Multi-modal data storage management method and system, electronic equipment and storage medium

Abstract

The invention relates to a multi-modal database management system. The system comprises a database construction module, a storage module and a query module. The database construction module determines a multi-modal composite key according to data contents and data types of to-be-stored data and constructs a multi-modal database, and the storage module obtains the to-be-stored data, performs coarse-grained classification aggregation and sorting processing on the to-be-stored data based on the multi-modal composite key, and stores the to-be-stored data into the multi-modal database according to a processing result. And the query module performs data query in the multi-modal database based on the multi-modal composite key. By means of the method and device, the problem that the efficiency is low when a database stores and processes various modal data at the same time is solved. The multi-modal data is classified, aggregated, sorted and stored, high performance of data writing is ensured, in addition, data compression is achieved due to continuous storage of the same kind of data, the data storage space is reduced, and the efficiency that a database processes the multi-modal data at the same time is improved.

Description

Technical Field

[0001] The present application relates to the field of data storage, and particularly to a multi-modal data storage method, system, electronic device, and storage medium. Background Art

[0002] With the development of artificial intelligence, Internet of Things, and big data technologies, databases need to meet the requirements of storing and querying multi-modal data (including structured data, time-series data, text data, vector data, etc.).

[0003] However, existing database systems have limitations in processing these data. For example, traditional relational databases (RDBMS) are mainly used to process structured data and are difficult to efficiently process other types of data such as time-series data, text data, and vector data. Time-series databases, text databases, vector databases, etc. adopt non-relational models to model and specifically optimize the data in their respective fields, enabling these databases to efficiently process time-series, text, and vector data respectively. However, they lack generality, that is, each type of database is difficult to efficiently process other types of data and even more difficult to support complex relational queries. In addition, these types of databases have poor compatibility with SQL. This will result in users needing to use multiple databases or related products such as computing platforms simultaneously in actual data management and analysis practices, with high comprehensive costs and low efficiency.

[0004] Currently, no effective solution has been proposed for the problem of low efficiency in simultaneously storing and processing multiple-modal data in databases in related technologies. Summary of the Invention

[0005] Embodiments of the present application provide a multi-modal database management method, system, electronic device, and storage medium to at least solve the problem of low efficiency in simultaneously storing and processing multiple-modal data in databases in related technologies.

[0006] In a first aspect, embodiments of the present application provide a multi-modal database management system, which includes:

[0007] A database construction module, configured to determine a multi-modal composite key according to the data content and data type of the data to be stored, and construct a multi-modal database;

[0008] A storage module, configured to obtain the data to be stored, perform coarse-grained classification aggregation and sorting processing on the data to be stored based on the multi-modal composite key, and store the data to be stored into the multi-modal database according to the processing result;

[0009] A query module, configured to perform data query in the multi-modal database based on the multi-modal composite key.

[0010] In some of these embodiments, the multimodal composite key includes a classification key, and the storage module includes:

[0011] A classification module, configured to perform coarse-grained classification and aggregation on data based on the classification key to obtain a multimodal data sequence;

[0012] A data storage module, configured to store the multimodal data sequence in order in the multimodal database.

[0013] In some of these embodiments, the query module includes:

[0014] A first query module, configured to determine a target multimodal data sequence according to the corresponding value of the classification key of the data to be queried;

[0015] A second query module, configured to further index and determine the target data in the target multimodal data sequence according to the corresponding values of other multimodal keys of the data to be queried, where the other multimodal keys are the keys in the multimodal composite key except the classification key.

[0016] In some of these embodiments, the multimodal composite key includes a classification key and a time sequence key, and the storage module includes:

[0017] A grouping module, configured to divide the data to be stored into different classified data groups based on the classification key, continuously store the classified data groups in an LSM tree, and arrange them in order according to the classification key;

[0018] A sorting module, configured to store the data inside the classified data group based on the time sequence key.

[0019] In some of these embodiments, the sorting module includes:

[0020] A column storage module, configured to store the data inside the classified data group by column based on the time sequence key;

[0021] A splitting module, configured to split each column in the classified data group into several blocks and compress and store them separately.

