52 results about "Index mapping" patented technology

The invention relates to the technical field of intelligent scoring, in particular to a teacher teaching skill multi-dimensional intelligent scoring method and system.The method comprises the following steps that based on a teaching behavior multi-dimensional evaluation system, scoring standard connectivity strength is recognized, behavior consistency and language stability are analyzed, a feature behavior set is extracted, adjacent areas are expanded, interference nodes are removed, and the scoring standard connectivity strength is obtained; marking isolated point and breakpoint complementation structures, combining small regions, removing fragments, extracting behavior coding sequences, clustering and distributing labels, identifying index mapping, and outputting a teacher teaching skill classification and partition identification label set. The judgment of behavior consistency and language expression stability on skill features is enhanced, the accuracy and real-time performance of an evaluation result are improved, particularly, the strength and deficiency of teachers are meticulously analyzed and fed back in a complex teaching scene, teacher growth and teaching quality optimization are effectively supported, the high efficiency and unbiasedness of an evaluation framework are improved, and the evaluation efficiency is improved. And valuable education decision support is provided.

The invention discloses a data processing system and method for big data modeling, and belongs to the technical field of data processing.The method specifically comprises the steps that feature tree operation records, geometric grids and constraint solving logs are collected in a computer aided design environment and combined into an initial data packet, and a design number and a time stamp are bound; respectively converting the three types of data into a topological sequential sequence, a graph data structure and a constraint conflict trajectory sequence, and unifying the topological sequential sequence, the graph data structure and the constraint conflict trajectory sequence into a sample unit format; generating a geometric semantic slice sequence and bidirectional index mapping on the graph data structure according to a topological time sequence, generating a multi-view rendering result for the slice, synthesizing a tensor, and keeping an index relationship from pixels to topological entities; performing time sequence alignment based on the multi-view rendering tensor, the topology time sequence and the constraint conflict trajectory sequence, generating a geometric sample index table, and performing fragmentation storage; and finally, exporting a uniform-format data packet containing a graph data structure, a time sequence and image features from the index table for direct reading of a big data modeling processor.

The embodiment of the invention discloses a password information processing method and device, electronic equipment and a storage medium, multiple groups of encryption key pairs are bound with a key effective time interval, a key corresponding to a playing time period is quickly matched in combination with an index mapping table, a user does not need to manually input multiple passwords, and the user experience is improved. The problem that the password needs to be input repeatedly after being replayed after being modified is solved; a key is stored by an asymmetric encryption algorithm, and is transmitted through a secure transmission channel, so that the key storage and transmission security is ensured, and the problem of insufficient security of existing key management is improved; the integrity of the decrypted data is verified by means of the verification key, and the equipment end is triggered to retransmit incomplete data fragments, so that the defects of lack of data integrity guarantee and an exception recovery mechanism are overcome; and meanwhile, the key query efficiency is improved by the index mapping table.

The invention relates to the field of data lookup, and discloses a locality perception lookup table lookup method, which comprises the following steps of: S1, reordering between rows: clustering and reordering the rows of an input vector and a weight matrix according to the same quantized value to enable elements with the same quantized value to be continuously distributed; for a common matrix GeMV scene, cache locality is improved through rearrangement and subsequent blocking operation, and repeated loading of a lookup table LUT is reduced. And the data locality of the GeMV scene is remarkably improved through inter-line reordering and a virtual mapping strategy. On the basis of reordering of input vector element values, elements with the same value are aggregated, so that the LUT line loading times are reduced to 256 times at most from being in direct proportion to the input dimension d, and frequent replacement of the LUT lines in the WRAM is avoided; meanwhile, a weight matrix is divided into BR * BC sub-matrixes, the characteristic that vertical adjacent rows share an accumulator is utilized, multi-row sub-matrix calculation is loaded at a time, and virtual rearrangement maintains the row matching relation between an input vector and the weight matrix through index mapping.

