CN112784787A - Device, system and method for analyzing and predicting user behavior based on deep learning - Google Patents
Device, system and method for analyzing and predicting user behavior based on deep learning Download PDFInfo
- Publication number
- CN112784787A CN112784787A CN202110127203.7A CN202110127203A CN112784787A CN 112784787 A CN112784787 A CN 112784787A CN 202110127203 A CN202110127203 A CN 202110127203A CN 112784787 A CN112784787 A CN 112784787A
- Authority
- CN
- China
- Prior art keywords
- behavior
- prediction
- user
- data
- user behavior
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/23—Updating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- 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"
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/20—Modifications to facilitate cooling, ventilating, or heating
- H05K7/20009—Modifications to facilitate cooling, ventilating, or heating using a gaseous coolant in electronic enclosures
- H05K7/20136—Forced ventilation, e.g. by fans
- H05K7/20145—Means for directing air flow, e.g. ducts, deflectors, plenum or guides
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05K—PRINTED CIRCUITS; CASINGS OR CONSTRUCTIONAL DETAILS OF ELECTRIC APPARATUS; MANUFACTURE OF ASSEMBLAGES OF ELECTRICAL COMPONENTS
- H05K7/00—Constructional details common to different types of electric apparatus
- H05K7/20—Modifications to facilitate cooling, ventilating, or heating
- H05K7/20218—Modifications to facilitate cooling, ventilating, or heating using a liquid coolant without phase change in electronic enclosures
- H05K7/20272—Accessories for moving fluid, for expanding fluid, for connecting fluid conduits, for distributing fluid, for removing gas or for preventing leakage, e.g. pumps, tanks or manifolds
Abstract
The invention discloses a device, a system and a method for analyzing and predicting user behaviors based on deep learning, which are used for acquiring data of website big data and extracting characteristics of the acquired data in time sequence to obtain corresponding user activity tracks; analyzing and undersampling the acquired sample data set, inputting the acquired subset into a decision tree model for training, and fusing the acquired single prediction models to acquire corresponding fusion models; inputting the corresponding user activity track into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target; and performing associated storage on the user behavior prediction target, and displaying and serving the constructed multi-dimensional behavior model by using a visualization technology, so that the operation performance can be improved.
Description
Technical Field
The invention relates to the technical field of deep learning, in particular to a device, a system and a method for analyzing and predicting user behaviors based on deep learning.
Background
Deep learning is the intrinsic law and expression level of the learning sample data, and the information obtained in the learning process is very helpful for the interpretation of data such as characters, images and sounds. The final aim of the method is to enable the machine to have the analysis and learning capability like a human, and to recognize data such as characters, images and sounds. Deep learning is a complex machine learning algorithm, and achieves the effect in speech and image recognition far exceeding the prior related art.
In the process of predicting the purchasing behavior of the user, the existing deep learning system needs to adopt a plurality of core processing units for calculation, and also needs to perform calculation processing on a large amount of data, so that the calculation performance is reduced.
Disclosure of Invention
The invention aims to provide a device, a system and a method for analyzing and predicting user behaviors based on deep learning, which improve the operation performance.
In order to achieve the above object, in a first aspect, the present invention provides a method for analyzing and predicting user behavior based on deep learning, including the following steps:
acquiring data of the website big data, and extracting characteristics of the acquired data in a time sequence to obtain a corresponding user activity track;
analyzing and undersampling the acquired sample data set, inputting the acquired subset into a decision tree model for training, and fusing the acquired single prediction models to acquire corresponding fusion models;
inputting the corresponding user activity track into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target;
and performing association storage on the user behavior prediction target, and displaying and serving the constructed multi-dimensional behavior model by using a visualization technology.
The method comprises the following steps of carrying out data acquisition on website big data, carrying out feature extraction on the acquired data in a time sequence, and obtaining corresponding user activity tracks, wherein the method comprises the following steps:
carrying out data acquisition on big data stored in any website by utilizing a web crawler technology;
carrying out data cleaning and data preprocessing on the acquired data;
and performing feature extraction on the preprocessed data in time sequence to obtain a corresponding user activity track.
Wherein, carry out feature extraction to the data after the preliminary treatment in the follow chronogenesis, obtain corresponding user's activity orbit, include:
acquiring corresponding user behaviors at adjacent moments in the preprocessed data;
and connecting all the user behaviors based on the internet surfing time sequence of the user, and constructing a corresponding user activity track.
The method for predicting the user behavior comprises the following steps of inputting the corresponding user activity track into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target, wherein the method comprises the following steps:
assigning the acquired corresponding attribute label and behavior label to the user activity track;
and inputting the user activity track into the fusion model for prediction to obtain a user behavior prediction target.
