A data analysis method, device, equipment and medium based on artificial intelligence

Abstract

The present application relates to the field of artificial intelligence technology, and in particular to a data analysis method, device, equipment and medium based on artificial intelligence. The method includes: obtaining data to be analyzed, classifying the data to be analyzed according to data types, and obtaining a data set corresponding to each data type; determining a preset result item corresponding to each data set based on a preset correspondence, where the preset correspondence is a correspondence between a data type and a preset result item; performing data analysis on a target data set based on the data type of the target data set, and obtaining a data analysis result containing preset result items of the target data set, where the target data set is any data set obtained by classifying the data to be analyzed. The present application can realize a comprehensive analysis of unstructured data.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an artificial intelligence-based data analysis method, device, equipment and medium. Background Art

[0002] With the advent of the big data era, all industries are facing unprecedented challenges in data processing and analysis. Traditional data analysis methods are inefficient when processing massive and diverse data types (such as text, images, videos, etc.) and it is difficult to ensure the real-time and comprehensiveness of the analysis results. Data analysis software in related technologies is often limited to specific fields, has limited processing capabilities for unstructured data, and is not comprehensive enough when processing large-scale unstructured data. Summary of the invention

[0003] In order to solve the problem that the existing technology is not comprehensive enough in processing and analyzing unstructured data, the present application provides a data analysis method, device, equipment and medium based on artificial intelligence.

[0004] In the first aspect, the present application provides a data analysis method based on artificial intelligence, which adopts the following technical solutions:

[0005] A data analysis method based on artificial intelligence, comprising:

[0006] Acquire data to be analyzed, classify the data to be analyzed according to data types, and obtain a data set corresponding to each data type;

[0007] Determine a preset result item corresponding to each data set based on a preset corresponding relationship, where the preset corresponding relationship is a corresponding relationship between a data type and a preset result item;

[0008] The target data set is analyzed based on its data type to obtain a data analysis result including preset result items of the target data set, wherein the target data set is any data set obtained by classifying the data to be analyzed.

[0009] In a preferred example, the present application may be further configured as follows: performing data analysis on the target data set based on the data type of the target data set to obtain a data analysis result containing preset result items of the target data set, including:

[0010] Constructing a feature matrix of the target data set;

[0011] A data analysis result of the target data set is generated based on the feature matrix, and preset result items of the data analysis result include a content analysis result and a sentiment analysis result.

[0012] In a preferred example, the present application may be further configured as follows: constructing the feature matrix of the target data includes:

[0013] Extracting an initial feature matrix of the target data set;

[0014] Calculate the similarity between every two feature points in the initial feature matrix to obtain a similarity matrix;

[0015] Converting the similarity matrix into a local feature matrix, wherein each data in the local feature matrix represents the connection strength between two feature points;

[0016] A global feature matrix is ​​constructed based on the local feature matrix, and the local feature matrix and the global feature matrix are combined into a feature matrix.

[0017] In a preferred example, the present application may be further configured as follows: the content analysis result includes a category analysis result, and the category analysis result of the target data set is generated based on the feature matrix, including:

[0018] Performing cluster analysis on the feature matrix to obtain multiple clusters;

[0019] Determine the density and size of each cluster, and calculate the weight value of a target cluster based on the density and size, wherein the target cluster is any one of the multiple clusters;

[0020] Calculating the similarity between the target cluster and each preset category, and calculating the probability that the target cluster belongs to each preset category based on the weight value of the target cluster and the similarity, and taking the preset category with the largest probability as the target preset category to which the target cluster belongs;

[0021] Based on the probability of each cluster in the plurality of clusters and the corresponding preset target category, the target category to which the target data set belongs is determined, and the target category serves as the category analysis result of the target data set.

