CN117539974A - Semantic indexing method, system, device and storage medium according to product route - Google Patents
Semantic indexing method, system, device and storage medium according to product route Download PDFInfo
- Publication number
- CN117539974A CN117539974A CN202311690804.4A CN202311690804A CN117539974A CN 117539974 A CN117539974 A CN 117539974A CN 202311690804 A CN202311690804 A CN 202311690804A CN 117539974 A CN117539974 A CN 117539974A
- Authority
- CN
- China
- Prior art keywords
- search
- service
- retrieval
- product
- index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 68
- 238000013507 mapping Methods 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 25
- 230000015654 memory Effects 0.000 claims abstract description 20
- 239000013598 vector Substances 0.000 claims description 27
- 238000010276 construction Methods 0.000 claims description 19
- 238000006243 chemical reaction Methods 0.000 claims description 13
- 230000003993 interaction Effects 0.000 claims description 11
- 238000004140 cleaning Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000005406 washing Methods 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 18
- 230000006870 function Effects 0.000 description 13
- 230000008569 process Effects 0.000 description 9
- 101100264195 Caenorhabditis elegans app-1 gene Proteins 0.000 description 5
- 230000008901 benefit Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000006467 substitution reaction Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 101100055496 Arabidopsis thaliana APP2 gene Proteins 0.000 description 2
- 101100016250 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) GYL1 gene Proteins 0.000 description 2
- 238000003491 array Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 101150053844 APP1 gene Proteins 0.000 description 1
- 101100189105 Homo sapiens PABPC4 gene Proteins 0.000 description 1
- 102100039424 Polyadenylate-binding protein 4 Human genes 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000008602 contraction Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/31—Indexing; Data structures therefor; Storage structures
- G06F16/316—Indexing structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/38—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/383—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Library & Information Science (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses a semantic indexing method, a semantic indexing system, a semantic indexing device and a semantic indexing storage medium according to product routing, wherein the method comprises the steps of obtaining a retrieval request for accessing a retrieval service; the search request comprises product identity information of a product to be accessed; determining a first instance of a search service according to the product identity information and the mapping table information, and sending the search request to the search service so that the search service obtains a search result according to the search request and the first instance; the first instance is used for representing a plurality of retrieval service instances in which index files are loaded, which correspond to the retrieval request, in the retrieval service. When searching, part of search service examples only need to load the index file corresponding to the part request, so that the memory overhead can be reduced, the resource utilization rate can be improved, the machine cost can be reduced, and the service performance and stability can be improved; the method and the device can be widely applied to the technical field of data processing.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a semantic indexing method, a semantic indexing system, a semantic indexing device and a semantic indexing storage medium according to product routing.
Background
In the AI field, content retrieval services typically rely on various forms of indexing, where Semantic indexing (also known as Semantic vector retrieval) is favored for its high accuracy and flexibility, which mainly includes Semantic tree indexing as well as Semantic bitmap indexing. The semantic tree index is mainly used for calculating feature codes of sentences corresponding to the words according to the arrangement sequence of the words in the sentences, and a chain type storage technology is utilized, so that an indexing method for realizing the corresponding semantic tree is obtained. And bitmap indexing is an indexing method using bit bits. For a traditional database table, it is mainly an indexing method created for a large number of columns of the same value. Both indexing methods occupy larger memory space, are unfavorable for popularization and use of the indexing method in a retrieval service center, and bring larger machine cost pressure under the condition of more corpus data. Meanwhile, the content retrieval center often serves a plurality of products, and a large burden is caused to a machine when semantic indexes of the plurality of products are simultaneously constructed, loaded and unloaded, so that the construction/updating of the indexes is long, the index management difficulty is high, and finally, the service performance is low and the stability is poor.
In view of the above, there is a need to solve the problems of the related art.
Disclosure of Invention
The present application aims to solve at least one of the technical problems existing in the related art to a certain extent.
Therefore, an object of the embodiments of the present application is to provide a semantic indexing method according to product routing, which can reduce memory overhead and improve resource utilization.
In order to achieve the technical purpose, the technical scheme adopted by the embodiment of the application comprises the following steps:
a semantic indexing method according to product routing, comprising: acquiring a retrieval request for accessing a retrieval service; the search request comprises product identity information of a product to be accessed; determining a first instance of a search service according to the product identity information and the mapping table information, and sending the search request to the search service so that the search service obtains a search result according to the search request and the first instance; the first instance is used for representing a plurality of retrieval service instances in which index files are loaded, which correspond to the retrieval request, in the retrieval service.
In addition, the semantic indexing method according to the product routing according to the above embodiment of the present application may further have the following additional technical features:
Further, in one embodiment of the present application, the step of sending a search request to the search service to cause the search service to obtain a search result according to the search request and the first instance includes: determining a plurality of retrieval service examples from an example list set in the retrieval service according to the product identity information and the mapping table information; the instance list set is used for representing all retrieval service instances in the retrieval service, wherein index files are loaded in the retrieval service instances; and sending the search request to the search service according to a product balancing strategy, so that the search service obtains a search result according to the search request and the plurality of search service instances.
Further, in one embodiment of the present application, the set of instance lists is obtained by: responding to a loading index notification sent by an index construction center, determining a plurality of second examples corresponding to each product identity information from a man-machine interaction platform and sending the second examples to the retrieval service, so that the retrieval service loads a first index to the second examples from the index construction center to obtain an example list set; the first index is constructed when the index construction center station sends out the loading index notification.
Further, in one embodiment of the present application, the index building middle stage builds the first index by: responding to an index construction instruction sent by a man-machine interaction platform, and inquiring related corpus of each product; washing vectorization is carried out on the related corpus to obtain a corpus vector set; and determining a first index of each product according to the corpus vector set and an annoy algorithm.