Inputting the user activity track into the fusion model for prediction to obtain a user behavior prediction target, wherein the method comprises the following steps:
inputting the user activity model into the fusion model, calculating the general attributes and behavior characteristics in the attribute tags and the behavior tags, and outputting corresponding execution probability;
and performing summation operation on all the execution probabilities, and comparing the obtained sum value with a set prediction threshold value to obtain a corresponding user behavior prediction target value.
The method comprises the following steps of performing association storage on the user behavior prediction target, and displaying and serving the constructed multidimensional behavior model by using a visualization technology, wherein the method comprises the following steps:
establishing a corresponding mapping relation among the user behavior prediction target, the attribute tag and the behavior tag, and storing the mapping relation;
and constructing a corresponding multi-dimensional behavior model based on the mapping relation, and displaying and serving the multi-dimensional behavior model by using a visualization technology.
In a second aspect, the present invention provides an apparatus for analyzing and predicting deep learning based user behavior, which is suitable for the method for analyzing and predicting deep learning based user behavior according to the first aspect,
the device for analyzing and predicting the user behavior based on the deep learning comprises a supporting component, a heat dissipation component and a circuit board, wherein the supporting component comprises a shell, a cover plate, a base plate and a supporting column, the shell is provided with a first groove, an air inlet hole and a heat dissipation hole, the air inlet hole and the heat dissipation hole are positioned on two sides of the first groove, the cover plate is detachably connected with the shell and covers the first groove, the base plate is fixedly connected with the shell and positioned in the shell, the supporting column is fixedly connected with the shell and positioned on one side of the shell, the heat dissipation component comprises a heat dissipation plate, a cooling pipe, a heat exchange fan, a circulating pump and a flow distribution plate, the heat dissipation plate is fixedly connected with the supporting column and positioned on one side of the supporting column, the cooling pipe is fixedly connected with the heat dissipation plate and positioned between the heat dissipation, the heat exchange tube is communicated with the cooling tube and is positioned at one side close to the heat dissipation hole, the circulating pump is communicated with the heat exchange tube and the cooling tube and is positioned at one side of the substrate, and the heat exchange fan is fixedly connected with the substrate and is positioned at one side of the substrate close to the heat dissipation hole;
the circuit board is fixedly connected with the support column and is positioned on one side of the support column, the circuit board comprises a data acquisition module, a training fusion module, a learning prediction module and a storage display module, and the data acquisition module, the training fusion module, the learning prediction module and the storage display module are sequentially connected.
In a third aspect, the present invention provides a deep learning based user behavior analysis and prediction system, which includes the deep learning based user behavior analysis and prediction apparatus according to the second aspect, and further includes an external interface, which is electrically connected to the circuit board and penetrates through the housing.
In a fourth aspect, the present invention provides a computer apparatus comprising a memory for storing program instructions and a processor for calling program instructions in the memory to perform some or all of the steps comprised in the method according to the first aspect.
In a fifth aspect, the present invention provides a computer readable storage medium having stored thereon a computer program comprising program instructions which, when executed by a processor, cause the processor to perform some or all of the steps of the method according to the first aspect.
The invention relates to a device, a system and a method for analyzing and predicting user behaviors based on deep learning, which are used for collecting data of website big data and extracting characteristics of the collected data in time sequence to obtain corresponding user activity tracks; analyzing and undersampling the acquired sample data set, inputting the acquired subset into a decision tree model for training, and fusing the acquired single prediction models to acquire corresponding fusion models; inputting the corresponding user activity track into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target; and performing associated storage on the user behavior prediction target, and displaying and serving the constructed multi-dimensional behavior model by using a visualization technology, so that the operation performance can be improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic step diagram of a method for analyzing and predicting user behavior based on deep learning according to a first embodiment of the present invention.
Fig. 2 is a detailed flow chart provided by the first embodiment of the present invention.
Fig. 3 is a schematic step diagram of a method for analyzing and predicting user behavior based on deep learning according to a second embodiment of the present invention.
Fig. 4 is a left side structural diagram of a system for analyzing and predicting user behavior based on deep learning according to the present invention.
Fig. 5 is a right-side structural diagram of a system for analyzing and predicting user behavior based on deep learning according to the present invention.
Fig. 6 is a schematic longitudinal cross-sectional view of a deep learning-based user behavior analysis and prediction system provided in the present invention.
Fig. 7 is a schematic transverse cross-sectional view of a deep learning-based user behavior analysis and prediction system according to the present invention.
Fig. 8 is a structural view of a circuit board provided by the present invention.
Fig. 9 is a schematic structural component diagram of a computer device provided by the present invention.
1-supporting component, 2-heat dissipation component, 3-circuit board, 4-external interface, 11-shell, 12-cover plate, 13-base plate, 14-supporting column, 15-mounting convex column, 16-mounting concave column, 17-non-slip pad, 18-sealing pad, 19-filter frame, 20-filter plate, 21-heat dissipation plate, 22-cooling pipe, 23-heat exchange pipe, 24-heat exchange fan, 25-circulating pump, 26-flow distribution plate, 27-protective cover, 31-data acquisition module, 32-training fusion module, 33-learning prediction module, 34-storage display module, 111-first groove, 112-air inlet hole, 113-heat dissipation hole, 221-pipe body, 222-U-shaped pipe, 223-bracket, etc, 261-board body, 262-board, 301-application program, 302-processor, 303-memory, 304-input unit, 305-display unit.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.