[0022] In a preferred example, the present application may be further configured as follows: the step of calculating the weight value of the target cluster based on the density and the size includes:

[0023] Sorting the plurality of clusters from small to large according to density to obtain a first sequence, and determining the first order of the target cluster in the first sequence;

[0024] Sorting the plurality of clusters in ascending order of size to obtain a second sequence, and determining a second order of the target cluster in the second sequence;

[0025] Calculating the sum of the first sequence and the second sequence as the target sequence of the target cluster;

[0026] The target positions corresponding to the plurality of clusters are normalized to obtain a weight value of each cluster.

[0027] In a preferred example, the present application may be further configured to: generate a sentiment analysis result of the target data set based on the feature matrix, including:

[0028] The feature matrix is ​​input into a preset sentiment analysis model to obtain a sentiment analysis result output by the sentiment analysis model, wherein the sentiment analysis result includes a sentiment label and a sentiment level.

[0029] In a preferred example, the present application may be further configured as follows: when the data type of the target data set is a video type, the preset result item further includes a target detection result, and the target detection result is obtained by performing data analysis on the target data set, including:

[0030] Dividing the target data set into multiple frame images with a time sequence;

[0031] Performing target detection on the multiple frames of images to identify the target in each frame of image;

[0032] The same targets detected in consecutive frames of the multiple frames are matched, and the matched targets are integrated according to a time series to obtain a motion trajectory of each target, and the motion trajectory of each target is used as the target detection result.

[0033] In a second aspect, the present application provides a data analysis device based on artificial intelligence, which adopts the following technical solution:

[0034] A data analysis device based on artificial intelligence, comprising:

[0035] A classification module is used to obtain the data to be analyzed, classify the data to be analyzed according to data types, and obtain a data set corresponding to each data type;

[0036] A determination module, used to determine a preset result item corresponding to each data set based on a preset corresponding relationship, wherein the preset corresponding relationship is a corresponding relationship between a data type and a preset result item;

[0037] The analysis module is used to perform data analysis on the target data set based on the data type of the target data set to obtain a data analysis result including preset result items of the target data set. The target data set is any data set obtained by classifying the data to be analyzed.

[0038] In a third aspect, the present application provides an electronic device, which adopts the following technical solution:

[0039] one or more processors;

[0040] Memory;

[0041] At least one application, wherein the at least one application is stored in a memory and configured to be executed by at least one processor, and the at least one application is configured to: execute the artificial intelligence-based data analysis method as described in any one of the first aspects.

[0042] In a fourth aspect, the present application provides a computer-readable storage medium, which adopts the following technical solution:

[0043] A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the artificial intelligence-based data analysis method as described in any one of the first aspects.

[0044] In a fifth aspect, the present application provides a computer program product, which adopts the following technical solution:

[0045] A computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program implements the artificial intelligence-based data analysis method as described in any one of the first aspects.

[0046] In summary, this application includes the following beneficial technical effects:

[0047] This application collects and integrates the data to be analyzed and classifies the data to obtain a data set with a clearer structure and clearer type, which is helpful for adopting more suitable analysis methods for different types of data in the future. The classified data sets can be processed in parallel, thereby improving the overall efficiency of data analysis. The preset corresponding relationship defines the analysis results that should be obtained for each data type, which helps to clarify the analysis goals and ensure that different types of data sets obtain consistent analysis results. Data analysis is performed based on the data type of the target data set to ensure the pertinence and accuracy of the analysis. For different types of data, the most suitable analysis algorithms and models are used to optimize the analysis effect and obtain more comprehensive data analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flowchart of a data analysis method based on artificial intelligence provided in an embodiment of the present application;

[0049] Figure 2 It is a structural schematic diagram of an artificial intelligence-based data analysis device provided in an embodiment of the present application;

[0050] Figure 3 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The following is combined with Figure 1 -Attached Figure 3 This application is described in further detail.

[0052] This specific embodiment is merely an explanation of the present application and is not a limitation of the present application. After reading this specification, those skilled in the art may make modifications to the present embodiment without any creative contribution as needed, but such modifications are protected by the patent law as long as they are within the scope of the claims of the present application.

