, video features, and/or sequence features, among others.
As an example, the quantity characteristics may include actual and estimated mileage, time, and cost of a network appointment order, among others. In addition, the quantitative feature may also include statistical information of the driver and passengers, such as the number of complaints that the driver has complained about over the past three months. Additionally, the quantitative feature may also include map information, such as the number of drifts occurring in the trip, and the like.
The category characteristics may include the time of occurrence of the complaint, the type of navigation used by the driver, and the travel congestion index, among others. The text features may include text converted from a recording in the trip, text reported by the driver after the end of the trip, passenger complaint text, and negotiation text of the driver and passenger, among others. The trajectory characteristics may include the points of the trajectory actually traveled by the driver, which may be recorded at predetermined time intervals (e.g., 3 seconds) and represented by a timestamp and latitude and longitude. The video features may include a sequence of numbers of people in the vehicle recorded at predetermined time intervals (e.g., 1 minute) provided by the video in the vehicle. Sequence characteristics may include yaw sequences, abnormal dwell sequences, and the like.
At 220, the computing device 110 may apply the plurality of features 130 to the fusion model 140 to generate a final result 150 indicating whether the target driver is responsible for the detour behavior. Fig. 3 illustrates a schematic diagram 300 of an example of a fusion model 140, according to some embodiments of the present disclosure.
The fusion model 140 includes a plurality of sub-models corresponding to the plurality of features 130, for example, sub-models 310-1 to 310-N (hereinafter, collectively referred to as "sub-models 310") corresponding to the features 130-1 to 130-N, respectively. The plurality of sub-models 310 generate a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features 130, respectively.
For example, the sub-model 310 may be a GRU (Gated current Unit) model, and the trajectory features may be applied to the sub-model 310 to determine whether an abnormal condition such as a road block is encountered during the journey, so as to determine responsibility for the target driver. As another example, the sub-model 310 may be a HAN (Hierarchical Attention Network) model, and textual features translated from the audio recordings in the trip may be applied to the sub-model 310 to determine whether the passenger is directed to the road to blame the target driver. As yet another example, the sub-model may be a TextCNN (Text Convolutional Neural Network) model, and textual features generated after the end of the trip may be applied to the sub-model 310 to determine information expressed by the target driver and the passenger after the end of the trip to disclaim the target driver. As still another example, the sub-model 310 may also be a Wide & Deep (generalized and depth) model corresponding to a number feature, a GRU model corresponding to a video feature, a GRU model corresponding to a sequence feature, and the like.
In some embodiments, the sub-model 310 may include multiple layers. For example, sub-model 310-1 may include layers 320-1 through 320-J (hereinafter, collectively referred to as "layers 320", where J is an integer greater than 1), and sub-model 310-N may include layers 340-1 through 340-K (hereinafter, collectively referred to as "layers 340", where K is an integer greater than 1). Of these layers, the last hidden layer may indicate whether the target driver is responsible for the detour behavior and therefore may be an intermediate result. For example, the last hidden layer 320-J of the sub-model 310-1 and the last hidden layer 340-K of the sub-model 310-N may each indicate a probability of whether the target driver is responsible for the detour behavior, and thus may be intermediate results.
In addition to the sub-models 310, the fusion model 140 also includes a feature intersection model 370. The feature intersection model 370 may be any suitable DNN (Deep Neural Network) model, such as Deep fm (Deep Factorization Machine) model, Deep & Cross (Deep intersection) model, XDeepFM (eXtreme depth Factorization Machine) model, and the like.
The feature intersection model 370 may generate a final result 150 indicating whether the target driver is responsible for the detour behavior based on a plurality of intermediate results (e.g., hidden layers 320-J and 340-K) generated by each of the plurality of sub-models 310. In some embodiments, the computing device 110 may merge multiple intermediate results to generate a merged result 360. For example, the computing device 110 may join the last hidden layers (e.g., hidden layers 320-J and 340-K) of the plurality of sub-models 310 to generate the merged result 360.