Further, in an embodiment of the present application, the step of performing washing vectorization on the relevant corpus to obtain a corpus vector includes: preprocessing the related corpus to obtain a first corpus; the preprocessing comprises at least one of case conversion, complex-form conversion or full-half angle conversion; correcting the first corpus and segmenting the first corpus to obtain a plurality of words; and inputting the words into a vectorization model to obtain a plurality of corpus vectors.
Further, in one embodiment of the present application, the step of determining the first instance of the search service according to the product identity information and the mapping table information includes: identifying a first keyword of the product identity information and an input keyword of an input side of the mapping table information; comparing the first keyword with the input keyword, and determining a plurality of retrieval service instances of which the first keyword is identical to the input keyword as a first instance.
Further, in one embodiment of the present application, the step of sending the search request to the search service according to a product balancing policy, where the search service obtains a search result according to the search request and the plurality of search service instances includes: according to a polling product balancing strategy, sequentially sending the plurality of sub-requests to the retrieval service, so that the retrieval service obtains a retrieval result according to the sub-requests and the plurality of retrieval service instances; or sending the plurality of sub-requests to the retrieval service according to a hash product balancing strategy, so that the retrieval service obtains a retrieval result according to the sub-requests and the plurality of retrieval service instances; or according to a random product balancing strategy, the plurality of sub-requests are randomly sent to the retrieval service, so that the retrieval service obtains a retrieval result according to the sub-requests and the plurality of retrieval service instances.
In another aspect, an embodiment of the present application provides a semantic indexing device according to product routing, including:
The acquisition module is used for acquiring a retrieval request for accessing the retrieval service; the search request comprises product identity information of a product to be accessed;
the processing module is used for determining a first instance of a search service according to the product identity information and the mapping table information, and sending the search request to the search service so that the search service obtains a search result according to the search request and the first instance;
the first instance is used for representing a plurality of retrieval service instances in which index files are loaded, which correspond to the retrieval request, in the retrieval service.
In another aspect, embodiments of the present application provide a computer device, including:
at least one processor;
at least one memory for storing at least one program;
the at least one program, when executed by the at least one processor, causes the at least one processor to implement the semantic indexing method according to product routing described above.
In another aspect, embodiments of the present application further provide a computer readable storage medium having stored therein a processor executable program, which when executed by a processor is configured to implement the above-described semantic indexing method according to product routing.
The advantages and benefits of the present application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the present application.
According to the semantic indexing method based on the product routing, the retrieval service instances of a plurality of index files loaded are determined according to the product identity information and mapping table information by acquiring a retrieval request with the product identity information for accessing the retrieval service. And the retrieval service instance loaded with the index file is sent to the retrieval service, so that the retrieval service can obtain a retrieval result corresponding to the retrieval request according to the retrieval request and a plurality of retrieval service instances loaded with the index file. The method can enable partial retrieval service examples to only load the index files corresponding to partial requests when retrieval is carried out, can reduce memory expenditure, can improve resource utilization rate, can reduce machine cost, and can improve service performance and stability.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following description is made with reference to the accompanying drawings of the embodiments of the present application or the related technical solutions in the prior art, it should be understood that, in the following description, the drawings are only for convenience and clarity to describe some embodiments in the technical solutions of the present application, and other drawings may be obtained according to these drawings without any inventive effort for those skilled in the art.
FIG. 1 is a schematic diagram of steps of a semantic indexing method according to product routing according to an embodiment of the present application;
FIG. 2 is a schematic diagram of steps for sending a search request to a search service, so that the search service obtains a search result according to the search request and a first instance;
FIG. 3 is a schematic diagram of mapping table information provided in an embodiment of the present application;
FIG. 4 is a schematic diagram of a step of constructing a first index by an index constructing middle stage according to an embodiment of the present application;
fig. 5 is a schematic diagram of a step of performing cleaning vectorization on a related corpus to obtain a corpus vector in the embodiment of the present application;
FIG. 6 is a schematic diagram of steps for determining a first instance of a search service according to product identity information and mapping table information according to an embodiment of the present application;
FIG. 7 is a schematic diagram of a semantic index retrieval service workflow according to product routing provided in an embodiment of the present application;
FIG. 8 is a schematic flow chart of a search service receiving request according to an embodiment of the present application;
FIG. 9 is a schematic structural diagram of a semantic indexing device according to product routing according to an embodiment of the present application;
Fig. 10 is a schematic structural diagram of a computer device of a semantic indexing system according to product routing according to an embodiment of the present application.
Detailed Description
The present application is further described below with reference to the drawings and specific examples. The described embodiments should not be construed as limitations on the present application, and all other embodiments, which may be made by those of ordinary skill in the art without the exercise of inventive faculty, are intended to be within the scope of the present application.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is to be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with one another without conflict.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the present application.
Before further describing embodiments of the present application in detail, the terms and expressions that are referred to in the embodiments of the present application are described, and are suitable for the following explanation.
Tree index: the tree index may mainly include a B tree index, a B-tree index, a b+ tree index, and a B x tree index. The index, like the table, also belongs to one of the segments (segments). The user data is stored in the storage device, and the storage device is as same as a watch and occupies disk space. However, the data storage form in the index is very different from the data storage form in the exterior. In understanding the index, one can imagine a book where the contents of the book correspond to the data of the table and the directory in front of the book corresponds to the index of the table. Meanwhile, under the general condition, the disk space occupied by the index is much smaller than that of the table, and the main function of the index is to speed up the searching speed of the data and also to ensure the uniqueness of the data.