The invention relates to a multi-line text searching and replacing method and system, and the method comprises the steps: carrying out the preprocessing of an original multi-line text, a multi-line searching content and a multi-line replacing content, removing the head and tail spaces of each line, and obtaining a corresponding logic text, a logic searching content and a logic replacing content, judging whether the logic search content contains a preset wildcard character or not, if the logic search content does not contain the preset wildcard character, carrying out simple character string matching on the logic text to obtain a logic index search result, converting the logic index search result into an original character index by utilizing a preset index mapping relation for positioning, and executing simple replacement operation, and if the logic search content contains the preset wildcard character, carrying out simple character string matching on the logic text to obtain a logic index search result; and if the logic text is matched with the fuzzy item list, regarding the logic search content as an integral matching block, performing sliding comparison according to rows in a row array corresponding to the logic text, if the matching succeeds, extracting the fuzzy item list and the starting and stopping row indexes of the matching block, performing positioning by utilizing an index mapping relation, and executing fuzzy replacement operation of inserting the fuzzy item list into a placeholder.

The invention discloses a multi-modal database-oriented adaptive index structure selection method, which comprises the following steps of: analyzing statistical characteristics such as data dimensions, variances, sparseness and distance distribution of each modal, calculating hidden dimensions of the modals, and describing effective data complexity of the modals; based on a preset index adaptation rule, automatically mapping each mode to an optimal index type in the candidate index structure set, and completing automatic construction of a local index and binding of the local index and a global routing structure according to the optimal index type; on the basis, the multi-modal query can be automatically routed to the corresponding index to execute retrieval according to the index mapping relation. According to the method, an index structure self-adaptive selection mechanism without manual configuration is realized, the index maintenance cost of the multi-modal database can be remarkably reduced, and the query efficiency and expandability in a complex retrieval scene are improved.

This application provides a rapid migration method and system based on customized data packages, as well as a readable storage medium. The rapid migration method based on customized data packages includes: extracting metadata from oil and gas exploration data using a general data acquisition module, employing the system's native API and file parsing algorithms; based on the metadata and the target operating system, formulating a migration strategy and monitoring the migration process through a control module, and generating a task execution list; dynamically adjusting the execution plan of the task execution list through a scheduling module, and generating a task execution priority list; and according to the task execution priority list, encapsulating and transmitting the data to the target system through a migration module, and unpacking the data according to the mapping index file and the unpacking priority list. Through the technical solution of this application, seamless data synchronization across operating systems can be achieved through customized data packaging and index mapping strategies, ensuring that the migrated data is immediately available and improving the execution efficiency of seismic exploration projects in multi-platform collaborative scenarios.

The invention provides an imaginary number hybrid simulation consistency data interaction method and system based on puppet entities, and the method comprises the steps: determining virtual simulator force resources participating in simulation, and generating a virtual force-puppet entity four-layer element index mapping table; according to the four-layer element index mapping table, exclusive data field intervals are divided for puppet entities of the same kind, data field identifiers are further dynamically distributed to members in the intervals, and data field isolation between the puppet entities in the digital space is completed; configuring transmission strategies for different theme protocol types according to different types of protocol data of different puppet entities in respective data interaction exclusive domains; the puppet entity carries out frequency division on the synchronization time macro-beat to the micro-beat capable of being scheduled and calculated in the simulator according to the simulation period requirement of the mapped simulator; and formulating a set of full life cycle management mechanism for the puppet entity, and performing unified management on the states of stages such as creation, initialization, operation, pause / recovery, termination and destruction of the puppet entity in the simulation process. According to the method, the data transmission overhead is effectively reduced, and the space-time consistency and situation consistency of imaginary number hybrid simulation are ensured.