[0053] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0055] It should be noted that in the optional embodiments of the present application, the object information and other related data involved, when the embodiments in the present application are applied to specific products or technologies, need to obtain the permission or consent of the object, and the collection, use and processing of the relevant data need to comply with the relevant laws, regulations and standards of the relevant countries and regions. In other words, if the embodiments of the present application involve data related to the object, it needs to be obtained with the authorization and consent of the object, the authorization and consent of the relevant departments, and in compliance with the relevant laws, regulations and standards of the country and region. If personal information is involved in the embodiments, the acquisition of all personal information needs to obtain the consent of the individual. If sensitive information is involved, the separate consent of the information subject needs to be obtained. The embodiments also need to be implemented with the authorization and consent of the object.

[0056] The present application embodiment provides a data analysis method based on artificial intelligence, such as Figure 1As shown, the method provided in the embodiment of the present application is executed by an electronic device, which can be a server or a terminal device, wherein the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc., but is not limited thereto. The terminal device and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiment of the present application. The method includes steps S101 to S103, wherein:

[0057] S101 , obtaining data to be analyzed, classifying the data to be analyzed according to data types, and obtaining a data set corresponding to each data type.

[0058] Specifically, the data source for obtaining the data to be analyzed can be a database, a file system or a network. The data initially obtained is cleaned to remove invalid or duplicate data, thereby ensuring the integrity and consistency of the data. The data types include text type, audio type, image type and video type. One or more identification features are defined in advance for each data type. The identification features can be file extension, MIME type, data format, etc.

[0059] Then, the data type of each piece of data in the data to be analyzed is determined based on the correspondence between the data type and the identification feature, and the data to be analyzed is classified according to the data type. In one possible case, the data to be analyzed is of one data type, and the number of data sets obtained is one; in another possible case, the data to be analyzed is of two or more data types, and the number of data sets obtained is multiple.

[0060] S102: Determine a preset result item corresponding to each data set based on a preset corresponding relationship, where the preset corresponding relationship is a corresponding relationship between a data type and a preset result item.

[0061] Specifically, the data analysis results include content analysis results and sentiment analysis results. For text types, audio types, and image types, the content analysis results include category analysis results, that is, the preset result items corresponding to the data types include category analysis results and sentiment analysis results; for video types, the content analysis results include category analysis results and target detection results, that is, the preset result items corresponding to the video type data type include category analysis results, target detection results, and sentiment analysis results.

[0062] S103 , performing data analysis on the target data set based on the data type of the target data set to obtain a data analysis result including preset result items of the target data set, where the target data set is any data set obtained by classifying the data to be analyzed.

[0063] A possible implementation of the embodiment of the present application is to perform data analysis on the target data set based on the data type of the target data set to obtain a data analysis result containing preset result items of the target data set, including:

[0064] Construct the feature matrix of the target data set;

[0065] The data analysis results of the target data set are generated based on the feature matrix, and the preset result items of the data analysis results include content analysis results and sentiment analysis results.

[0066] In this embodiment, the corresponding feature extraction method can be selected according to the data type, and the features extracted from the target data set are represented as an initial feature matrix, and then the local feature matrix and the overall feature matrix are extracted from the initial feature matrix, and the local feature matrix and the global feature matrix are combined into a feature matrix. Then, by clustering the feature matrix, the category analysis result of the target data set is obtained; the sentiment analysis result can be analyzed by a preset sentiment analysis model; when the target data set is a video type, the target detection result of the target data set can be obtained by dividing the frame image and the target detection analysis.

[0067] A possible implementation of the embodiment of the present application is to construct a feature matrix of target data, including:

[0068] Extract the initial feature matrix of the target data set;

[0069] Calculate the similarity between every two feature points in the initial feature matrix to obtain a similarity matrix;

[0070] The similarity matrix is ​​converted into a local feature matrix, where each data in the local feature matrix represents the connection strength between two feature points;

[0071] A global feature matrix is ​​constructed based on the local feature matrix, and the local feature matrix and the global feature matrix are combined into a feature matrix.