In this way, a plurality of features in different modalities can be used for restoring the scene in the journey where the detour behavior occurs, so that the loss of feature information is reduced, and the accuracy of the final result is obviously improved.
Operations performed in the use phase of the fusion model 140 are described above, and operations performed in the training phase of the fusion model 140 will be described below with reference to fig. 4 and 5. Specifically, the fusion model 140 may be trained in advance before the fusion model 140 is used to justify the target driver. The training of the fusion model 140 can be divided into two phases: a sub-model training phase and a joint training phase.
FIG. 4 illustrates a flow diagram of a method 400 for training a sub-model 310 according to some embodiments of the present disclosure. In training the submodel 310, the computing device 110 may obtain, at 410, training features corresponding to the submodel 310 for respective ones of the plurality of submodels 310. The training features are associated with the driver's historical detour behavior. Similar to the features 130, the training features may include quantity features, category features, text features, trajectory features, video features, or sequence features, among others. In addition, the computing device 110 may also obtain realistic results associated with the training features. For example, the training features may be labeled with real results. The real results may indicate whether the driver is responsible for historical detour behavior.
At 420, the computing device 110 may generate a prediction indicating whether the driver is responsible for the historical detour behavior by applying the training features to the sub-models 310. For example, assuming that the submodel 310 is a HAN model, the textual features 130 converted from the recording on the trip may be applied to the submodel 310 to determine whether the passenger is directing the route, and thus whether the driver is responsible for historical detour behavior.
At 430, the computing device 110 may train the sub-model 310 to minimize a difference between the predicted outcome and a true outcome indicating whether the driver is responsible for the historical detour behavior. For example, the computing device 110 may iteratively train the sub-model 310 to make the accountability results of the sub-model 310 to the driver as close as possible to the true results.
In this way, multiple features in different modalities may be utilized to train the corresponding sub-models, respectively. The parameters of the trained sub-models may be used as initialization parameters for the fusion model 140. Thus, the training of the submodel 310 may act as a pre-training of the fusion model 140, thereby speeding up the convergence of the fusion model 140.
The fusion model 140 may then be jointly trained, that is, the sub-models 310 and the feature intersection models 370. As described above, the feature intersection model 370 may be any suitable DNN model, such as a Deep fm model, Deep & Cross model, XDeepFM model, and the like. Since the feature intersection model 370 fuses the intermediate results from multiple sub-models, the fit capability of the fused model 140 may be improved. In addition, since the joint training trains both the sub-model 310 and the feature intersection model 370, the parameters of the sub-model 310 may be further updated. Fig. 5 illustrates a flow diagram of a method 500 for joint training, in accordance with some embodiments of the present disclosure.
At 510, the computing device 110 may obtain a plurality of training features. A plurality of training features are associated with the driver's historical detour behavior. Similar to the features 130, the training features may include quantity features, category features, text features, trajectory features, video features, or sequence features, among others. In addition, the computing device 110 may also obtain realistic results associated with the training features. For example, the training features may be labeled with real results. The real results may indicate whether the driver is responsible for historical detour behavior. In some embodiments, the training features used in the joint training phase may be different from the training features used in the sub-model training phase, and thus various features may be leveraged to better train the fusion model 140. Alternatively, the training features used in the joint training phase may also be the same as the training features used in the sub-model training phase, as the invention is not limited thereto.
At 520, the computing device 110 may generate a plurality of intermediate predictions indicating whether the driver is responsible for the historical detour behavior by applying each of the plurality of training features to a corresponding one of the plurality of sub-models 310. As described above, for example, the sub-model 310 may be a GRU model, and trajectory features may be applied to the sub-model 310 to determine whether an abnormal condition such as a shut-off is encountered during travel to justify the driver. As another example, the sub-model 310 may be a HAN model, and textual features converted from the recording on the trip may be applied to the sub-model 310 to determine whether the passenger is directing his/her way to blame the driver. In some embodiments, the sub-model 310 may include multiple layers. Among these layers, the last hidden layer may indicate whether the driver is responsible for historical detour behavior and therefore may serve as an intermediate prediction.