Techniques for vector database indexing are described. In an example, a plurality of unstructured data is obtained. Vector embeddings are performed on each unstructured data from amongst the plurality of unstructured data to generate a plurality of vectors. Each vector from amongst the plurality of vectors is indicative of attributes of the unstructured data from amongst the plurality of unstructured data. The plurality of unstructured data and the plurality of vectors is stored in the vector database. A vector index corresponding to the plurality of vectors based on a first indexing mechanism is generated where the vector index includes at least one cluster of vectors with similar attributes. An index mapping is performed to store the vector index into a relational database.

The application belongs to the technical field of computer software and information security, and particularly relates to a pseudo-random dictionary generation and distribution method and system based on index mapping. The method comprises the following steps: obtaining a character dictionary and a target string length to obtain an initial character sequence; randomly rearranging the initial character sequence to generate a disordered character dictionary; calculating a total combination space of the string according to the disordered character dictionary and a character repetition constraint condition; mapping a numerical index into a target string on demand based on the total combination space of the string through a pseudo-random dictionary mapping method; grouping the numerical index to form a serial number pool, and persistently storing the index sequence in each serial number pool after randomly disturbing the index sequence; and taking out an actual index value from the current serial number pool according to a distribution cursor during the running period, converting the actual index value into a final string result through a mapping method, and outputting the final string result. The application realizes the unique generation and safe distribution of the string through the index mapping and double disordering mechanism, avoids the running period repetition verification, and prevents the number from being speculated.

The application discloses a kind of embedded device font library dynamic replacement system and method based on Unicode and GB2312 encoding mapping, belong to embedded system character coding and font library management technical field.Its method is by constructing lightweight Unicode-GB2312 bidirectional index mapping table, realizes Unicode character on-demand loading and mixed rendering under the premise of reserving original font library.Hash table is used to accelerate query, virtual area code extension and incremental updating mechanism, so that storage occupancy is reduced by more than 95%, font library updating efficiency is improved by 100 times, and progressive adaptation of full Unicode character set is supported.The application can be widely applied to intelligent terminal, industrial equipment and other scenes, and solves the contradiction between multi-language compatibility and resource limitation.

The invention discloses a large-model-based case material intelligent homing system and method, and relates to the technical field of data storage. After case type feature information is extracted through a large language model, a unique code is analyzed according to a preset coding rule and is bound with a homing target; the method comprises the following steps: classifying and storing case materials, constructing a field index mapping table containing a unique case identifier and a storage partition address, creating a distributed intelligent homing dual-channel and a homing shunt, accurately distinguishing cases to be homed, supplementing materials and invalid submission based on the index mapping table, performing corresponding processing, synchronously cleaning index table completion records, and completing the homing process. Homing delay and data disorder caused by mixed storage are avoided; the problems of unreasonable resource scheduling and server waste are solved by extracting a cross-regional space-time rate threshold to judge abnormality, predicting the number of abnormalities in combination with a sliding window and dynamically configuring a dual-channel memory proportion. The retrieval and eye-returning efficiency is greatly improved, the eye-returning process in a cross-domain scene is ensured to be stable and efficient, and resource utilization is remarkably optimized.

The embodiment of the application provides a data processing method, device and electronic equipment, the method comprises: obtaining to be stored data, the to be stored data comprises at least two key-value pair data, each key-value pair data in at least two key-value pair data comprises: string and the value corresponding to string;Determine any one key-value pair data in at least two key-value pair data, according to the string size ordering of key-value pair data, determine the storage location of each character in the corresponding key-value bucket in string, and store each character of string to the corresponding key-value bucket;According to the storage order of the value corresponding to string, determine the storage location of the value corresponding to string in value partition and store the value corresponding to string to value partition;According to the storage location of each character in the corresponding key-value bucket in string and the storage location of the value corresponding to string in value partition, establish index mapping table.