In the description of the present invention, it is to be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on the orientations or positional relationships illustrated in the drawings, and are used merely for convenience in describing the present invention and for simplicity in description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed in a particular orientation, and be operated, and thus, are not to be construed as limiting the present invention. Further, in the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.
Referring to fig. 1, a first embodiment of the present invention provides a method for analyzing and predicting user behavior based on deep learning, including the following steps:
s101, carrying out data acquisition on the website big data, and carrying out feature extraction on the acquired data in time sequence to obtain a corresponding user activity track.
Specifically, firstly, a web crawler technology is utilized to crawl and collect web big data stored in any or specific website, all user information and user behavior information in the current website are obtained, a corresponding sample data set is constructed, and then data cleaning and data preprocessing are carried out on the collected sample data set: and performing abnormal data cleaning (including null value and abnormal value processing and abnormal user behavior data cleaning) on the collected user information and user behavior information, and performing standardized processing according to specified data requirements.
The user information comprises user basic attribute data, commodity data and user comment data. In this embodiment, the basic attribute data of the user includes gender, age, and registration time; the user behavior data comprises operation time, operation types, objects, products and product types; the user comment data comprise the number of user comments, the number of good comments, the number of bad comments and the time of final comment; the commodity data comprises commodity codes, names, category codes, categories and commodity attributes; and carrying out abnormal value processing on the user data and the commodity data through the variance. The data preprocessing comprises the following steps: data format standardization, null value and illegal value statistics, and consistency detection of user, product and behavior data.
And (3) carrying out feature extraction on the preprocessed data in time sequence to obtain a corresponding user activity track: acquiring corresponding user behaviors at adjacent moments in the preprocessed data; and encoding the user behavior; and connecting all the user behaviors based on the internet surfing time sequence of the user, and constructing a corresponding user activity track.
The detailed process comprises the following steps: after preprocessing is carried out on data acquired by using a crawler technology, corresponding user behaviors of the same user information at two adjacent moments are extracted, the user behaviors are coded or marked, and then all the user behaviors are connected based on a user internet access time sequence according to the coded user behaviors, so that a complete user behavior track can be obtained.
S102, analyzing and undersampling the acquired sample data set, inputting the acquired subset into a decision tree model for training, and fusing the acquired single prediction models to acquire corresponding fusion models.
Specifically, as shown in fig. 2, the method includes the following steps:
and S1021, carrying out positive and negative sample classification on the acquired sample data set, and obtaining a plurality of corresponding subsets through undersampling.
Specifically, based on actual behavior data or training purposes, all data in the obtained sample data set are marked and classified to generate positive and negative sample data, unbalance conditions of the positive and negative samples are automatically analyzed, the sample data are subjected to appropriate undersampling processing according to a preset positive and negative proportion, and a plurality of positive and negative sample subsets are generated. If the user has generated the current behavior data within the prediction date, the sample data is a positive sample, and otherwise, the sample data is a negative sample.
Aiming at the unbalance of positive and negative samples possibly existing in sample data, if the data are directly used, the training result has bias, so that the invention generates a plurality of sample subsets by the positive and negative samples through dynamic undersampling processing, and each subset is independently trained. The dynamic undersampling processing of the invention is to extract a part of samples with overlarge number of samples by a certain method so as to coordinate the unbalanced proportion between the samples and other samples. In this embodiment, a certain number of subsamples are extracted from the negative samples in a random extraction manner, and are combined with the positive samples to form a new sample set for learning training, and a plurality of classifiers are trained by repeating the generation process of the new sample subset.
And S1022, inputting the subsets into a decision tree model for training, and fusing the obtained single prediction models to obtain corresponding fusion models.
Specifically, the plurality of subsets are divided into a test set and a verification set, the test set is input into a decision tree model for training to obtain a plurality of single prediction models and corresponding first prediction values, the first prediction values are compared with a preset first threshold value, if the first prediction values are smaller than the first threshold value, parameters in the decision tree model are adjusted by using a gradient descent method, then the test set is input into the decision tree model after the parameters are adjusted until the first prediction values are larger than or equal to the first threshold value, and then the verification set is input into the decision tree model for verification to obtain a plurality of second prediction values; then calculating an average value of the second predicted values, comparing the average value with a preset second threshold value, if the average value is smaller than the second threshold value, dividing the test set and the verification set again, and performing training prediction again; and if the average value is greater than or equal to the second threshold value, dividing a corresponding weight for each second predicted value according to a normal distribution principle, and fusing parameters in the decision tree model based on the weights to obtain a corresponding fusion model.