At 530, the computing device 110 may merge the plurality of intermediate predicted results to generate a merged result. For example, the computing device 110 may join the last hidden layer of the plurality of submodels 310 to generate a merged result.
At 540, the computing device 110 may generate a final predicted result indicating whether the driver is responsible for the historical detour behavior by applying the merged result to the feature intersection model 370. At 550, the computing device 110 may jointly train the plurality of sub-models 310 and the feature intersection model 370 to minimize a difference between the final predicted result and a true result indicating whether the driver is responsible for the historical detour behavior. For example, the computing device 110 may iteratively conduct joint training to bring the outcome of disclaimer performed by the fusion model 140 on the driver as close as possible to the true outcome.
In this way, the scheme can make full use of various features in different modes to reduce the loss of feature information caused in the model training process. In addition, the scheme can improve the model performance, realize the end-to-end model and reduce the complexity of model deployment.
Fig. 6 illustrates a block diagram of an apparatus 600 for data processing according to some embodiments of the present disclosure. For example, the apparatus 600 may be disposed in a computing device 110. As shown in fig. 6, the apparatus 600 includes an acquisition module 610 configured to acquire a plurality of characteristics associated with a detour behavior of a target driver; and a generating module 620 configured to apply the plurality of features to a fusion model to generate a final result indicating whether the target driver is responsible for the detour behavior, the fusion model including a plurality of sub-models corresponding to the plurality of features, the plurality of sub-models respectively generating a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features, and a feature intersection model generating the final result based on the plurality of intermediate results.
In some embodiments, the apparatus 600 further comprises a first training module configured to train a plurality of sub-models.
In some embodiments, the first training module comprises: for a respective submodel of the plurality of submodels, a first training feature acquisition module configured to acquire training features corresponding to the submodel, the training features associated with historical detour behavior of the driver; a prediction result generation module configured to generate a prediction result indicating whether the driver is responsible for the historical detour behavior by applying the training features to the sub-models; and a submodel training module configured to train the submodel to minimize a difference between the predicted outcome and a real outcome indicating whether the driver is responsible for the historical detour behavior.
In some embodiments, the apparatus 600 further comprises a second training module configured to jointly train the plurality of submodels and the feature intersection model.
In some embodiments, the second training module comprises: a second training feature acquisition module configured to acquire a plurality of training features, the plurality of training features being associated with historical detour behavior of the driver; an intermediate prediction result generation module configured to generate a plurality of intermediate prediction results indicating whether the driver is responsible for the historical detour behavior by applying each of the plurality of training features to a corresponding one of the plurality of sub-models; a merging module configured to merge the plurality of intermediate prediction results to generate a merged result; a final prediction result generation module configured to generate a final prediction result indicating whether the driver is responsible for the historical detour behavior by applying the merged result to the feature intersection model; and a joint training module configured to jointly train the plurality of sub-models and the feature intersection model to minimize a difference between the final predicted result and a true result indicating whether the driver is responsible for the historical detour behavior.
In some embodiments, the plurality of features includes at least two of: a quantity feature, a category feature, a text feature, a track feature, a video feature, and a sequence feature.
FIG. 7 shows a schematic block diagram of an electronic device 700 that may be used to implement embodiments of the present disclosure. Device 700 may be used to implement apparatus 600 of fig. 6. As shown, device 700 includes a Central Processing Unit (CPU)701 that may perform various appropriate actions and processes in accordance with computer program instructions stored in a Read Only Memory (ROM)702 or computer program instructions loaded from a storage unit 708 into a Random Access Memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The CPU 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.