The invention provides a power plant equipment multi-modal fault causal chain retrieval and self-evolution method and system, and the method comprises the steps: obtaining equipment multi-modal heterogeneous data, extracting features through coding and projection, and outputting potential physical states and static semantic features; constructing a hybrid neurodynamic equation, and describing a physical quantity continuous flow by using a hypergraph convolution differential equation with an incidence matrix; driving hypergraph evolution based on formalized Ricci curvature, calculating a Wasserstein distance of a node pair by adopting Sinkhorn entropy regularization, and updating a hyperedge weight through a curvature flow equation; designing a physical topology double constraint loss function; variational neural symbol reasoning is introduced, and gradient propagation of discrete logic variables is achieved through GumbelSoftmax; aligning the physical state with the query vector by utilizing Poincare sphere index mapping in the hyperbolic retrieval space; an invariant causal prediction framework is adopted to carry out anti-fact intervention verification, knowledge base self-evolution is realized based on feedback, and fault causal chain retrieval which is physically consistent, feasible in calculation and deeply fused with multi-modal information is realized.

The embodiment of the invention discloses a meta-object code generation method and device, electronic equipment and a storage medium. According to the technical scheme, before the first source code file is compiled, the first source code file is pre-analyzed, the first connection statement in the source code is recognized, and the first connection statement is used for connecting the first function and the second function; obtaining a corresponding first function index and a second function index based on the first connection statement, and generating index mapping information based on the first function index and the second function index; and in the process of compiling the first source code file, generating a meta-object code corresponding to the first connection statement based on the index mapping information. By means of the technical means, the code size can be effectively controlled while the function binding efficiency is improved, and the dual requirements for program efficiency and size in the resource-constrained environment are met.

Aspects of the subject disclosure may include, for example, a technique for processing a decision network that includes obtaining an index encoding a mapping from potential values of an input parameter to decision parameters of the network's predicates, wherein the mapping associates potential values of the input parameter with decision parameters affected by those potential values; evaluating decision parameters affected by specified values of the input parameter, including identifying each decision parameter to which the index maps at least one specified values of the input parameter, and setting the values of those decision parameters in accordance with the input parameter's specified values; and analyzing the decision network, including evaluating the predicates of one or more of the decision nodes based on the values of the predicates' decision parameters, and determining, based on the values of the evaluated predicates and a topology of the decision network, that a particular terminal node encodes the network's output. Other embodiments are disclosed.

This invention discloses an adaptive index structure selection method for multimodal databases. By analyzing the statistical characteristics of each modality, such as data dimensionality, variance, sparsity, and distance distribution, the hidden dimension of each modality is calculated, and its effective data complexity is characterized. Based on preset index adaptation rules, each modality is automatically mapped to the optimal index type in the candidate index structure set, thereby automatically constructing local indexes and binding them to the global routing structure. Furthermore, multimodal queries can be automatically routed to the corresponding index for retrieval based on the index mapping relationship. This invention achieves an adaptive index structure selection mechanism that requires no manual configuration, significantly reducing the index maintenance cost of multimodal databases and improving query efficiency and scalability in complex retrieval scenarios.

The invention provides a data storage device and an operating method thereof. A data storage device may include a memory device and a memory controller. The memory controller is configured to extract a key value from the modified logical address and map an index generated based on the modified logical address and the key value to a physical address of the memory device.

The invention provides a code positioning method and device, electronic equipment and a storage medium, and relates to the technical field of data processing. The method comprises the following steps of: dividing a structured data region and a common text region of an input mixed text, respectively processing, analyzing a target code library, establishing index mapping between the text and a code position, carrying out query matching based on the mapping, and returning code position information. The problems that in the prior art, formatting fails or semantic errors are caused due to the fact that structured data and unstructured texts cannot be distinguished, and log checking is low in efficiency due to lack of a log exclusive searching and indexing mechanism can be solved. The technical effects that intelligent formatting is achieved, original typesetting of a common text is reserved, and associated codes are rapidly positioned to improve the development efficiency are achieved.