S103, inputting the corresponding user activity track into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target.
Specifically, the obtained corresponding attribute label and behavior label are assigned to the user activity track; inputting the user activity track into the fusion model for prediction to obtain a user behavior prediction target: inputting the user activity model into the fusion model, calculating the general attributes and behavior characteristics in the attribute tags and the behavior tags, and outputting corresponding execution probability; and performing summation operation on all the execution probabilities, and comparing the obtained sum value with a set prediction threshold value to obtain a corresponding user behavior prediction target value.
The detailed process comprises the following steps: assigning the obtained corresponding attribute label and behavior label to the user activity track, namely marking the user activity track, inputting the user activity track into a search fusion model for user behavior prediction, performing probability calculation on the attribute label and the behavior label on the user activity track in the fusion model, executing the execution probability of the current behavior under the current user attribute, then performing summation operation on all the execution probabilities on the user activity track, and comparing the obtained sum with a set prediction threshold value to obtain a corresponding user behavior prediction target value, namely, if the sum is greater than or equal to the prediction threshold value, indicating that the current user can execute the predicted behavior target; if the sum is smaller than the prediction threshold, the behavior target of the current user, which cannot be suffocated, is predicted, and the prediction time and the operation efficiency are reduced and the operation performance is improved through accurate analysis obtained through accurate modeling of historical data.
And S104, performing associated storage on the user behavior prediction target, and displaying and serving the constructed multi-dimensional behavior model by using a visualization technology.
Specifically, a corresponding mapping relation is established among the user behavior prediction target, the attribute tag and the behavior tag, and meanwhile, the user behavior prediction target, the attribute tag and the behavior tag are stored in combination with the internet surfing time sequence of the user; and constructing a corresponding multi-dimensional behavior model based on all the mapping relations, and displaying and serving the multi-dimensional behavior model by using a visualization technology. The specific process of constructing the multi-dimensional behavior model comprises the following steps: and obtaining all mapping relations corresponding to the current researched user based on any user, dividing the dimensionality according to the user behavior, and then sequencing and stacking the mapping relations according to the pyramid model and the data size corresponding to each mapping relation to obtain the pyramid-shaped multi-dimensional behavior model.
Referring to fig. 3, a second embodiment of the present invention provides a method for analyzing and predicting user behavior based on deep learning, including the following steps:
the descriptions of the specific implementation manners of S201 to S202 are the same as those of the specific implementation manners of S101 to S102 in the first embodiment of the present invention, and therefore, the descriptions thereof are omitted here.
S203, constructing a corresponding attribute tag database and a behavior tag database, and updating the attribute tag database and the behavior tag database.
Specifically, feature extraction is carried out according to sample data sets collected from different websites to obtain corresponding user information and user behaviors, meanwhile, weight division is carried out on authenticity magnitude values of the websites according to the magnitude of data authenticity values of the different websites, the website with higher data authenticity has higher weight values, then a corresponding attribute tag database is constructed based on the weight values and the user information, and corresponding behavior tag data are constructed according to the weight values and the user behaviors; in order to ensure the real-time performance of the attribute tag database and the behavior tag database, data are crawled for different websites in real time by using a crawler technology, and then the attribute tag database and the behavior tag database are updated after feature extraction is carried out according to the crawled data.
The descriptions of the specific implementation of S204-S205 are the same as the descriptions of the specific implementation of S103-S104 in the first embodiment of the present invention, and therefore, the descriptions thereof are omitted here.
Referring to fig. 4 to 8, the present invention provides a device for analyzing and predicting user behavior based on deep learning, including:
support assembly 1, radiator unit 2 and circuit board 3, support assembly 1 includes shell 11, apron 12, base plate 13 and support column 14, shell 11 has first recess 111, inlet port 112 and louvre 113 are located the both sides of first recess 111, apron 12 with shell 11 can dismantle the connection, and cover first recess 111, base plate 13 with shell 11 fixed connection, and be located in shell 11, support column 14 with shell 11 fixed connection, and be located one side of shell 11, radiator unit 2 includes heating panel 21, cooling tube 22, heat exchange tube 23, heat exchange fan 24, circulating pump 25 and flow distribution plate 26, heating panel 21 with support column 14 fixed connection, and be located one side of support column 14, cooling tube 22 with heating panel 21 fixed connection, the heat exchange tube 23 is communicated with the cooling tube 22 and is positioned at one side close to the heat dissipation hole 113, the circulating pump 25 is communicated with the heat exchange tube 23 and the cooling tube 22 and is positioned at one side of the substrate 13, and the heat exchange fan 24 is fixedly connected with the substrate 13 and is positioned at one side of the substrate 13 close to the heat dissipation hole 113;
the circuit board 3 with support column 14 fixed connection, and be located one side of support column 14, the circuit board 3 includes data acquisition module 31, training fuses module 32, study prediction module 33 and storage display module 34, data acquisition module 31 training fuses module 32 study prediction module 33 with storage display module 34 connects gradually.