Various components in the device 700 are connected to the I/O interface 705, including: an input unit 706 such as a keyboard, a mouse, or the like; an output unit 707 such as various types of displays, speakers, and the like; a storage unit 708 such as a magnetic disk, optical disk, or the like; and a communication unit 709 such as a network card, modem, wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
Various processes and processes described above, such as methods 200, 400, and/or 500, may be performed by processing unit 701. For example, in some embodiments, methods 200, 400, and/or 500 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as storage unit 708. In some embodiments, part or all of a computer program may be loaded onto and/or installed onto device 700 via ROM 702 and/or communications unit 709. When the computer program is loaded into RAM 703 and executed by CPU 701, one or more steps of methods 200 and/or 500 described above may be performed. Alternatively, in other embodiments, CPU 701 may be configured to perform methods 200, 400, and/or 500 in any other suitable manner (e.g., by way of firmware).
The present disclosure may be methods, apparatus, systems, and/or computer program products. The computer program product may include a computer-readable storage medium having computer-readable program instructions embodied thereon for carrying out various aspects of the present disclosure.
The computer readable storage medium may be a tangible device that can hold and store the instructions for use by the instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic memory device, a magnetic memory device, an optical memory device, an electromagnetic memory device, a semiconductor memory device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a Static Random Access Memory (SRAM), a portable compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD), a memory stick, a floppy disk, a mechanical coding device, such as punch cards or in-groove projection structures having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media as used herein is not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through a fiber optic cable), or electrical signals transmitted through electrical wires.
The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or to an external computer or external storage device via a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. The network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in the respective computing/processing device.
The computer program instructions for carrying out operations of the present disclosure may be assembler instructions, Instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider). In some embodiments, the electronic circuitry that can execute the computer-readable program instructions implements aspects of the present disclosure by utilizing the state information of the computer-readable program instructions to personalize the electronic circuitry, such as a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA).
Various aspects of the present disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-readable program instructions.
These computer-readable program instructions may be provided to a processing unit of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processing unit of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer-readable program instructions may also be stored in a computer-readable storage medium that can direct a computer, programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.
The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer, other programmable apparatus or other devices implement the functions/acts specified in the flowchart and/or block diagram block or blocks.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The foregoing description of the embodiments of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (14)
1. A data processing method, characterized by:
obtaining a plurality of characteristics associated with a detour behavior of a target driver; and
applying the plurality of features to a fusion model to generate a final result indicating whether the target driver is responsible for the detour behavior, the fusion model including a plurality of sub-models corresponding to the plurality of features and a feature intersection model, the plurality of sub-models respectively generating a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features, the feature intersection model generating the final result based on the plurality of intermediate results.
2. The method of claim 1, further comprising:
training the plurality of sub-models.
3. The method of claim 2, wherein training the plurality of sub-models comprises:
for a respective submodel of the plurality of submodels,
acquiring training characteristics corresponding to the submodels, wherein the training characteristics are associated with historical detour behaviors of drivers;
generating a prediction indicating whether the driver is responsible for the historical detour behavior by applying the training features to the sub-model; and
training the sub-models to minimize a difference between the predicted outcome and a true outcome indicating whether the driver is responsible for the historical detour behavior.
4. The method of claim 1, further comprising:
jointly training the plurality of sub-models and the feature intersection model.
5. The method of claim 4, wherein jointly training the plurality of sub-models and the feature intersection model comprises:
obtaining a plurality of training features, the plurality of training features associated with a driver's historical detour behavior;
generating a plurality of intermediate predictions indicating whether the driver is responsible for the historical detour behavior by applying each of the plurality of training features to a corresponding one of the plurality of sub-models;
merging the intermediate prediction results to generate a merged result;
generating a final predicted result indicating whether the driver is responsible for the historical detour behavior by applying the merged result to the feature intersection model; and
jointly training the plurality of sub-models and the feature intersection model to minimize a difference between the final predicted outcome and a true outcome indicating whether the driver is responsible for the historical detour behavior.