However, the index is an optional data structure, and the user may or may not select to create the index for a certain table. This is because once the index is created, it means that additional effort (i.e., maintenance of the index structure) and overhead in terms of storage must be handled when the oracle pair performs DML (including INSERT, UPDATE, DELETE). The improvement in query performance that results from creating the index needs to be considered when creating the index, whether it is worth the additional overhead that is incurred.
Physically, an index can be generally divided into: partition and non-partition indexes, regular B-tree indexes, bitmap indexes, flip indexes, etc. Among them, the B-tree index belongs to the most common index. The B-tree index is a typical tree structure that contains mainly the components:
1) Leaf node): the containing entries point directly to rows of data in the table.
2) Branch node (Branch node): the entries contained point to other branch nodes or leaf nodes in the index.
3) Root node: a B-tree index has only one root node, which is actually the branch node at the top of the tree.
In the related art, in the AI field, a content retrieval service generally relies on various forms of indexes, wherein Semantic indexes (also called Semantic vector indexes) are favored due to their higher accuracy and flexibility, and the Semantic indexes mainly include Semantic tree indexes and Semantic bitmap indexes. The semantic tree index is mainly used for calculating feature codes of sentences corresponding to the words according to the arrangement sequence of the words in the sentences, and a chain type storage technology is utilized, so that an indexing method for realizing the corresponding semantic tree is obtained. And bitmap indexing is an indexing method using bit bits. For a traditional database table, it is mainly an indexing method created for a large number of columns of the same value. Both indexing methods occupy larger memory space, are unfavorable for popularization and use of the indexing method in a retrieval service center, and bring larger machine cost pressure under the condition of more corpus data. Meanwhile, the content retrieval center often serves a plurality of products, and a large burden is caused to a machine when semantic indexes of the plurality of products are simultaneously constructed, loaded and unloaded, so that the construction/updating of the indexes is long, the index management difficulty is high, and finally, the service performance is low and the stability is poor.
In order to solve the problems of long time consumption for constructing/updating indexes, high index management difficulty, low service performance and poor stability caused by the fact that a large load is caused on a machine when semantic indexes of a plurality of products in a retrieval center are simultaneously constructed, loaded and unloaded in the related technology, the application provides a semantic indexing method according to product routing.
Referring to fig. 1, fig. 1 is a schematic diagram of a semantic indexing method according to a product route according to an embodiment of the present application, where a search request with product identity information for accessing a search service may be obtained, and a plurality of search service instances loaded with an index file may be determined according to the product identity information and mapping table information. And the retrieval service instance loaded with the index file is sent to the retrieval service, so that the retrieval service can obtain a retrieval result corresponding to the retrieval request according to the retrieval request and a plurality of retrieval service instances loaded with the index file. The method can enable partial retrieval service examples to only load the index files corresponding to partial requests when retrieval is carried out, can reduce memory expenditure, can improve resource utilization rate, can reduce machine cost, and can improve service performance and stability. The method can perfect the whole flow from construction to loading to providing service of the semantic index. The method can improve the problems that the semantic indexes occupy high memory, a plurality of semantic indexes are difficult to maintain and service is difficult to popularize and use in the retrieval process, can reduce the time consumption of the construction of the semantic indexes, can improve the retrieval performance of the middle stage, can improve the flexibility and the stability of the whole index flow, and can enhance the maintainability of the architecture. The method may include, but is not limited to, steps S101-S102.
S101, acquiring a retrieval request for accessing a retrieval service; the retrieval request includes product identity information for the product to be accessed.
It is understood that in the embodiment of the present application, the search service may be a micro service that is independently provided in other hardware housing devices of the present application. The retrieval request may come from any one of the hardware devices or servers connected to the retrieval service through the gateway. The number of search requests may be one or more, and when the number of search requests is plural, the processor may process the plural search requests in order. The search request may include product identity information of the product to be accessed and search key information carried by the request. According to the keyword information, the retrieval service can obtain a retrieval result corresponding to the keyword information. The product to be accessed can be a certain application in the search service, a certain search engine, and the like.
In some optional embodiments of the present application, the gateway may obtain, from a server or a hardware device connected to the gateway, a search request for accessing a search service, where the search request may carry product identity information of a product to be accessed. After the gateway itself has processed the data, it can determine the instance of executing the search service and send the search request to the search service.
It will be appreciated that in some alternative embodiments of the present application, the gateway may obtain a retrieval request to access the retrieval service via an internal data processing module or other processing chip integrated processing circuitry. The processing circuit and the data processing module can be connected with any hardware device or server for sending the retrieval request, and the retrieval request for accessing the retrieval service can be obtained through wired connection or wireless connection. It should be noted that the limited connection manner may include a connection between a mobile device and a processing module, and may also include a connection between a processing module and a hardware device, and a wired connection between other now known or later developed devices and a processing module; the wireless connection may include, but is not limited to, 3G/4G/5G connection, wiFi connection, bluetooth connection, wiMAX connection, zigbee connection, UWB (Ultra Wide Band) connection, and other now known or later developed wireless connection.
S102, determining a first instance of a search service according to product identity information and mapping table information, and sending a search request to the search service so that the search service obtains a search result according to the search request and the first instance; the first instance is used for representing a plurality of retrieval service instances which are loaded with index files and correspond to the retrieval request in the retrieval service.
It will be appreciated that in some embodiments of the present application, the mapping table information may be preset in the processor of the gateway. The mapping table information may be a table of information in which one instance of the identity product information and the search service are mapped to each other, or a table of information in which two or more instances of the identity product information and the search service are mapped to each other. And the first instance may be one or more instances. And since the search request can be one or more, the search result can be one or more correspondingly.