[0072] In this embodiment, the feature extraction method can be determined according to the data type. If the target data set is text-type data, the text features can be extracted by natural language processing technology, and the extracted text features can be converted into numerical vector representations to form an initial feature matrix. If the target data set is audio-type data, Mel-frequency cepstral coefficients, chromaticity features, spectral contrast, etc. can be extracted, and the extracted features can be organized into an initial feature matrix, where each row represents an audio clip and each column represents a feature. If the target data set is image-type or video-type data, a corner detection algorithm, a scale-invariant feature transform, or an accelerated robust feature can be used to extract features, and the extracted features can be organized into an initial feature matrix, where each row represents an image or video frame, and each column represents a feature. Among them, the feature extraction method used for each data type and the extracted feature type can be flexibly selected according to actual needs, and this embodiment does not make specific restrictions.

[0073] Each row in the initial feature matrix represents a feature vector of a data point, and each column represents a feature. The method for calculating the similarity between each two data points in the initial feature matrix can be Euclidean distance or cosine similarity. For example, when there are n data points, the similarity between each two data points is calculated to obtain an n×n matrix, where the element S ij Indicates the similarity between the i-th and j-th data points. Then, a local feature matrix is ​​constructed based on the similarity matrix. The local feature matrix can be represented by an adjacency matrix. Specifically, the similarity matrix is ​​divided into two cases: a symmetric matrix and an asymmetric matrix. When the similarity matrix is ​​an asymmetric matrix, the similarity matrix is ​​converted into a symmetric matrix. By taking S ij and S ji The average value is realized, and the element M in the symmetric matrix ij =(S ij +S ji ) / 2, the values ​​in the symmetric similarity matrix are converted into weights in the adjacency matrix, which involves converting similarities into probabilities so that the sum of the connection strengths of each data point is 1. In the adjacency matrix, each element W ij represents the connection strength between the i-th and j-th data points, , where exp refers to the exponential function and the denominator is the sum of the exponential function values ​​of all elements in the i-th row, ensuring that the sum of each row is 1.

[0074] The degree matrix D is constructed based on the local feature matrix. The degree matrix D is a diagonal matrix, in which each diagonal element D ij It represents the sum of the elements in the local feature matrix of the i-th row, that is, the total weight connected to the i-th data point. The obtained degree matrix D is used as the global feature matrix, and the difference between the global feature matrix D and the local feature matrix W is calculated to obtain the Laplace matrix, that is, the feature matrix.

[0075] In a possible implementation of the embodiment of the present application, the content analysis result includes a category analysis result, and the category analysis result of the target data set is generated based on the feature matrix, including:

[0076] Perform cluster analysis on the feature matrix to obtain multiple clusters;

[0077] Determine the density and size of each cluster, and calculate the weight value of the target cluster based on the density and size, where the target cluster is any one of the multiple clusters;

[0078] Calculate the similarity between the target cluster and each preset category, and calculate the probability that the target cluster belongs to each preset category based on the weight value and similarity of the target cluster, and take the preset category with the largest probability as the target preset category to which the target cluster belongs;

[0079] Based on the probability of each cluster in the plurality of clusters and the corresponding preset target category, the target category to which the target data set belongs is determined, and the target category is used as the category analysis result of the target data set.

[0080] In this embodiment, a clustering algorithm (K-means, hierarchical clustering, DBSCAN, etc.) is used to perform cluster analysis on the feature matrix to obtain multiple clusters, so that data points with similar features are divided into the same cluster. For each cluster, its density and size are calculated, the average value of the distance between each point in the cluster is calculated as the density, the number of each data point in the cluster is used as the size, and the weight value is calculated based on the density and size of the cluster. The weight value reflects the stability and importance of the cluster. The greater the density and the larger the size, the higher the weight value.