6. The method of claim 1, wherein the plurality of features comprises at least two of:
the characteristics of the quantity are such that,
the characteristics of the categories are used to determine,
the characteristics of the text are such that,
the characteristics of the track are that,
video features, and
sequence characteristics.
7. An apparatus for data processing, comprising:
an acquisition module configured to acquire a plurality of characteristics associated with a detour behavior of a target driver; and
a generation module configured to apply the plurality of features to a fusion model to generate a final result indicating whether the target driver is responsible for the detour behavior, the fusion model including a plurality of sub-models corresponding to the plurality of features and a feature intersection model, the plurality of sub-models respectively generating a plurality of intermediate results indicating whether the target driver is responsible for the detour behavior based on the plurality of features, the feature intersection model generating the final result based on the plurality of intermediate results.
8. The apparatus of claim 7, further comprising:
a first training module configured to train the plurality of sub-models.
9. The apparatus of claim 8, wherein the first training module comprises:
for a respective submodel of the plurality of submodels,
a first training feature acquisition module configured to acquire training features corresponding to the submodels, the training features being associated with historical detour behavior of the driver;
a prediction result generation module configured to generate a prediction result indicating whether the driver is responsible for the historical detour behavior by applying the training features to the sub-model; and
a sub-model training module configured to train the sub-model to minimize a difference between the predicted outcome and a real outcome indicating whether the driver is responsible for the historical detour behavior.
10. The apparatus of claim 7, further comprising:
a second training module configured to jointly train the plurality of sub-models and the feature intersection model.
11. The apparatus of claim 10, wherein the second training module comprises:
a second training feature acquisition module configured to acquire a plurality of training features associated with a driver's historical detour behavior;
an interim prediction results generation module configured to generate a plurality of interim prediction results indicating whether the driver is responsible for the historical detour behavior by applying each of the plurality of training features to a corresponding one of the plurality of sub-models;
a merging module configured to merge the plurality of intermediate prediction results to generate a merged result;
a final prediction result generation module configured to generate a final prediction result indicating whether the driver is responsible for the historical detour behavior by applying the merged result to the feature intersection model; and
a joint training module configured to jointly train the plurality of sub-models and the feature intersection model to minimize a difference between the final predicted outcome and a true outcome indicating whether the driver is responsible for the historical detour behavior.
12. The apparatus of claim 7, wherein the plurality of features comprises at least two of:
the characteristics of the quantity are such that,
the characteristics of the categories are used to determine,
the characteristics of the text are such that,
the characteristics of the track are that,
video features, and
sequence characteristics.
13. An electronic device, the electronic device comprising:
one or more processors; and
memory storing one or more programs that, when executed by the one or more processors, cause the electronic device to implement the method of any of claims 1-6.
14. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011547888.2A CN112561108A (en) | 2020-12-24 | 2020-12-24 | Data processing method, device, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011547888.2A CN112561108A (en) | 2020-12-24 | 2020-12-24 | Data processing method, device, equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112561108A true CN112561108A (en) | 2021-03-26 |
Family
ID=75032544
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011547888.2A Pending CN112561108A (en) | 2020-12-24 | 2020-12-24 | Data processing method, device, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112561108A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018205561A1 (en) * | 2017-05-09 | 2018-11-15 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for processing an abnormal order |
| CN109035768A (en) * | 2018-07-25 | 2018-12-18 | 北京交通大学 | A kind of taxi detours the recognition methods of behavior |
| CN109033966A (en) * | 2018-06-25 | 2018-12-18 | 北京嘀嘀无限科技发展有限公司 | Detour detection model training method and device, and detour detection method and device |