[0022] In some of these embodiments, the query module includes:

[0023] A first indexing module, configured to determine a target classified data group according to the corresponding value of the classification key of the data to be queried;

[0024] A second indexing module, configured to determine a target block in the target classified data group according to the corresponding value of the time sequence key of the data to be queried.

[0025] In some of these embodiments, the multimodal composite key further includes a text key and a vector key, and the query module further includes:

[0026] A third indexing module, configured to perform an inverted index in the target classified data group according to the text key corresponding value of the data to be queried; and / or

[0027] A fourth indexing module, configured to perform a vector index in the target classified data group according to the vector key corresponding value of the data to be queried.

[0028] In a second aspect, an embodiment of the present application provides a multimodal database management method, the method including:

[0029] Determine a multimodal composite key according to the data content and data type of the data to be stored, and construct a multimodal database;

[0030] Obtain the data to be stored, perform coarse-grained classification aggregation and sorting processing on the data to be stored based on the multimodal composite key, and store the data to be stored into the multimodal database according to the processing result;

[0031] Perform data query in the multimodal database based on the multimodal composite key.

[0032] In a third aspect, an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the computer program, the multimodal database management method described in the second aspect above is implemented.

[0033] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the multimodal database management method described in the second aspect above is implemented.

[0034] Compared with the related art, in the multimodal database management system provided by the embodiment of the present application, a database construction module determines a multimodal composite key according to the data content and data type of the data to be stored, and constructs a multimodal database. A storage module obtains the data to be stored, performs coarse-grained classification aggregation and sorting processing on the data to be stored based on the multimodal composite key, and stores the data to be stored into the multimodal database according to the processing result. A query module performs data query in the multimodal database based on the multimodal composite key, solving the problem of low efficiency in simultaneously storing and processing multiple-modal data in the database. By classifying, aggregating, and sorting the multimodal data for storage, the high performance of data writing is ensured. In addition, the continuous storage of the same type of data enables data compression to be achieved, reducing the data storage space and improving the efficiency of the database in simultaneously processing multiple-modal data. Description of the Drawings

[0035] The accompanying drawings described herein are used to provide a further understanding of the present application and form a part of the present application. The schematic embodiments and descriptions thereof of the present application are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings:

[0036] Figure 1 is a structural block diagram of a multimodal database management system according to an embodiment of the present application;

[0037] Figure 2 is a schematic diagram of data storage based on the LSM tree according to an embodiment of the present application;

[0038] Figure 3 is a schematic diagram of an index structure according to an embodiment of the present application;

[0039] Figure 4 is a flowchart of a multimodal database management method according to an embodiment of the present application;

[0040] Figure 5 is a schematic diagram of the internal structure of an electronic device according to an embodiment of the present application. Detailed implementation manners

[0041] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be described and explained below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments provided in the present application without making creative efforts fall within the scope of protection of the present application.

[0042] Obviously, the accompanying drawings in the following description are only some examples or embodiments of the present application. For those of ordinary skill in the art, without making creative efforts, the present application can also be applied to other similar scenarios based on these drawings. In addition, it can also be understood that although the efforts made in this development process may be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in the present application, some design, manufacturing or production changes based on the technical content disclosed in the present application are only conventional technical means and should not be understood as the content disclosed in the present application being insufficient.

[0043] Referring to "embodiment" in the present application means that the specific features, structures or characteristics described in connection with the embodiment can be included in at least one embodiment of the present application. The phrase appears in various positions in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those of ordinary skill in the art explicitly and implicitly understand that the embodiments described in the present application can be combined with other embodiments without conflict.