In some possible embodiments of the application, the gateway may determine one or more instances of the search service according to the received product identity information and mapping table information preset in the processor, and the gateway may send the search request to the search service at the same time, so that the search service may obtain a search result according to the search request and the first instance; the first instance may be used to characterize a search service and a plurality of search service instances in which the index file is loaded corresponding to the search request.
Illustratively, the search request of the present application is taken as 1 example, and the search request may carry product identity information a, where the product identity information a may correspond to 3 search service instances that have already loaded with index files. When 1 search request for accessing the search service is received, the request carries product identity information A, according to the product identity information A, the corresponding 3 search service instances with index files loaded in the search service can be determined, the gateway can send the request to one instance in the 3 search service instances with index files loaded according to the algorithm of the gateway, and the search result corresponding to the search request is finally obtained through the product instance and the search request.
Further, referring to fig. 2, fig. 2 is a schematic diagram of steps for sending a search request to a search service, so that the search service obtains a search result according to the search request and the first instance in the embodiment of the present application. In fig. 2, the step of sending the search request to the search service in step S102 to enable the search service to obtain the search result according to the search request and the first instance may specifically include, but is not limited to, step S201 to step S202.
S201, determining a plurality of retrieval service examples from an example list set in the retrieval service according to product identity information and mapping table information; the instance list set is used for representing all retrieval service instances in the retrieval service, wherein the retrieval service instances are loaded with the index file.
It will be appreciated that the instance list set may be all of the retrieval service instances in the retrieval service that have loaded the index file. And the plurality of search service instances may be any number of search service instances of all the search service instances that have loaded the index file.
In some embodiments of the present application, the gateway determines, according to the product identity information and the mapping table information in the search request, a plurality of search service instances from all search service instances in the search services, on which the index file has been loaded, where the instances may determine a final search result with the search request.
Illustratively, referring to fig. 3, fig. 3 is a schematic diagram of mapping table information. It can be seen from fig. 3 that one product identity information APP1 may correspond to two search service instances being engine-instance1 and engine-instance2, respectively, and another product identity information APP2 may correspond to three search service instances being engine-instance2, engine-instance3, and engine-instance4, respectively.
It will be appreciated that the search service instances corresponding to each product may be partially the same or completely different, for example, the product identity information APP2 may correspond to 3 search service instances, namely, engine-instance1, engine-instance2, and engine-instance4, and may also be engine-instance 5, engine-instance 6, and engine-instance 7.
S202, according to the product balancing strategy, a search request is sent to a search service, so that the search service obtains a search result according to the search request and a plurality of search service instances.
It can be appreciated that the product balancing policy of the present application may be preset in the processor of the gateway. The purpose of the product balancing policy is to send search requests to each different search service instance of the search service in a balanced manner such that each search request corresponds to each search service instance.
In some possible embodiments of the present application, the gateway sends the search request to the search service according to a preset product balancing policy, so that each search service instance in the plurality of search service instances may correspond to one or more different search requests. Thereby, the search service obtains different search results according to different search requests.
Further, in some possible embodiments of the present application, the step of obtaining the instance list set may include, but is not limited to, step S301.
S301, determining a plurality of second examples corresponding to each product identity information from a man-machine interaction platform in response to a loading index notification sent by an index construction center, and sending the second examples to a retrieval service, so that the retrieval service loads a first index to the second examples from the index construction center to obtain an example list set; the first index is constructed when the index construction center station sends out a loading index notification.
It is understood that the first index may be an index corresponding to different products determined by different product corpora. The second instance may be any one or more product instances that do not load the first index.
In some possible embodiments of the present application, the gateway receives the loading index notification sent by the index building center, and may determine, from the man-machine interaction platform, a plurality of instances of unloaded indexes corresponding to each product identity information, and send the instances to the retrieval service. After the retrieval service receives a plurality of instances of the unloaded index, the first index can be loaded to a plurality of second instances from the index building center table, and after the operation is repeated for a plurality of times, all instance list sets can be obtained.
It should be noted that, the first index of the present application may be constructed when the index middle stage issues a notification of loading the index. And after the instance list set is obtained, the index building middle station can update the state of each instance and perform external service through the gateway.
Further, referring to fig. 4, fig. 4 is a schematic diagram of a step of constructing a first index by a middle stage of index construction. In fig. 4, the index building center may build the first index through, but not limited to, steps S401-S403.
S401, responding to an index construction instruction sent by the man-machine interaction platform, and inquiring related corpus of each product.
It can be understood that the man-machine interaction platform can establish connection with the index construction center after receiving the trigger condition, and the man-machine interaction platform can send the index construction instruction to the index construction center after receiving the trigger condition, wherein the trigger condition can be product addition, product identity information change or other various trigger conditions capable of triggering the man-machine interaction platform to send the index construction instruction to the index construction center.
In some possible embodiments of the present application, in response to an index building instruction sent by the man-machine interaction platform, the index building middle stage may query a relevant corpus of each product, where the relevant corpus may be a field describing product relevance.
It should be noted that, the related corpus of the product may be preset in the database of the index building center, and the related corpus of the product may be obtained by calling the database to query in the prior art, or may be a product corpus description of the product itself.
S402, cleaning vectorization is carried out on the related corpus to obtain a corpus vector set.
It can be understood that the corpus vector set may be a corpus vector set formed by a plurality of words obtained by cleaning the related corpus, and vectors corresponding to each word.
In some possible embodiments of the present application, the index building middle stage may wash and vectorize the related corpus to obtain vectors of a plurality of words corresponding to the related corpus. Any two words can be different, and a plurality of words can form a complete related corpus.