| CN110458649A (en) * | 2019-07-11 | 2019-11-15 | 北京三快在线科技有限公司 | Information recommendation method, device, electronic equipment and readable storage medium storing program for executing |
| CN111242182A (en) * | 2020-01-03 | 2020-06-05 | 联想(北京)有限公司 | Data processing method and device and electronic equipment |
| CN111833137A (en) * | 2020-01-14 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Order processing method, order processing device, computer equipment and storage medium |
| CN111833074A (en) * | 2020-01-06 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Dispute responsibility confirmation method and device for network appointment vehicle and computer readable storage medium |
| CN111863029A (en) * | 2019-09-19 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Audio-based event detection method and system |
| CN111860927A (en) * | 2019-12-20 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Model training method, service request processing method, device, equipment and medium |
-
2020
- 2020-12-24 CN CN202011547888.2A patent/CN112561108A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018205561A1 (en) * | 2017-05-09 | 2018-11-15 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for processing an abnormal order |
| CN109033966A (en) * | 2018-06-25 | 2018-12-18 | 北京嘀嘀无限科技发展有限公司 | Detour detection model training method and device, and detour detection method and device |
| CN109035768A (en) * | 2018-07-25 | 2018-12-18 | 北京交通大学 | A kind of taxi detours the recognition methods of behavior |
| CN110458649A (en) * | 2019-07-11 | 2019-11-15 | 北京三快在线科技有限公司 | Information recommendation method, device, electronic equipment and readable storage medium storing program for executing |
| CN111863029A (en) * | 2019-09-19 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Audio-based event detection method and system |
| CN111860927A (en) * | 2019-12-20 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Model training method, service request processing method, device, equipment and medium |
| CN111242182A (en) * | 2020-01-03 | 2020-06-05 | 联想(北京)有限公司 | Data processing method and device and electronic equipment |
| CN111833074A (en) * | 2020-01-06 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Dispute responsibility confirmation method and device for network appointment vehicle and computer readable storage medium |
| CN111833137A (en) * | 2020-01-14 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Order processing method, order processing device, computer equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 姚迪: "时空数据语义理解：技术与应用", 《软件学报》, vol. 29, no. 7, 31 December 2018 (2018-12-31), pages 2018 - 2045 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108921200B (en) | Method, apparatus, device and medium for classifying driving scene data | |
| CN111656735B (en) | Testing embedded systems and applications using hardware-in-the-loop services (HILAAS) | |
| US10620011B2 (en) | Autonomous vehicle routing | |
| US10252461B2 (en) | Cognitive-based driving anomaly detection based on spatio-temporal landscape-specific driving models | |
| US10775178B2 (en) | Spatio-temporal re-routing of navigation | |
| US20210123753A1 (en) | Enhanced path mapping based on safety consideration | |
| US11181387B2 (en) | Dynamic routing system | |
| CN109872360A (en) | Localization method and device, storage medium, electric terminal | |
| US10916132B2 (en) | Vehicle dash cam sign reading | |
| CN108663062B (en) | Path planning method and system | |
| CN112230632B (en) | Method, apparatus, device and storage medium for automatic driving | |
| CN111896020B (en) | Method for information processing, electronic device, and storage medium | |
| CN117725159A (en) | Data processing and model training method and device and electronic equipment | |
| CN112561108A (en) | Data processing method, device, equipment and medium | |
| WO2023045916A1 (en) | Ensuring proactive availability of vehicles for providing connected vehicle services | |
| CN112750045A (en) | Method, mobile device, and computer-readable storage medium for generating insurance information | |
| US10739155B2 (en) | Cognitive route quality-learning service | |
| CN112685531B (en) | Vehicle matching method and device, computing device and computer-readable storage medium | |
| US10783782B1 (en) | Vehicle management | |
| US11702101B2 (en) | Automatic scenario generator using a computer for autonomous driving | |
| US11030890B2 (en) | Local driver pattern based notifications | |
| US20240037825A1 (en) | Restoration of a kinetic event using video | |
| US20220309943A1 (en) | Proactive training via virtual reality simulation | |
| US20230358640A1 (en) | System and method for simulating autonomous vehicle testing environments | |
| US20220327930A1 (en) | Cooperative operation of vehicles |
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 |