[0081] Calculate the similarity between the target cluster and each preset category. The similarity can be calculated by cosine similarity or Euclidean distance. For any preset category, weight the target cluster and the similarity between the target cluster and the preset category. The sum is used as the probability that the target cluster belongs to the preset category. The weight of the target cluster and the similarity between the target cluster and the preset category can be flexibly set according to actual experience, and this embodiment does not specifically limit it. Obtain the probability that the target cluster belongs to each preset category, and determine the preset category with the highest probability as the target preset category to which the target cluster belongs, thereby obtaining a target preset category corresponding to each cluster, and taking the probability that each cluster belongs to the target preset category as the target probability.

[0082] Furthermore, for any preset target category, the sum of target probabilities of each cluster belonging to the preset target category is determined, and the one with the largest sum of target probabilities among the preset target categories is taken as the target category.

[0083] A possible implementation of the embodiment of the present application is to calculate the weight value of the target cluster based on density and size, including:

[0084] Arrange the multiple clusters in ascending order according to density to obtain a first sequence, and determine the first order of the target cluster in the first sequence;

[0085] Sort the multiple clusters in ascending order of size to obtain a second sequence, and determine a second position of the target cluster in the second sequence;

[0086] Calculate the sum of the first and second order as the target order of the target cluster;

[0087] The target positions corresponding to the multiple clusters are normalized to obtain the weight value of each cluster.

[0088] In the first order, the more forward the position of the target cluster, the smaller the first order of the target cluster, indicating that the density of the target cluster is smaller. In the second order, the more forward the position of the target cluster, the smaller the second order of the target cluster, indicating that the size of the target cluster is smaller. The first order and the second order are added to obtain the target order of the target cluster. The target order combines density and size to reflect the relative position of the target cluster in the two dimensions of density and size.

[0089] The target positions of all clusters are normalized to convert the target position values ​​to a uniform range, such as 0-1, find the minimum and maximum values ​​of the target positions in all clusters, and then subtract the minimum value from the target position of each cluster, and divide it by the difference between the maximum and minimum values ​​to get the normalized weight value.

[0090] A possible implementation of the embodiment of the present application generates a sentiment analysis result of a target data set based on a feature matrix, including:

[0091] The feature matrix is ​​input into the preset sentiment analysis model to obtain the sentiment analysis results output by the sentiment analysis model. The sentiment analysis results include sentiment labels and sentiment levels.

[0092] In this embodiment, the preset sentiment analysis model is pre-built based on machine learning or deep learning, and is used to identify and analyze sentiment in a data set. The training process includes: collecting a large amount of data with sentiment labels and sentiment level labels. The labels can be manually labeled or generated based on rules or algorithms. The acquired data is divided into a training set, a validation set, and a test set. The training set is used to train the model, the validation set is used to adjust model parameters and select the best model, and the test set is used to evaluate model performance, so as to finally obtain a trained preset sentiment analysis model.

[0093] In a possible implementation of the embodiment of the present application, when the data type of the target data set is a video type, the preset result item further includes a target detection result, and the target detection result is obtained by performing data analysis on the target data set, including:

[0094] Divide the target dataset into multiple frame images with time series;

[0095] Perform target detection on multiple frames of images and identify the target in each frame;

[0096] The same targets detected in consecutive frames of the multi-frame images are matched, and the matched targets are integrated according to the time series to obtain the motion trajectory of each target. The motion trajectory of each target is used as the target detection result.

[0097] In this embodiment, the target data set is divided into a series of separate frame images according to the time series, so that the video data is decoded into a series of image frames, each frame representing a time point in the video. The model selected for the target detection task can be a model based on deep learning (such as YOLO, SSD, Faster R-CNN, etc.) or a traditional computer vision method (such as HOG+SVM, background subtraction, etc.).

[0098] Use the selected target detection model to perform target detection on each frame of the video, locate and identify the target of interest (such as people, cars, animals, etc.) in the image. For each target detected in each frame, record its position (such as bounding box coordinates) and category (such as people, cars, etc.). For images in consecutive frames, use matching algorithms (such as position-based nearest neighbor matching, feature-based matching, etc.) to identify and match the same target, including comparing the position, size, shape and other features of the targets in adjacent frames to determine whether they belong to the same target. Once the same target in consecutive frames is matched, these targets can be integrated in time series to form the motion trajectory of each target. Among them, the motion trajectory of the target may be broken due to errors in target detection, occlusion, or the target entering and leaving the field of view. Based on this, some trajectory smoothing techniques (such as Kalman filtering, particle filtering, etc.) can be used to predict and fill in the missing parts of the trajectory.