[0044] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the ordinary meanings understood by those with ordinary skills in the technical field to which this application belongs. The similar words such as "a", "an", "one kind", "the" involved in this application do not indicate a quantity limitation and may represent singular or plural. The terms "include", "comprise", "have" and any variations thereof involved in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may further include unlisted steps or units, or may further include other steps or units inherent to these processes, methods, products or devices. The similar words such as "connect", "be connected", "couple" involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "plurality" involved in this application means two or more. "And / or" describes the association relationship of associated objects and indicates that three relationships may exist. For example, "A and / or B" may represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0045] This embodiment provides a multimodal database management system, which is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, terms such as "module", "unit", "sub-unit", etc. may be a combination of software and / or hardware that can implement a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0046] Figure 1 is a structural block diagram of the multimodal database management system according to an embodiment of the present application, as Figure 1 shown, the system includes: a database construction module 11, a storage module 12, and a query module 13.

[0047] The database construction module 11 is used to determine a multimodal composite key according to the data content and data type of the data to be stored, and construct a multimodal database.

[0048] In this embodiment, based on the traditional two-dimensional data table and column-store database, a multimodal composite key is introduced. The multimodal composite key is a subset of all columns in the column-store table created by the user and is specified by the user when creating the table.

[0049] Optionally, in this embodiment, the multimodal database is a time series database (TSDB). The multimodal composite key includes, but is not limited to, a classification key, a time series key, a text key, and a vector key. It should be noted that the time series key, the text key, and the vector key are collectively referred to as multimodal keys. The classification key is mandatory and the multimodal keys can be set according to user selection.

[0050] The classification key is a key used to perform coarse-grained classification and aggregation on data. It is specified by the user and is mandatory. The classification key can correspond to the column definitions of one or more columns in a traditional columnar database. In the time series data scenario, the classification key can be the device ID of the Internet of Things, the stock code in the financial field, etc.; in the text data scenario, the classification key can be the ID of a certain creator.

[0051] The time series key is a key used to sort and index time series data. It can be set according to user selection. The time series key is usually a time-type column in a traditional columnar database, representing the occurrence time of a piece of data.

[0052] The text key is a key used to index text data. It can be set according to user selection. The text key is usually a text column that needs to be text-indexed in a traditional columnar database.

[0053] The vector key is a key used to index vector data. It can be set according to user selection. The vector key is usually a column that needs to be vector-indexed in a traditional columnar database.

[0054] The storage module 12 is used to obtain the data to be stored, perform coarse-grained classification, aggregation, and sorting processing on the data to be stored based on the multimodal composite key, and store the data to be stored in the multimodal database according to the processing result.

[0055] In some of these embodiments, the multimodal composite key includes a classification key, and the storage module 12 includes:

[0056] The classification module is used to perform coarse-grained classification and aggregation on the data based on the classification key to obtain a multimodal data sequence.

[0057] After the data is coarsely classified and aggregated through the classification key and then stored, this enables the compact storage and continuous writing of multimodal data, overcoming the problems of poor writing performance and inability to be efficiently compressed caused by the scattered storage of multimodal data in the traditional columnar-based relational model.

[0058] The data storage module is used to store the multimodal data sequence in order in the multimodal database.

[0059] In the multimodal composite key model of this embodiment, multimodal data is aggregated and stored according to classification keys, which ensures high performance in data writing and low space occupancy in data storage. In addition, to accelerate indexing, the classification keys are stored in sequential order.

[0060] The query module 13 is used to query data in the multimodal database based on the multimodal composite key.

[0061] In some of the embodiments, the query module 13 includes:

[0062] The first query module is used to determine the target multimodal data sequence according to the corresponding value of the classification key of the data to be queried.

[0063] Given a classification value, a continuous multimodal data sequence can be quickly found, which provides a good logical basis for the vectorization operation and batch aggregation of multimodal data. It also enables the TSDB using this model to perform high-performance analysis and query on time-series data.

[0064] The second query module is used to further index and determine the target data in the target multimodal data sequence according to the corresponding values of other multimodal keys of the data to be queried, where the other multimodal keys are the keys in the multimodal composite key except the classification key.

[0065] In the query scenario, a continuous and compactly stored multimodal data sequence can be quickly retrieved through the classification key, and then a piece of data can be determined by indexing inside the multimodal data sequence through the multimodal key.