In this embodiment, the supporting component 1 includes a housing 11, a cover plate 12, a substrate 13 and a supporting column 14, the housing 11 has a first groove 111, an air inlet 112 and a heat dissipating hole 113, the air inlet 112 and the heat dissipating hole 113 are located at two sides of the first groove, the first groove 111 is used for placing a computing component, air can enter through the air inlet 112 and be discharged from the heat dissipating hole 113 for heat dissipation, the cover plate 12 is detachably connected with the housing 11 and covers the first groove 111, the cover plate 12 is used for sealing the housing 11 to protect internal components, the substrate 13 is fixedly connected with the housing 11 and is located in the housing 11, the substrate 13 is used as a main body for installing the computing component, the supporting column 14 is fixedly connected with the housing 11 and is located at one side of the housing 11, the supporting column 14 can raise an installation height, thereby a certain interval is provided between the operation component and the substrate 13, thereby facilitating heat dissipation, the heat dissipation component 2 includes a heat dissipation plate 21, a cooling pipe 22, a heat exchange pipe 23, a heat exchange fan 24, a circulation pump 25 and a flow distribution plate 26, the heat dissipation plate 21 is fixedly connected with the supporting column 14 and is located at one side of the supporting column 14 for contacting with the operation component to conduct heat, the cooling pipe 22 is fixedly connected with the heat dissipation plate 21 and is located between the heat dissipation plate 21 and the substrate 13, the heat exchange pipe 23 is communicated with the cooling pipe 22 and is located at one side close to the heat dissipation hole 113, the circulation pump 25 is communicated with the heat exchange pipe 23 and the cooling pipe 22 and is located at one side of the substrate 13, and cooling liquid is driven to circularly flow between the cooling pipe 22 and the heat exchange pipe 23 through the circulation pump 25 to conduct heat transfer, the heat exchange fan 24 is fixedly connected to the substrate 13 and located on one side of the substrate 13 close to the heat dissipation hole 113, the heat exchange fan 24 is started to suck air from the air inlet 112 and discharge the air from the heat dissipation hole 113, so that the heat exchange tube 23 and the first groove 111 can be simultaneously dissipated, the heat dissipation efficiency is improved, the structure is more compact and the occupied space is reduced, the heat dissipation capability of the operation component placed on the support column 14 can be greatly improved, the operation can be stably performed in a high-load state, the problem that the operation performance is reduced due to poor heat dissipation effect caused by high heat generated by the operation of the existing system is solved, the data acquisition module 31, the training fusion module 32, the learning prediction module 33 and the storage display module 34 respectively execute all contents in the first embodiment and the second embodiment, the operation time and the operation efficiency can be reduced, the problem that the operation performance of the existing system is reduced due to high heating and poor heat dissipation effect can be solved.
Further, the support assembly 1 further includes a mounting convex column 15 and a mounting concave column 16, the mounting convex column 15 is fixedly connected to the housing 11 and located at one side of the housing 11, and the mounting concave column 16 is fixedly connected to the substrate 13 and located between the substrate 13 and the cover plate 12.
In this embodiment, a single device can be supported by the mounting convex column 15, and then when the calculation amount is large and a plurality of devices need to be used together, the cover plate 12 can be taken down, and then the mounting concave column 16 on other devices is connected with the mounting convex column 15, so that the occupied space can be saved, and the assembly and the placement are more stable.
Further, the support assembly 1 further includes a non-slip pad 17, and the non-slip pad 17 is in threaded connection with the mounting boss 15 and is located on one side of the mounting boss 15.
In the present embodiment, the friction between the mounting post 15 and the ground may be increased by the non-slip pad 17, so that the placement may be more stable.
Further, the support assembly 1 further comprises a sealing gasket 18, wherein the sealing gasket 18 is fixedly connected with the housing 11 and is located between the housing 11 and the cover plate 12.
In the present embodiment, the gasket 18 is provided between the housing 11 and the cover plate 12, so that the sealing performance can be enhanced to prevent foreign matter from entering therethrough.
Furthermore, the support assembly 1 further comprises a filter rack 19 and a filter plate 20, the filter rack 19 is fixedly connected with the housing 11 and is located on one side of the housing 11 close to the air inlet 112, and the filter plate 20 is detachably connected with the filter rack 19 and is located in the filter rack 19.
In this embodiment, the filter frame 19 is disposed at one side of the air inlet hole 112, and the filter plate 20 is detachably connected thereto, so that air entering the housing 11 can be filtered to prevent dust from entering the housing 11.
Further, the cooling pipe 22 includes a pipe body 221, multiple sections of U-shaped pipes 222 and a bracket 223, the bracket 223 is fixedly connected to the substrate 13 and is located on one side of the substrate 13, the multiple sections of U-shaped pipes 222 are communicated and are fixedly connected to the bracket 223, and the pipe body 221 is communicated with the U-shaped pipes 222 and the cooling pipe 22 and is located between the U-shaped pipes 222 and the cooling pipe 22.