The embodiment of the invention provides a PIR preprocessing method and system based on two-party privacy full permutation, and belongs to the technical field of computers. The method comprises the following steps: receiving an index mapping request sent by a client, executing Hash mapping based on the index mapping request, and establishing an initial slot relationship of each data item; receiving a full-permutation index which is sent by the client and is generated based on the initial slot position relations, generating an associated random number sequence corresponding to the initial slot position relations, and determining mask values of the initial slot position relations; performing privacy rearrangement on each data item based on the full arrangement index, the initial slot position relationship and the corresponding mask value to obtain a privacy arrangement storage structure; private permutation storage structure storage is executed for PIR query invocation. According to the scheme of the invention, the privacy rearrangement of the database is completed through lightweight Hash and a full arrangement mechanism in a single server environment, and the privacy security of the query process is kept while the preprocessing calculation and communication overhead is remarkably reduced.

The application provides a hybrid batch adaptive switching method and device based on a Cocos rendering pipeline, a storage node, a static node and a dynamic node are constructed for each model group, when a scene is loaded or a round is switched, model nodes are collected to the storage node to perform grid merging to generate a merged model mounted on the static node to participate in rendering in a static batch mode, rendering batches and CPU overhead are greatly reduced; when a model state change event is received, the merged model is restored to an independent node according to index mapping data maintained in the storage node and is migrated to the dynamic node to perform dynamic batch, and the independent editability of the model is restored; after the state change processing is completed, the effective nodes are migrated back to the storage node to perform static batch again, so that reversible switching is realized between the two batch modes according to actual needs at runtime.

Updating fingerprints for deduplicated data segments in a cluster network by calculating new fingerprint hash values using a new hash process. A new index mapping from the present hash process to the new hash process is created, as is a new fingerprint index mapping from the updated hash process to the containers. The process generates new fingerprints with the new fingerprint hash, and calculates the new fingerprint hash values of the new data segments using the new hash process. Relevant data structures are updated using the current fingerprint hash values with the new fingerprint hash values. New containers are created as needed to store the new fingerprint hash values. Old data elements like containers and fingerprints are deleted during the upgrading process to reduce space overheads.

The invention discloses an index storage method and device for an access control policy, equipment and a medium, and the method comprises the steps: obtaining at least one to-be-processed policy rule in an access control architecture, and constructing a multi-dimensional attribute vector corresponding to each to-be-processed policy rule; performing clustering processing on all the multi-dimensional attribute vectors to obtain at least one cluster, and constructing a to-be-stored strategy template corresponding to each cluster; and performing index mapping on each to-be-stored strategy template, generating a target index corresponding to each to-be-stored strategy template, and performing uplink storage on the target index to the block chain. On the premise that the quality of the policy rules is not affected, the number of the policy rules is remarkably reduced, and the unique index corresponding to the policy rules is established so that the corresponding policy rules can be called according to the index information when the access request is received.

The application discloses a hydrological model deep learning rate setting method for large-scale multi-sub-basins, establishes a data warehouse supporting high-concurrency reading, constructs a structured index mapping table to realize direct addressing of data, reconstructs continuous hydrological physical processes into a vectorized whole-process differentiable computation graph, adapts parallel computing characteristics of a graphics processing unit by using instant compilation and operator fusion technology, constructs a double-parallel high-dimensional parameter tensor containing a basin dimension and a parameter group dimension, realizes concurrent forward evolution of a mass of basin-parameter groups by using multi-path random initialization and tensor broadcasting technology, calculates a gradient by using a reverse mode automatic differentiation mechanism, updates parameters by using a deterministic optimization algorithm, constructs a performance evaluation matrix based on a training set, filters out an optimal parameter set, and exports the optimal parameter set after verification by using a verification set. The application breaks through the efficiency bottleneck of traditional serial computation, considers convergence speed and global optimization accuracy, and realizes full-process automation of large-scale basin parameter rate setting.