[0099] Output the motion trajectory of each target in an appropriate format, including the target ID, category, position in each frame (bounding box coordinates), timestamps of appearance and disappearance, etc. In order to more intuitively display the target detection results, the target's motion trajectory can be superimposed on the original video frame and a video or image sequence with trajectory markers can be generated.

[0100] The above-mentioned embodiment introduces a data analysis method based on artificial intelligence from the perspective of method flow, and the following embodiment introduces a data analysis device based on artificial intelligence from the perspective of a virtual module or a virtual unit. Please refer to the following embodiments for details.

[0101] The present application embodiment provides a data analysis device based on artificial intelligence, such as Figure 2 As shown, the device may include:

[0102] The classification module 201 is used to obtain the data to be analyzed, classify the data to be analyzed according to the data type, and obtain the data set corresponding to each data type;

[0103] A determination module 202, configured to determine a preset result item corresponding to each data set based on a preset corresponding relationship, where the preset corresponding relationship is a corresponding relationship between a data type and a preset result item;

[0104] The analysis module 203 is used to perform data analysis on the target data set based on the data type of the target data set to obtain a data analysis result including preset result items of the target data set. The target data set is any data set obtained by classifying the data to be analyzed.

[0105] In a preferred example, the present application may be further configured as follows: when the analysis module 203 performs data analysis on the target data set based on the data type of the target data set to obtain a data analysis result including a preset result item of the target data set, it is specifically used to:

[0106] Construct the feature matrix of the target data set;

[0107] The data analysis results of the target data set are generated based on the feature matrix, and the preset result items of the data analysis results include content analysis results and sentiment analysis results.

[0108] In a preferred example, the present application can be further configured as follows: when the analysis module 203 executes the construction of the feature matrix of the target data, it is specifically used to:

[0109] Extract the initial feature matrix of the target data set;

[0110] Calculate the similarity between every two feature points in the initial feature matrix to obtain a similarity matrix;

[0111] The similarity matrix is ​​converted into a local feature matrix, where each data in the local feature matrix represents the connection strength between two feature points;

[0112] A global feature matrix is ​​constructed based on the local feature matrix, and the local feature matrix and the global feature matrix are combined into a feature matrix.

[0113] In a preferred example, the present application can be further configured as follows: when the analysis module 203 generates the category analysis result of the target data set based on the feature matrix, it is specifically used to:

[0114] Perform cluster analysis on the feature matrix to obtain multiple clusters;

[0115] Determine the density and size of each cluster, and calculate the weight value of the target cluster based on the density and size, where the target cluster is any one of the multiple clusters;

[0116] Calculate the similarity between the target cluster and each preset category, and calculate the probability that the target cluster belongs to each preset category based on the weight value and similarity of the target cluster, and take the preset category with the largest probability as the target preset category to which the target cluster belongs;

[0117] Based on the probability of each cluster in the plurality of clusters and the corresponding preset target category, the target category to which the target data set belongs is determined, and the target category is used as the category analysis result of the target data set.

[0118] In a preferred example, the present application can be further configured as follows: when the analysis module 203 calculates the weight value of the target cluster based on density and size, it is specifically used to:

[0119] Arrange the multiple clusters in ascending order according to density to obtain a first sequence, and determine the first order of the target cluster in the first sequence;

[0120] Sort the multiple clusters in ascending order of size to obtain a second sequence, and determine a second position of the target cluster in the second sequence;

[0121] Calculate the sum of the first and second order as the target order of the target cluster;

[0122] The target positions corresponding to the multiple clusters are normalized to obtain the weight value of each cluster.