Illustratively, after the corpus is divided into several words, the words may be vectorized. The vectorization of the words is a numeric process of the words, the numeric of the words can be vectorized by using the existing model, for example, the related corpus can be divided into a plurality of words which are [ "in", "social security", "reimbursement", "middle" ], and decimal values of [0.0025752163,0.8979353,0.97538406,4.2164777E-4] can be obtained after vectorization of the words, and the decimal values can be used as a corpus vector set.
S403, determining a first index of each product according to the corpus vector set and the annoy algorithm.
In some possible embodiments of the invention, the annoy algorithm is a similar nearest neighbor algorithm in the prior art. The algorithm can be implemented by establishing a plurality of binary trees, each binary tree being randomly segmented; the query method comprises the following steps: inserting the root node of each tree into a priority queue; searching each binary tree in the priority queue, wherein each binary tree can obtain a corpus vector set with the maximum Top K; deleting the repeated corpus vector set; calculating the similarity or distance between each vector in the corpus vector set and the query point; the index of the vector corresponding to the nearest similarity is taken as a first index.
Further, referring to fig. 5, fig. 5 is a schematic diagram of a step of performing cleaning vectorization on a related corpus to obtain a corpus vector. In fig. 5, the step of performing washing vectorization on the relevant corpus to obtain a corpus vector may include, but is not limited to, step S501 to step S503.
S501, preprocessing related corpus to obtain a first corpus; the preprocessing includes at least one of case conversion, complex conversion, or full-half angle conversion.
S502, correcting errors and word segmentation are carried out on the first corpus, and a plurality of words are obtained.
S503, inputting a plurality of words into a vectorization model to obtain a plurality of corpus vectors.
It may be understood that the first corpus may be a preprocessed related corpus, and the plurality of words may be a word set obtained by word segmentation of the preprocessed related corpus according to a certain order. The vectorization model may be an existing model, which may be preset in the processor of the index build midst.
In some possible embodiments of the present application, the index building middle stage may perform at least one of case conversion, complex-body conversion, or full-half-angle conversion on the relevant corpus, and then obtain a processed relevant corpus. The processed related corpus can be segmented to obtain a plurality of words, the words can be any word or words formed by a plurality of words in the related corpus, and the words are input into a preset vectorization model by an index construction center to obtain a plurality of corpus vectors.
Illustratively, the corpus of the input index build center is "in a set-up ". The corpus is output as' in the setting and canceling after text pretreatment (case and complex, full half angle and other conversions); after error correction, the method can output 'in social security reimbursement'; then, after word segmentation, words such as [ "in", "social security", "reimbursement", "middle" ] and the like can be output. The vectorization of words is a numeric process of words, the distance can be calculated by an algorithm after the numeric process to evaluate the similarity between words, the vectorization can be performed by using a model, and the words such as [ "in", "social security", "reimbursement", "middle" ] can be obtained after vectorization to obtain decimal numerical representation similar to [0.0025752163,0.8979353,0.97538406,4.2164777E-4 ].
Further, referring to fig. 6, fig. 6 is a schematic diagram of steps for determining a first instance of a search service according to product identity information and mapping table information. In fig. 6, the step of determining the first instance of the search service according to the product identity information and the mapping table information may include, but is not limited to, steps S601 to S602.
S601, identifying a first keyword of the product identity information and an input keyword of an input side of the mapping table information.
S602, comparing the first keyword with the input keyword, and determining a plurality of retrieval service instances, in which the first keyword is identical to the input keyword, as a first instance.
It is understood that the product identity information may include an english field or a digital field. The product identity information of each product may only have a part of information difference, the first keyword may be digital information of the product identity information difference, the mapping table information may include an input side and an output side, the output side is a product retrieval service instance of the product identity information, and the input side may be complete and complete product identity information, or may be product identity information with only keywords
In some possible embodiments of the present application, when determining the first instance of the search service, it may be determined that a plurality of search service instances, in which the first keyword is identical to the input keyword, are the first instance by identifying the keyword of the product identity information and the input keyword of the input side of the mapping table information, by comparing whether the two keywords are identical to each other.
Further, the step of sending the search request to the search service according to the product balancing policy so that the search service obtains the search result according to the search request and the plurality of search service instances may include, but is not limited to, step S701, step S702, or step S703.
S701, sequentially sending a plurality of sub-requests to the search service according to the polling product balancing strategy, so that the search service obtains a search result according to the sub-requests and a plurality of search service instances.
In some possible embodiments of the present application, the retrieval request may include several sub-requests. The gateway can sequentially send a plurality of sub-requests to the retrieval service according to the self polling product balancing strategy, so that the retrieval service obtains a retrieval result according to the sub-requests and a plurality of retrieval service instances.
Illustratively, the gateway receives 3 retrieval requests, and only two retrieval service instances are a and B, respectively. The gateway can send the first sub-request to the search service according to the polling product balancing strategy of the gateway, so that the search service obtains a search result corresponding to the first sub-request and the search service instance A according to the first sub-request. And then the second sub-request is sent to the retrieval service, so that the retrieval service obtains a retrieval result corresponding to the second sub-request and B according to the sub-request and the retrieval service instance B. And then the third sub-request is sent to the retrieval service, so that the retrieval service obtains a retrieval result corresponding to the third sub-request and the retrieval service instance A according to the third sub-request.
S702, according to the hash product balancing strategy, a plurality of sub-requests are sent to the retrieval service, so that the retrieval service obtains a retrieval result according to the sub-requests and a plurality of retrieval service instances.