In the present embodiment, the U-shaped tubes 222 in a plurality of stages are fixed by the bracket 223, and then the U-shaped tubes 222 in a plurality of stages are brought into contact with the heat dissipation plate 21 so that the contact area with the heat dissipation plate 21 can be increased to improve the heat dissipation efficiency, and then the tank pipe 221 guides the liquid absorbing heat to the cooling pipe 22 to cool the liquid.
Further, the supporting assembly 1 further includes a protective cover 27, and the protective cover 27 is fixedly connected to the housing 11 and covers the heat dissipation hole 113.
In this embodiment, the protection cover 27 is disposed at one side of the housing 11, so that the heat exchange fan 24 can be protected from damage due to external entering.
Further, the flow distribution plate 26 includes a plate body 261 and a plurality of support plates 262, the plate body 261 is fixedly connected to the housing 11 and is located at a side of the housing 11 close to the air inlet 112, and the plurality of support plates 262 is fixedly connected to the plate body 261 and is located at a side of the plate body 261.
In the present embodiment, the plurality of brackets 262 are supported by the plate body 261, so that after air enters from the air inlet 112, the air is blocked by the plurality of brackets 262 and can be dispersed into the space inside the housing 11, and thus the heat dissipation effect can be enhanced.
Referring to fig. 4 to 8, the present invention further provides a system for analyzing and predicting user behavior based on deep learning, which includes an external interface 4 of a device for analyzing and predicting user behavior based on deep learning, wherein the external interface 4 is electrically connected to the circuit board 3 and penetrates through the housing 11.
In the present embodiment, the external interface 4 can be connected to an external device, thereby enabling data transfer.
In an embodiment of the present invention, an application program is stored on a computer-readable storage medium, and when the application program is executed by a processor, the method for analyzing and predicting user behavior based on deep learning in any one of the above embodiments is implemented. The computer-readable storage medium includes, but is not limited to, any type of disk including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks, ROMs (Read-Only memories), RAMs (random access memories), EPROMs (EraSable Programmable Read-Only memories), EEPROMs (Electrically EraSable Programmable Read-Only memories), flash memories, magnetic cards, or optical cards. That is, a storage device includes any medium that stores or transmits information in a form readable by a device (e.g., a computer, a cellular phone), and may be a read-only memory, a magnetic or optical disk, or the like.
The embodiment of the invention also provides a computer application program which runs on a computer and is used for executing the method for analyzing and predicting the user behavior based on deep learning in any one of the embodiments.
Fig. 9 is a schematic structural diagram of a computer device in the embodiment of the present invention.
An embodiment of the present invention further provides a computer device, as shown in fig. 9. The computer apparatus includes a processor 302, a memory 303, an input unit 304, a display unit 305, and the like. Those skilled in the art will appreciate that the device configuration means shown in fig. 9 do not constitute a limitation of all devices and may include more or less components than those shown, or some components in combination. The memory 303 may be used to store the application 301 and various functional modules, and the processor 302 executes the application 301 stored in the memory 303, thereby performing various functional applications of the device and data processing. The memory may be internal or external memory, or include both internal and external memory. The memory may comprise read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), flash memory, or random access memory. The external memory may include a hard disk, a floppy disk, a ZIP disk, a usb-disk, a magnetic tape, etc. The disclosed memory includes, but is not limited to, these types of memory. The disclosed memory is by way of example only and not by way of limitation.
The input unit 304 is used for receiving input of signals and receiving keywords or images input by a user. The input unit 304 may include a touch panel and other input devices. The touch panel can collect touch operations of a user on or near the touch panel (for example, operations of the user on or near the touch panel by using any suitable object or accessory such as a finger, a stylus and the like) and drive the corresponding connecting device according to a preset program; other input devices may include, but are not limited to, one or more of a physical keyboard, function keys (e.g., play control keys, switch keys, etc.), a trackball, a mouse, a joystick, and the like. The display unit 305 may be used to display information input by a user or information provided to the user and various menus of the terminal device. The display unit 305 may take the form of a liquid crystal display, an organic light emitting diode, or the like. The processor 302 is a control center of the terminal device, connects various parts of the entire device using various interfaces and lines, and performs various functions and processes data by running or executing software programs and/or modules stored in the memory 303 and calling data stored in the memory.
As one embodiment, the computer device includes: one or more processors 302, a memory 303, one or more applications 301, wherein the one or more applications 301 are stored in the memory 303 and configured to be executed by the one or more processors 302, the one or more applications 301 configured to perform a method for deep learning based user behavior analysis and prediction in any of the above embodiments.