[0123] In a preferred example, the present application can be further configured as follows: when the analysis module 203 generates the sentiment analysis result of the target data set based on the feature matrix, it is specifically used to:

[0124] The feature matrix is ​​input into the preset sentiment analysis model to obtain the sentiment analysis results output by the sentiment analysis model. The sentiment analysis results include sentiment labels and sentiment levels.

[0125] In a preferred example, the present application can be further configured as follows: when the analysis module 203 performs data analysis on the target data set to obtain the target detection result, it is specifically used to:

[0126] Divide the target dataset into multiple frame images with time series;

[0127] Perform target detection on multiple frames of images and identify the target in each frame;

[0128] The same targets detected in consecutive frames of the multi-frame images are matched, and the matched targets are integrated according to the time series to obtain the motion trajectory of each target. The motion trajectory of each target is used as the target detection result.

[0129] An embodiment of the present application provides a computer program product, including a computer program. When the computer program is executed by a processor, the contents shown in the aforementioned embodiment of the data analysis method based on artificial intelligence are implemented.

[0130] An electronic device is provided in an embodiment of the present application, such as Figure 3 As shown, Figure 3 The electronic device 300 shown includes: a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, such as through a bus 302. Optionally, the electronic device 300 may also include a transceiver 304. It should be noted that in actual applications, the transceiver 304 is not limited to one, and the structure of the electronic device 300 does not constitute a limitation on the embodiments of the present application.

[0131] The processor 301 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. The processor 301 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0132] The bus 302 may include a path to transmit information between the above components. The bus 302 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 302 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3Only one thick line is used in the diagram, but it does not mean that there is only one bus or only one type of bus.

[0133] The memory 303 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0134] The memory 303 is used to store the application code for executing the solution of the present application, and the execution is controlled by the processor 301. The processor 301 is used to execute the application code stored in the memory 303 to implement the contents shown in the above-mentioned embodiment of the data analysis method based on artificial intelligence.

[0135] Figure 3 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0136] An embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer-readable storage medium is run on a computer, the computer can execute the contents shown in the aforementioned embodiment of the artificial intelligence-based data analysis method.

[0137] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0138] The above are only some implementation methods of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A data analysis method based on artificial intelligence, characterized in that: include: Acquire data to be analyzed, classify the data to be analyzed according to data types, and obtain a data set corresponding to each data type; Determine a preset result item corresponding to each data set based on a preset corresponding relationship, where the preset corresponding relationship is a corresponding relationship between a data type and a preset result item; Performing data analysis on the target data set based on the data type of the target data set to obtain a data analysis result including preset result items of the target data set, wherein the target data set is any data set obtained by classifying the data to be analyzed; The step of performing data analysis on the target data set based on the data type of the target data set to obtain a data analysis result containing preset result items of the target data set includes: Constructing a feature matrix of the target data set; Generate a data analysis result of the target data set based on the feature matrix, wherein the preset result items of the data analysis result include a content analysis result and a sentiment analysis result; The content analysis result includes a category analysis result, and the category analysis result of the target data set is generated based on the feature matrix, including: Performing cluster analysis on the feature matrix to obtain multiple clusters; Determine the density and size of each cluster, and calculate the weight value of a target cluster based on the density and size, wherein the target cluster is any one of the multiple clusters; Calculating the similarity between the target cluster and each preset category, and calculating the probability that the target cluster belongs to each preset category based on the weight value of the target cluster and the similarity, and taking the preset category with the largest probability as the target preset category to which the target cluster belongs; Based on the probability of each cluster in the plurality of clusters and the corresponding preset target category, determining the target category to which the target data set belongs, the target category serving as the category analysis result of the target data set; The calculating the weight value of the target cluster based on the density and the size includes: Sorting the plurality of clusters from small to large according to density to obtain a first sequence, and determining the first order of the target cluster in the first sequence; Sorting the plurality of clusters in ascending order of size to obtain a second sequence, and determining a second order of the target cluster in the second sequence; Calculating the sum of the first sequence and the second sequence as the target sequence of the target cluster; The target positions corresponding to the plurality of clusters are normalized to obtain a weight value of each cluster.