The application belongs to the technical field of data processing and statistical analysis, in order to solve the technical problems of rigid index system, low data processing efficiency and lack of dynamic early warning in the prior art, a digital economy index calculation method and system based on multi-source heterogeneous data are provided, through standardized index mapping logic and adaptive weight adjustment algorithm, combined with automatic processing and model calculation system based on digital economy index, the whole process automatic processing of digital economy index from multi-source heterogeneous data collection, standardized splitting, adaptive model calculation to classification verification display is realized.

The invention discloses a first-in first-out data structure based on key value pair search and a processing method, relates to data processing and data storage technologies, and comprises a three-layer closed-loop architecture controlled by a hash index, an annular array and metadata. Wherein the Hash index layer is used for receiving an external key value operation request and positioning a data position in an annular array through index mapping; the annular array storage layer is used for physical storage and sequence management of data, an annular array of a continuous memory is adopted, and FIFO sequence management is achieved through a compact data unit and head and tail pointers; and the metadata control layer is used for maintaining the running state of the structure and providing parameter support for operation execution, and the metadata control layer adopts atomic variables to store core state parameters and realizes lock-free synchronization through CAS operation. The storage structure is optimized, the precursor / subsequent pointers of the double linked list are cancelled, a compact storage mode is adopted, and the memory utilization rate in a small data scene is improved.

The application discloses a kind of progressive retrieval enhancement generation method, system, terminal and medium based on reasoning chain.The generation method includes: based on source text data Construction Heterogeneous Knowledge Graph, Including the document corpus in external knowledge source is blocked processing, utilize big model from each document block Entity and relationship are extracted to construct basic knowledge graph, and the bidirectional index mapping between graph element and document block is established;Progressive decomposition is carried out to the retrieval question, and the query is decomposed into multiple interdependent sub-problems by big model, and the sub-problems and their dependency relationship are organized into a graph-structured logical execution graph;Progressive retrieval solving is carried out to the decomposed sub-problem, and the reasoning chain is constructed and the final retrieval answer is generated.The method disclosed in the application bridges the semantic gap, blocks the propagation of errors and provides a multi-hop question and answer solution with explicit reasoning guidance, improving the accuracy of multi-hop problem retrieval.

The invention provides an intelligent input recommendation method and system based on multi-source corpora and parallel indexing, and belongs to the technical field of intelligent input recommende.The method comprises the steps that core corpora are obtained, and weighted word frequency calculation is conducted on the core corpora to obtain final word frequency; extracting a plurality of first character prefixes of all vocabularies in the core corpus, and grouping according to the length of the character prefixes to obtain a plurality of prefix groups; distributing an independent thread for each prefix group and constructing index mapping in the independent thread; obtaining input associative query content issued by a user, and performing degradation matching according to the independent thread based on the input associative query content to obtain a matching result; according to the method, the core pain point of the association function of the existing input method can be solved, and the method has the characteristics of reproducibility, easiness in landing and wide adaptation, and is suitable for multi-scene input requirements of office, vehicle-mounted, embedded and the like.

The invention provides a three-dimensional space tile indexing method and system, and the method comprises the steps: generating a topological complementary code according to a preset topological complementary coding rule based on a topological relation between a reference position determined for a geometric object and a three-dimensional space tile when a tile index is established for the geometric object in a three-dimensional space, the topological complementary coding is based on a reference position, and represents a spatial intersection relationship between a geometric object and the interior and each boundary surface of the tile; according to the method, a unique main index tile is determined for a geometric object based on topological complementary coding and an index judgment rule, and the index judgment rule ensures that the geometric object in a public boundary area of adjacent tiles is identified as the main index tile by only one tile by analyzing information about a complementary direction boundary in the topological complementary coding, so that repeated indexing is avoided; establishing and maintaining index mapping between the geometric object and the main index tile; and in response to a spatial data subscription / query request, retrieving and returning matched geometric object data based on the indicated tile set and topology conditions.