[0066] Through the above system, the database construction module 11 determines the multimodal composite key according to the data content and data type of the data to be stored, constructs the multimodal database, the storage module 12 obtains the data to be stored, performs coarse-grained classification aggregation and sorting processing on the data to be stored based on the multimodal composite key, and stores the data to be stored into the multimodal database according to the processing result. The query module 13 queries data in the multimodal database based on the multimodal composite key, solving the problem of low efficiency in the database for simultaneously storing and processing multiple-modal data. By classifying, aggregating, and sorting the storage of multimodal data, high performance in data writing is ensured. In addition, the continuous storage of similar data also enables data compression to be achieved, reducing the data storage space and improving the efficiency of the database for simultaneously processing multiple-modal data.

[0067] By setting the multimodal composite key, the problems that the current database is not good at efficiently processing and storing structured data, time-series data, text data, and vector data simultaneously, and that the current database cannot support multimodal and complex relationship queries simultaneously are solved.

[0068] The multi-modal composite key model can utilize the powerful query analysis capabilities in the relational model to implement complex query logics, such as multi-table association, nested queries, etc. Generally speaking, the multi-modal composite key model reuses the advantages of the relational model in complex queries, which also enables the TSDB to fully support standard SQL compatibility.

[0069] In some of these embodiments, the multi-modal composite key includes a classification key and a time series key, and the storage module 12 includes:

[0070] A grouping module, configured to divide the data to be stored into different classified data groups based on the classification key, continuously store the classified data groups in the LSM tree, and arrange them in an orderly manner according to the classification key;

[0071] A sorting module, configured to store the data inside the classified data group based on the time series key.

[0072] In this embodiment, the above data storage format based on the multi-modal composite key model is implemented through the storage architecture of the LSM tree (LSM-Tree).

[0073] Figure 2 It is a schematic diagram of data storage based on the LSM tree according to an embodiment of the present application. As Figure 2 shown, the multi-modal composite key model requires that multi-modal data stored continuously can be located through the classification key. Therefore, this storage structure organizes and stores the data centered around the classification key, that is, the data is divided into different classified data groups according to the classification key, and the data within the same group is stored continuously. In addition, to ensure the indexing performance of the classification key value, the classified data groups are continuously stored in the storage unit (Level File) and arranged in an orderly manner according to the classification key.

[0074] In some of these embodiments, the sorting module includes:

[0075] A column storage module, configured to store the data inside the classified data group column by column based on the time series key;

[0076] A splitting module, configured to split each column in the classified data group into several blocks and compress and store them separately.

[0077] The storage mode centered around the classification key ensures that data with the same classification key is aggregated by row in the classified data group. Continuing to refer to Figure 2 , inside the classified data group, to ensure the retrieval performance of the data, the data is sorted according to the time series key in the multi-modal composite key. To ensure high-performance query and compact storage, the data is stored in a columnar storage manner, that is, the data of the first column is stored continuously first, and then the data of the second column is stored continuously. In addition, for the convenience of compression and indexing, when storing the data in a columnar storage manner, a single column is split into several blocks (block) and compressed and stored separately.

[0078] In some of these embodiments, the query module 13 includes:

[0079] A first index module for determining a target classified data group according to the corresponding value of the classification key of the data to be queried.

[0080] A second index module for determining a target block in the target classified data group according to the corresponding value of the time series key of the data to be queried.

[0081] Based on the above data organization method, the present embodiment designs a corresponding index structure. Figure 3 It is a schematic diagram of an index structure according to an embodiment of the present application. As Figure 3 shown, corresponding to the centralized data aggregation mode based on the classification key, the index is organized around the classification key. One index entry corresponds to a classified data group in the data area. Therefore, when retrieving, given the value of a classification key, and then traversing the index entries can efficiently find the corresponding classified data group.

[0082] The index entry is responsible for indexing the data inside the classified data group. The granularity of the index is the block. By means such as a zone map or recording the smallest time series key value in each block, high-performance retrieval of the block is achieved.