The invention relates to a device, a system and a method for analyzing and predicting user behaviors based on deep learning, which are used for collecting data of website big data and extracting characteristics of the collected data in time sequence to obtain corresponding user activity tracks; analyzing and undersampling the acquired sample data set, inputting the acquired subset into a decision tree model for training, and fusing the acquired single prediction models to acquire corresponding fusion models; inputting the corresponding user activity track into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target; and performing associated storage on the user behavior prediction target, and displaying and serving the constructed multi-dimensional behavior model by using a visualization technology, so that the operation performance can be improved.
While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A method for analyzing and predicting user behavior based on deep learning is characterized by comprising the following steps:
acquiring data of the website big data, and extracting characteristics of the acquired data in a time sequence to obtain a corresponding user activity track;
analyzing and undersampling the acquired sample data set, inputting the acquired subset into a decision tree model for training, and fusing the acquired single prediction models to acquire corresponding fusion models;
inputting the corresponding user activity track into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target;
and performing association storage on the user behavior prediction target, and displaying and serving the constructed multi-dimensional behavior model by using a visualization technology.
2. The method for analyzing and predicting user behavior based on deep learning of claim 1, wherein the step of collecting data of website big data and extracting features of the collected data from time sequence to obtain corresponding user activity track comprises:
carrying out data acquisition on big data stored in any website by utilizing a web crawler technology;
carrying out data cleaning and data preprocessing on the acquired data;
and performing feature extraction on the preprocessed data in time sequence to obtain a corresponding user activity track.
3. The method of claim 2, wherein the step of performing feature extraction on the preprocessed data in time sequence to obtain a corresponding user activity track comprises:
acquiring corresponding user behaviors at adjacent moments in the preprocessed data;
and connecting all the user behaviors based on the internet surfing time sequence of the user, and constructing a corresponding user activity track.
4. The method for analyzing and predicting user behavior based on deep learning of claim 1, wherein the step of inputting the corresponding user activity trajectory into the fusion model for prediction based on the corresponding attribute label and behavior label to obtain a user behavior prediction target comprises:
assigning the acquired corresponding attribute label and behavior label to the user activity track;
and inputting the user activity track into the fusion model for prediction to obtain a user behavior prediction target.
5. The method of deep learning based user behavior analysis and prediction as claimed in claim 4, wherein inputting the user activity trajectory into the fusion model for prediction to obtain a user behavior prediction objective comprises:
inputting the user activity model into the fusion model, calculating the general attributes and behavior characteristics in the attribute tags and the behavior tags, and outputting corresponding execution probability;
and performing summation operation on all the execution probabilities, and comparing the obtained sum value with a set prediction threshold value to obtain a corresponding user behavior prediction target value.
6. The method for analyzing and predicting the user behavior based on the deep learning of claim 1, wherein the step of performing the associative storage on the user behavior prediction target and performing the display and the service on the constructed multidimensional behavior model by using a visualization technology comprises the following steps:
establishing a corresponding mapping relation among the user behavior prediction target, the attribute tag and the behavior tag, and storing the mapping relation;
and constructing a corresponding multi-dimensional behavior model based on the mapping relation, and displaying and serving the multi-dimensional behavior model by using a visualization technology.
7. An apparatus for analyzing and predicting deep learning based user behavior, which is suitable for the method of analyzing and predicting deep learning based user behavior according to any one of claims 1 to 6,
the device for analyzing and predicting the user behavior based on the deep learning comprises a supporting component, a heat dissipation component and a circuit board, wherein the supporting component comprises a shell, a cover plate, a base plate and a supporting column, the shell is provided with a first groove, an air inlet hole and a heat dissipation hole, the air inlet hole and the heat dissipation hole are positioned on two sides of the first groove, the cover plate is detachably connected with the shell and covers the first groove, the base plate is fixedly connected with the shell and positioned in the shell, the supporting column is fixedly connected with the shell and positioned on one side of the shell, the heat dissipation component comprises a heat dissipation plate, a cooling pipe, a heat exchange fan, a circulating pump and a flow distribution plate, the heat dissipation plate is fixedly connected with the supporting column and positioned on one side of the supporting column, the cooling pipe is fixedly connected with the heat dissipation plate and positioned between the heat dissipation, the heat exchange tube is communicated with the cooling tube and is positioned at one side close to the heat dissipation hole, the circulating pump is communicated with the heat exchange tube and the cooling tube and is positioned at one side of the substrate, and the heat exchange fan is fixedly connected with the substrate and is positioned at one side of the substrate close to the heat dissipation hole;
the circuit board is fixedly connected with the support column and is positioned on one side of the support column, the circuit board comprises a data acquisition module, a training fusion module, a learning prediction module and a storage display module, and the data acquisition module, the training fusion module, the learning prediction module and the storage display module are sequentially connected.
8. A deep learning based user behavior analysis and prediction system comprising the deep learning based user behavior analysis and prediction apparatus of claim 7,
the circuit board is electrically connected with the circuit board and penetrates through the shell.