2. The artificial intelligence-based data analysis method according to claim 1, characterized in that: The step of constructing a feature matrix of the target data includes: Extracting an initial feature matrix of the target data set; Calculate the similarity between every two feature points in the initial feature matrix to obtain a similarity matrix; Converting the similarity matrix into a local feature matrix, wherein each data in the local feature matrix represents the connection strength between two feature points; A global feature matrix is ​​constructed based on the local feature matrix, and the local feature matrix and the global feature matrix are combined into a feature matrix.

3. The data analysis method based on artificial intelligence according to claim 1, characterized in that: Generating a sentiment analysis result of the target data set based on the feature matrix includes: The feature matrix is ​​input into a preset sentiment analysis model to obtain a sentiment analysis result output by the sentiment analysis model, wherein the sentiment analysis result includes a sentiment label and a sentiment level.

4. The artificial intelligence-based data analysis method according to claim 1, characterized in that: When the data type of the target data set is a video type, the preset result item further includes a target detection result, and the target detection result is obtained by performing data analysis on the target data set, including: Dividing the target data set into multiple frame images with a time sequence; Performing target detection on the multiple frames of images to identify the target in each frame of image; The same targets detected in consecutive frames of the multiple frames are matched, and the matched targets are integrated according to a time series to obtain a motion trajectory of each target, and the motion trajectory of each target is used as the target detection result.

5. A data analysis device based on artificial intelligence, characterized in that: include: A classification module is used to obtain the data to be analyzed, classify the data to be analyzed according to data types, and obtain a data set corresponding to each data type; A determination module, used to determine a preset result item corresponding to each data set based on a preset corresponding relationship, wherein the preset corresponding relationship is a corresponding relationship between a data type and a preset result item; An analysis module, configured to perform data analysis on a target data set based on a data type of the target data set to obtain a data analysis result including preset result items of the target data set, wherein the target data set is any data set obtained by classifying the data to be analyzed; When the analysis module performs data analysis on the target data set based on the data type of the target data set to obtain a data analysis result including preset result items of the target data set, the analysis module is specifically used to: Constructing a feature matrix of the target data set; Generate a data analysis result of the target data set based on the feature matrix, wherein the preset result items of the data analysis result include a content analysis result and a sentiment analysis result; The content analysis result includes a category analysis result, and when the analysis module generates the category analysis result of the target data set based on the feature matrix, the analysis module is specifically used to: Performing cluster analysis on the feature matrix to obtain multiple clusters; Determine the density and size of each cluster, and calculate the weight value of a target cluster based on the density and size, wherein the target cluster is any one of the multiple clusters; Calculating the similarity between the target cluster and each preset category, and calculating the probability that the target cluster belongs to each preset category based on the weight value of the target cluster and the similarity, and taking the preset category with the largest probability as the target preset category to which the target cluster belongs; Based on the probability of each cluster in the plurality of clusters and the corresponding preset target category, determining the target category to which the target data set belongs, the target category serving as the category analysis result of the target data set; When the analysis module calculates the weight value of the target cluster based on the density and the size, it is specifically used to: Sorting the plurality of clusters from small to large according to density to obtain a first sequence, and determining the first order of the target cluster in the first sequence; Sorting the plurality of clusters in ascending order of size to obtain a second sequence, and determining a second order of the target cluster in the second sequence; Calculating the sum of the first sequence and the second sequence as the target sequence of the target cluster; The target positions corresponding to the plurality of clusters are normalized to obtain a weight value of each cluster.

6. An electronic device, characterized in that: include: at least one processor; Memory; At least one application, wherein the at least one application is stored in a memory and configured to be executed by at least one processor, and the at least one application is configured to: execute the artificial intelligence-based data analysis method described in any one of claims 1-4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed in a computer, the computer is caused to execute the artificial intelligence-based data analysis method described in any one of claims 1 to 4.

Images