[0083] In some of these embodiments, the multi-modal composite key further includes a text key and a vector key. The query module 13 further includes:

[0084] A third index module for performing an inverted index in the target classified data group according to the corresponding value of the text key of the data to be queried.

[0085] A fourth index module for performing a vector index in the target classified data group according to the corresponding value of the vector key of the data to be queried.

[0086] In addition, an inverted index and a vector index are also stored in the storage unit for further retrieving the multi-modal data using the inverted index or the vector index after locating a certain partition data group through the classification key.

[0087] Based on aggregating the data around the classification key, through the row-column hybrid storage mode, the data can be efficiently compressed, and the TSDB can also perform complex query optimization and vectorized execution, further improving the query performance of the TSDB for multi-modal data.

[0088] It should be noted that the above-mentioned various modules can be functional modules or program modules, and can be implemented either by software or by hardware. For the modules implemented by hardware, the above-mentioned various modules can be located in the same processor; or the above-mentioned various modules can also be located in different processors in any combination form.

[0089] This embodiment provides a multimodal database management method. Figure 2 It is a flowchart of the multimodal database management method according to an embodiment of the present application, as Figure 2 shown, and this process includes the following steps:

[0090] Step S201: Determine a multimodal composite key according to the data content and data type of the data to be stored, and construct a multimodal database.

[0091] Step S202: Obtain the data to be stored, perform coarse-grained classification aggregation and sorting processing on the data to be stored based on the multimodal composite key, and store the data to be stored into the multimodal database according to the processing result.

[0092] Step S203: Perform data query in the multimodal database based on the multimodal composite key.

[0093] In some embodiments, the multimodal composite key includes a classification key, and step S202 includes:

[0094] Step S2021: Perform coarse-grained classification aggregation on the data based on the classification key to obtain a multimodal data sequence.

[0095] Step S2022: Store the multimodal data sequence in the multimodal database in an ordered manner.

[0096] In some embodiments, step S203 includes:

[0097] Step S2031: Determine the target multimodal data sequence according to the corresponding value of the classification key of the data to be queried.

[0098] Step S2032: Further index and determine the target data in the target multimodal data sequence according to the corresponding values of other multimodal keys of the data to be queried, where the other multimodal keys are the keys in the multimodal composite key except the classification key.

[0099] In some embodiments, the multimodal composite key includes a classification key and a time sequence key, and step S202 includes:

[0100] Step S301: Divide the data to be stored into different classification data groups based on the classification key, continuously store the classification data groups in the LSM tree, and arrange them in an orderly manner according to the classification key.

[0101] Step S302: Store the data inside the classification data group based on the time sequence key.

[0102] In some embodiments, step S302 includes:

[0103] Step S3021: Store the data inside the classification data group column by column based on the time sequence key.

[0104] Step S3022: Split each column in the classified data group into several blocks and store them separately in a compressed manner.

[0105] In some embodiments, step S203 includes:

[0106] Step S401: Determine the target classified data group according to the corresponding value of the classification key of the data to be queried.

[0107] Step S402: Determine the target block in the target classified data group according to the corresponding value of the time sequence key of the data to be queried.

[0108] In some embodiments, the multimodal composite key further includes a text key and a vector key, and step S203 further includes:

[0109] Step S403: Perform an inverted index in the target classified data group according to the corresponding value of the text key of the data to be queried.

[0110] Step S404: Perform a vector index in the target classified data group according to the corresponding value of the vector key of the data to be queried.

[0111] Through the above steps, determine the multimodal composite key according to the data content and data type of the data to be stored, construct a multimodal database, obtain the data to be stored, perform coarse-grained classification aggregation and sorting processing on the data to be stored based on the multimodal composite key, store the data to be stored into the multimodal database according to the processing result, and perform data query in the multimodal database based on the multimodal composite key, which solves the problem of low efficiency in simultaneously storing and processing multiple modal data in the database. By classifying, aggregating and sorting the multimodal data for storage, the high performance of data writing is ensured. In addition, continuous storage of the same type of data also enables data compression, reduces the data storage space, and improves the efficiency of the database in simultaneously processing multiple modal data.