9. A computer device comprising a memory for storing program instructions and a processor for invoking the program instructions in the memory to perform some or all of the steps included in the method of any of claims 1-6.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program comprising program instructions which, when executed by a processor, cause the processor to carry out part or all of the steps of the method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110127203.7A CN112784787B (en) | 2021-01-29 | 2021-01-29 | Device, system and method for analyzing and predicting user behavior based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110127203.7A CN112784787B (en) | 2021-01-29 | 2021-01-29 | Device, system and method for analyzing and predicting user behavior based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112784787A true CN112784787A (en) | 2021-05-11 |
| CN112784787B CN112784787B (en) | 2022-06-17 |
Family
ID=75759819
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110127203.7A Active CN112784787B (en) | 2021-01-29 | 2021-01-29 | Device, system and method for analyzing and predicting user behavior based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112784787B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110147149A (en) * | 2019-05-22 | 2019-08-20 | 重庆工业职业技术学院 | A kind of computer hardware rapid cooling and maintenance component and its cool-down method |
| CN110555717A (en) * | 2019-07-29 | 2019-12-10 | 华南理工大学 | method for mining potential purchased goods and categories of users based on user behavior characteristics |
| CN211123917U (en) * | 2019-12-30 | 2020-07-28 | 广州龙诚电子实业有限公司 | Intelligent temperature control heat dissipation computer case |
| CN112085525A (en) * | 2020-09-04 | 2020-12-15 | 长沙理工大学 | User network purchasing behavior prediction research method based on hybrid model |
-
2021
- 2021-01-29 CN CN202110127203.7A patent/CN112784787B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110147149A (en) * | 2019-05-22 | 2019-08-20 | 重庆工业职业技术学院 | A kind of computer hardware rapid cooling and maintenance component and its cool-down method |
| CN110555717A (en) * | 2019-07-29 | 2019-12-10 | 华南理工大学 | method for mining potential purchased goods and categories of users based on user behavior characteristics |
| CN211123917U (en) * | 2019-12-30 | 2020-07-28 | 广州龙诚电子实业有限公司 | Intelligent temperature control heat dissipation computer case |
| CN112085525A (en) * | 2020-09-04 | 2020-12-15 | 长沙理工大学 | User network purchasing behavior prediction research method based on hybrid model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112784787B (en) | 2022-06-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108664538B (en) | Automatic identification method and system for suspected familial defects of power transmission and transformation equipment | |
| US8200010B1 (en) | Image segmentation by clustering web images | |
| US20120045132A1 (en) | Method and apparatus for localizing an object within an image | |
| US20120020563A1 (en) | Systems and Methods for Automated Extraction of Measurement Information in Medical Videos | |
| US9652695B2 (en) | Label consistency for image analysis | |
| WO2021174812A1 (en) | Data cleaning method and apparatus for profile, and medium and electronic device | |
| EP3333757B1 (en) | Predictive anomaly detection | |
| US20160078087A1 (en) | Query features and questions | |
| CN113254354A (en) | Test case recommendation method and device, readable storage medium and electronic equipment | |
| CN113449187A (en) | Product recommendation method, device and equipment based on double portraits and storage medium | |
| CN108846117A (en) | The duplicate removal screening technique and device of business news flash | |
| CN103530312A (en) | User identification method and system using multifaceted footprints | |
| CN112070506A (en) | Risk user identification method, device, server and storage medium | |
| CN109389050A (en) | A kind of flow chart connection relationship recognition methods | |
| CN111489262A (en) | Policy information detection method and device, computer equipment and storage medium | |
| CN112784787B (en) | Device, system and method for analyzing and predicting user behavior based on deep learning | |
| Wang et al. | Research on thermal state diagnosis of substation equipment based on infrared image | |
| Gao et al. | Quantifying animal trajectories using spatial aggregation and sequence analysis: a case study of differentiating trajectories of multiple species | |
| CN113569118B (en) | Self-media pushing method, device, computer equipment and storage medium | |
| CN112133449A (en) | Method and device for predicting induced hepatotoxicity, computer equipment and storage medium | |
| KR102056709B1 (en) | Method and computer program for extracting table record from spread sheet | |
| CN111414519A (en) | Information processing method, device, computer system and readable storage medium | |
| CN113343306B (en) | Differential privacy-based data query method, device, equipment and storage medium | |
| CN115062200A (en) | User behavior mining method and system based on artificial intelligence | |
| CN113806492A (en) | Record generation method, device and equipment based on semantic recognition and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20221209 Address after: Room 306-5, Building 2, Yintan Cultural Innovation Park, No. 88 Yapeng Road, Jianye District, Nanjing, Jiangsu Province, 210000 Patentee after: Nanjing Hongxin Digital Technology Co.,Ltd. Address before: 211800 room 2412, block B, Kechuang headquarters building, 320 pubin Avenue, Jiangpu street, Pukou District, Nanjing City, Jiangsu Province Patentee before: Nanjing zhishuyun Information Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right |