In some possible embodiments of the present application, the policy preset in the gateway may be a hash product balancing policy, and the plurality of sub-requests are sent to the search service, so that the search service obtains the search result according to the sub-requests and the plurality of search service instances.
Illustratively, when there are 3 instances in the search service, numbered 0-2. Extracting the request ID of each request, modulo 3 the request ID of each request to obtain a remainder between 0 and 2, and sending the sub-request corresponding to the remainder 0 to the search service so that the search service obtains a first search result according to the sub-request and the first search service instance. And sending the sub-request corresponding to the remainder 1 to the search service, so that the search service obtains a second search result according to the sub-request and the second search service instance. And sending the sub-request corresponding to the remainder 2 to the retrieval service, so that the retrieval service obtains a third retrieval result according to the sub-request and a third retrieval service instance.
S703, randomly sending a plurality of sub-requests to the retrieval service according to a random product balancing strategy, so that the retrieval service obtains a retrieval result according to the sub-requests and a plurality of retrieval service instances.
In some possible embodiments of the present application, the gateway may randomly send a plurality of sub-requests to the search service according to a random product balancing policy, so that the search service obtains a search result according to the sub-requests and the plurality of search service instances.
Illustratively, the gateway receives 3 search requests and the number of search service instances is 3, and transmits any sub-request to the search service, so that the search service obtains the first search result according to any sub-request and the first search service instance. And sending any one sub-request of the remaining two sub-requests to the retrieval service, so that the retrieval service obtains a second retrieval result according to the sub-request and the second retrieval service instance. And sending the remaining last sub-request to the retrieval service so that the retrieval service obtains a third retrieval result according to the sub-request and a third retrieval service instance.
It will be appreciated that the order in which the 3 search requests are sent and the order of the 3 instances are not limited herein.
The semantic indexing method according to the product route of the present application is described below with reference to the accompanying drawings:
it should be noted that, the number of search requests in this embodiment is 1, the product identity information is a product ID, and the 2 search service instances corresponding to each product ID are respectively engine-instance1 and engine-instance2.
First, referring to FIG. 7, FIG. 7 is a schematic diagram of a semantic index retrieval service workflow according to product routing. In fig. 7, the man-machine platform may maintain a mapping relationship between a set of products and a search service instance list, for example, a product with a lower left corner product ID of app1 may correspond to two instances of engine-instance1 and engine-instance 2; triggering an index construction flow of the product app1 on a man-machine platform, calling an index construction middle stage by the man-machine platform, inquiring relevant corpus of app1 by the index construction middle stage, cleaning vectorization, and carrying out index construction; when the index construction of the index construction middle stage is completed, the index file of the app1 can be saved to the nas disk, and the index is loaded by notifying a retrieval service through a gateway; the gateway can forward the index loading notification of the product app1 to two instances of the engine-instance1 and the engine-instance2 according to the mapping relation of the retrieval service instance list; after the loading of the retrieval service instance is completed, the service state can be changed, and the service is provided for the outside, so that the index construction or updating process is completed.
After completing the index construction process, referring to fig. 8, fig. 8 is a schematic diagram of a process of receiving a request by a search service. In fig. 8, all requests to access the retrieval service carry identity information APPID in the request header. After receiving the request, the gateway filters the request according to the mapping relation between the product maintained by the man-machine and the retrieval service instance to obtain an available service instance list, and the service instances load the semantic index of the product; and selecting an instance from the instance list according to the polling load balancing strategy of the gateway to finally forward the request to the retrieval service, and obtaining a final retrieval result by the retrieval service according to the specific request and the selected instance.
In summary, the content retrieval platform usually serves a plurality of products, and the conventional manner of loading the semantic indexes of all the products into the memory brings huge machine cost pressure, and is difficult to manage and maintain. The semantic indexing method according to the product routing, which is provided by the application, is characterized in that a retrieval service instance can only load semantic indexes of part of products, maintain a mapping relation between a product and the service instance, and route according to product identity information when a gateway forwards a request, so that the request can be accurately forwarded to the service instance loaded with the semantic indexes of the product, and the main advantages of the scheme are as follows:
1. The memory overhead of the semantic index is reduced, the resource utilization rate is improved, and the machine cost is reduced.
2. The indexes of different products are distributed on different examples, so that semantic indexes of partial products can be flexibly constructed/loaded/unloaded, the time consumption of index updating is shortened, and the flexibility and maintainability of the indexes are enhanced.
3. The capacity expansion and contraction can be carried out on the retrieval service instance of a certain product according to the flow condition, so that the service elasticity is enhanced, the concurrency risk is reduced, and the service stability is improved.
4. The method and the device enable the processes of index construction/loading/unloading and the like to be lighter, shorten the time consumption, and enhance the flexibility and maintainability.
In addition, referring to fig. 9, in fig. 9, the embodiment of the application further provides a semantic indexing device according to product routing. The semantic indexing device according to the product route may include: the acquisition module 1001 and the processing module 1002. Wherein the obtaining module 1001 may be configured to obtain a retrieval request for accessing a retrieval service. The retrieval request may include product identity information for the product to be accessed. The processing module 1002 may be configured to determine a first instance of the search service according to the product identity information and the mapping table information, and may send a search request to the search service, so that the search service may obtain a search result through internal algorithm processing according to the search request and the first instance; the first instance may be used to characterize a search service and a plurality of search service instances in which the index file is loaded corresponding to the search request.
It should be noted that the acquiring module may be any integrated circuit module or a micro processor module obtained by integrating a chip with a processing function and its peripheral circuit through the existing integration technology. The processing module may be any integrated circuit module or micro-processor module obtained by integrating a chip with a processing function and its peripheral circuits in the prior art. And the processing module may also include one or more memories. One or more memories may be used to store data such as mapping table information or search results.