[0112] It should be noted that the steps shown in the above process or the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0113] This embodiment also provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0114] Optionally, the above electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0115] Optionally, in this embodiment, the above-mentioned processor may be configured to execute the following steps through a computer program:

[0116] S1. Determine a multimodal composite key according to the data content and data type of the data to be stored, and construct a multimodal database.

[0117] S2. Obtain the data to be stored, perform coarse-grained classification aggregation and sorting processing on the data to be stored based on the multimodal composite key, and store the data to be stored into the multimodal database according to the processing result.

[0118] S3. Perform data query in the multimodal database based on the multimodal composite key.

[0119] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiment and optional implementation manners, and will not be elaborated herein.

[0120] In one embodiment, Figure 5 is a schematic internal structure diagram of an electronic device according to an embodiment of the present application. As Figure 5 shown, an electronic device is provided. The electronic device may be a server, and its internal structure diagram may be as Figure 5 shown. The electronic device includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the electronic device is used to store data. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a multimodal database management method.

[0121] Those skilled in the art can understand that Figure 5 the structure shown in

[0122] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0123] Those skilled in the art should understand that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0124] The above embodiments only represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A multimodal database management system, characterized in that: The system comprises: A database construction module, used to determine a multimodal composite key according to the data content and data type of the data to be stored, and to construct a multimodal database; A storage module, used for acquiring the data to be stored, performing coarse-grained classification, aggregation and sorting processing on the data to be stored based on the multimodal composite key, and storing the data to be stored in the multimodal database according to the processing result; A query module is used to perform data query in the multimodal database based on the multimodal composite key.

2. The system according to claim 1, characterized in that The multimodal composite key includes a classification key, and the storage module includes: A classification module, used for performing coarse-grained classification and aggregation on the data based on the classification key to obtain a multimodal data sequence; A data storage module is used to store the multimodal data sequence in order in the multimodal database.

3. The system according to claim 2, characterized in that The query module includes: A first query module, used to determine a target multimodal data sequence according to a corresponding value of a classification key of the data to be queried; The second query module is used to further index and determine the target data in the target multimodal data sequence according to the corresponding values ​​of other multimodal keys of the data to be queried, wherein the other multimodal keys are the keys in the multimodal composite key except the classification key.

4. The system according to claim 1, characterized in that The multimodal composite key includes a classification key and a time sequence key, and the storage module includes: A grouping module, used for dividing the data to be stored into different classified data groups based on the classification key, storing the classified data groups continuously in the LSM tree, and arranging them in order according to the classification key; A sorting module is used to store the data within the classified data group based on the time sequence key.

5. The system according to claim 4, characterized in that The sorting module comprises: A column storage module, used for storing the data in the classified data group by column based on the time sequence key; The segmentation module is used to segment each column in the classified data group into a number of blocks for compression and storage respectively.

6. The system according to claim 5, characterized in that The query module includes: A first indexing module, used to determine a target classification data group according to a classification key corresponding value of the data to be queried; The second indexing module is used to determine a target block in the target classification data group according to a corresponding value of a time sequence key of the data to be queried.

7. The system according to claim 6, characterized in that The multimodal composite key further includes a text key and a vector key, and the query module further includes: A third indexing module is used to perform an inverted index in the target classification data group according to the text key corresponding value of the data to be queried; and / or The fourth indexing module is used to perform vector indexing in the target classified data group according to the vector key corresponding value of the data to be queried.

8. A multimodal database management method, characterized in that: The method comprises: Determine a multimodal composite key according to the data content and data type of the data to be stored, and build a multimodal database; Acquire the data to be stored, perform coarse-grained classification, aggregation and sorting processing on the data to be stored based on the multimodal composite key, and store the data to be stored in the multimodal database according to the processing result; Data query is performed in the multimodal database based on the multimodal composite key.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the multimodal database management method according to claim 8 is implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the multimodal database management method as claimed in claim 8 is implemented.

Images