In some embodiments of the present application, the acquisition module 1001 may be disposed in the same gateway or device with a processor as the processing module 1002. The acquisition module 1001 may acquire a search request for accessing a search service through a chip inside its own processor. The processing module 1002 may receive the request, determine a first instance of the search service according to the product identity information and the mapping table information in the request, and send the search request to the search service, so that the search service obtains a search result according to the search request and a plurality of search service instances that have loaded the index file. The acquisition module 1001 may be any module that interfaces with a gateway or a processor internal to the device. The acquisition module 1001 may transmit the acquired data to the processor of the processing module 1002 through a wired or wireless connection with the processor. The processor of the processing module 1002 may perform data processing through an internal chip, to finally obtain a search result. The specific device connection manner and device arrangement of the acquisition module 1001 and the processing module 1002 are not limited.
It can be understood that the content in the embodiment of the semantic indexing method according to the product route shown in fig. 1 is applicable to the embodiment of the semantic indexing device according to the product route, and the functions of the embodiment of the semantic indexing device according to the product route are the same as those of the embodiment of the semantic indexing method according to the product route shown in fig. 1, and the beneficial effects achieved by the embodiment of the semantic indexing device according to the product route shown in fig. 1 are the same as those achieved by the embodiment of the semantic indexing device according to the product route shown in fig. 1.
Referring to fig. 10, the embodiment of the application further discloses a computer device, including:
at least one processor 1011;
at least one memory 1012 for storing at least one program;
the at least one program, when executed by the at least one processor 1011, causes the at least one processor 1012 to implement the semantic indexing method embodiment according to product routing as shown in fig. 1.
It will be appreciated that the embodiments of the semantic indexing method according to the product route shown in fig. 1 are applicable to the embodiments of the present computer device, and the functions of the embodiments of the present computer device are the same as those of the embodiments of the semantic indexing method according to the product route shown in fig. 1, and the advantages achieved are the same as those achieved by the embodiments of the semantic indexing method according to the product route shown in fig. 1.
The embodiment of the application also discloses a computer readable storage medium, in which a processor executable program is stored, which when executed by a processor is used to implement the semantic indexing method embodiment according to product routing as shown in fig. 1.
It can be appreciated that the content of the embodiment of the semantic indexing method according to the product route shown in fig. 1 is applicable to the embodiment of the computer readable storage medium, and the functions of the embodiment of the computer readable storage medium are the same as those of the embodiment of the semantic indexing method according to the product route shown in fig. 1, and the advantages achieved by the embodiment of the semantic indexing method according to the product route shown in fig. 1 are the same as those achieved by the embodiment of the semantic indexing method according to the product route shown in fig. 1.
In some alternative embodiments, the functions/acts noted in the block diagrams may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may in fact be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Furthermore, the embodiments presented and described in the flowcharts of this application are provided by way of example in order to provide a more thorough understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed, and in which sub-operations described as part of a larger operation are performed independently.
Furthermore, while the present application is described in the context of functional modules, it should be appreciated that, unless otherwise indicated, one or more of the functions and/or features may be integrated in a single physical device and/or software module or one or more of the functions and/or features may be implemented in separate physical devices or software modules. It will also be appreciated that a detailed discussion of the actual implementation of each module is not necessary to an understanding of the present application. Rather, the actual implementation of the various functional modules in the apparatus disclosed herein will be apparent to those skilled in the art from consideration of their attributes, functions and internal relationships. Thus, those of ordinary skill in the art will be able to implement the present application as set forth in the claims without undue experimentation. It is also to be understood that the specific concepts disclosed are merely illustrative and are not intended to be limiting upon the scope of the application, which is to be defined by the appended claims and their full scope of equivalents.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium upon which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.
It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
In the foregoing description of the present specification, descriptions of the terms "one embodiment/example", "another embodiment/example", "certain embodiments/examples", and the like, are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that: many changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the application, the scope of which is defined by the claims and their equivalents.
While the preferred embodiment of the present invention has been described in detail, the present invention is not limited to the embodiments, and one skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention, and these equivalent modifications or substitutions are intended to be included in the scope of the present invention as defined by the appended claims
In the description of the present specification, reference to the terms "one embodiment," "another embodiment," or "certain embodiments," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that: many changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the application, the scope of which is defined by the claims and their equivalents.
Claims (10)
1. A method of semantic indexing according to product routing, comprising:
acquiring a retrieval request for accessing a retrieval service; the search request comprises product identity information of a product to be accessed;
determining a first instance of a search service according to the product identity information and the mapping table information, and sending the search request to the search service so that the search service obtains a search result according to the search request and the first instance;
the first instance is used for representing a plurality of retrieval service instances in which index files are loaded, which correspond to the retrieval request, in the retrieval service.
2. A method of semantic indexing according to product routing according to claim 1, wherein the step of sending a search request to the search service to cause the search service to obtain a search result based on the search request and the first instance comprises:
Determining a plurality of retrieval service examples from an example list set in the retrieval service according to the product identity information and the mapping table information; the instance list set is used for representing all retrieval service instances in the retrieval service, wherein index files are loaded in the retrieval service instances;
and sending the search request to the search service according to a product balancing strategy, so that the search service obtains a search result according to the search request and the plurality of search service instances.
3. A method of semantic indexing according to product routing according to claim 2, wherein the set of instance lists is obtained by:
responding to a loading index notification sent by an index construction center, determining a plurality of second examples corresponding to each product identity information from a man-machine interaction platform and sending the second examples to the retrieval service, so that the retrieval service loads a first index to the second examples from the index construction center to obtain an example list set;
the first index is constructed when the index construction center station sends out the loading index notification.
4. A semantic indexing method according to claim 3, wherein the index building center builds the first index by:
Responding to an index construction instruction sent by a man-machine interaction platform, and inquiring related corpus of each product;
washing vectorization is carried out on the related corpus to obtain a corpus vector set;
and determining a first index of each product according to the corpus vector set and an annoy algorithm.
5. The method for semantic indexing according to product routing according to claim 4, wherein the step of performing cleaning vectorization on the relevant corpus to obtain corpus vectors comprises:
preprocessing the related corpus to obtain a first corpus; the preprocessing comprises at least one of case conversion, complex-form conversion or full-half angle conversion;
correcting the first corpus and segmenting the first corpus to obtain a plurality of words;
and inputting the words into a vectorization model to obtain a plurality of corpus vectors.
6. A method of semantic indexing according to product routing according to claim 1, wherein the step of determining a first instance of a search service based on the product identity information and mapping table information comprises:
identifying a first keyword of the product identity information and an input keyword of an input side of the mapping table information;
Comparing the first keyword with the input keyword, and determining a plurality of retrieval service instances of which the first keyword is identical to the input keyword as a first instance.
7. A method of semantic indexing according to product routing according to claim 2, wherein the search request comprises a plurality of sub-requests, the step of sending the search request to the search service according to a product balancing policy to cause the search service to obtain a search result according to the search request and the plurality of search service instances, comprising:
according to a polling product balancing strategy, sequentially sending the plurality of sub-requests to the retrieval service, so that the retrieval service obtains a retrieval result according to the sub-requests and the plurality of retrieval service instances;
or sending the plurality of sub-requests to the retrieval service according to a hash product balancing strategy, so that the retrieval service obtains a retrieval result according to the sub-requests and the plurality of retrieval service instances;
or according to a random product balancing strategy, the plurality of sub-requests are randomly sent to the retrieval service, so that the retrieval service obtains a retrieval result according to the sub-requests and the plurality of retrieval service instances.
8. A semantic indexing apparatus according to product routing, comprising:
the acquisition module is used for acquiring a retrieval request for accessing the retrieval service; the search request comprises product identity information of a product to be accessed;
the processing module is used for determining a first instance of a search service according to the product identity information and the mapping table information, and sending the search request to the search service so that the search service obtains a search result according to the search request and the first instance;
the first instance is used for representing a plurality of retrieval service instances in which index files are loaded, which correspond to the retrieval request, in the retrieval service.
9. A computer device, comprising:
at least one processor;
at least one memory for storing at least one program;
the at least one program, when executed by the at least one processor, causes the at least one processor to implement the semantic indexing method according to product routing as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium having stored therein a program executable by a processor, characterized in that: the processor executable program when executed by a processor is for implementing a semantic indexing method according to product routing as claimed in any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311690804.4A CN117539974A (en) | 2023-12-07 | 2023-12-07 | Semantic indexing method, system, device and storage medium according to product route |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311690804.4A CN117539974A (en) | 2023-12-07 | 2023-12-07 | Semantic indexing method, system, device and storage medium according to product route |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117539974A true CN117539974A (en) | 2024-02-09 |
Family
ID=89791765
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311690804.4A Pending CN117539974A (en) | 2023-12-07 | 2023-12-07 | Semantic indexing method, system, device and storage medium according to product route |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117539974A (en) |
-
2023
- 2023-12-07 CN CN202311690804.4A patent/CN117539974A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102521334B (en) | Data storage and query method based on classification characteristics and balanced binary tree | |
| JP6928677B2 (en) | Data processing methods and equipment for performing online analysis processing | |
| CN105824957A (en) | Query engine system and query method of distributive memory column-oriented database | |
| CN109669925B (en) | Management method and device of unstructured data | |
| WO2023131218A1 (en) | Graph data storage | |
| CN104239377A (en) | Platform-crossing data retrieval method and device | |
| US20210271658A1 (en) | Data edge platform for improved storage and analytics | |
| CN105335481A (en) | Large scale character string text suffix index building method and device | |
| US10372736B2 (en) | Generating and implementing local search engines over large databases | |
| EP4423628A1 (en) | Method and system for data query | |
| CN112102840A (en) | Semantic recognition method, device, terminal and storage medium | |
| KR101955376B1 (en) | Processing method for a relational query in distributed stream processing engine based on shared-nothing architecture, recording medium and device for performing the method | |
| AU2019241002B2 (en) | Transaction processing method and system, and server | |
| CN110908996B (en) | Data processing method and device | |
| WO2021012211A1 (en) | Method and apparatus for establishing index for data | |
| CN117539974A (en) | Semantic indexing method, system, device and storage medium according to product route | |
| CN111767287A (en) | Data import method, device, equipment and computer storage medium | |
| CN111125216A (en) | Method and device for importing data into Phoenix | |
| CN106933882A (en) | A kind of big data incremental calculation method and device | |
| CN114238390A (en) | Data warehouse optimization method, device, equipment and storage medium | |
| CN112835932A (en) | Batch processing method and device of service table and nonvolatile storage medium | |
| JP2000339332A (en) | Medium recording retrieval index, method and device for updating retrieval index and medium recording its program | |
| US10380090B1 (en) | Nested object serialization and deserialization | |
| CN113726342B (en) | Segmented difference compression and inert decompression method for large-scale graph iterative computation | |
| CN111258955A (en) | File reading method and system, storage medium and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |