51 results about "Query log" patented technology

The invention discloses an intelligent data query system and method based on natural language processing, and relates to the technical field of natural language processing and database query. According to the method, natural language input and database mode information are received, a historical query log is combined to construct mode knowledge representation, a query skeleton is generated on the basis, a fine-tuned large language model is called to generate candidate SQL statements, error detection and ambiguity recognition are carried out on the candidate statements, and the query result is obtained. And if necessary, triggering interaction clarification and updating a query result according to user feedback. And meanwhile, the wrong clauses are locally repaired through a gating mechanism, and the clauses are returned and regenerated when multiple times of repair fails, so that the correctness of the query statement is ensured. And finally, after the SQL is executed in the database, a result is fed back to the user. Besides, the system records generation and repair tracks in the operation process, and continuously optimizes the model based on reward shaping, comparative learning and self-game training, so as to improve the generalization ability in different business scenes.

A system for cognitive inferencing for large language models (LLMs), the system including: a integration layer communicatively connected to a static LLM, wherein the integration layer includes a processor configured to receive one or more user requests from a user, wherein the one or more user requests include a user interaction with the static LLM, generate one or more cognitive inference (CI) units from the one or more user requests, assign a confidence-falsifiability (CF) delta for each of the one or more CI units using a Socratic engine and append the one or more CI units and the CF delta to a cognitive inference (CI) log associated with the user, wherein the CI log serves as an inference engine for the static LLM, wherein the integration layer is configured to query the CI log upon a subsequent user interaction with the static LLM.

Systems and methods for generating virtual assistant proactive queries improving virtual assistant-user interaction. Initial user prompts, including trigger conditions, corresponding to one or more initial user sessions are received. The initial user prompts, and trigger conditions, are stored in a query log database. Pattern recognition is performed on the stored initial user prompts, and the corresponding trigger conditions, to determine a proactive prompt for a subsequent user session. A proactive response is generated from the proactive prompt. Prior to receiving a subsequent user prompt, the proactive response is provided during the subsequent user session upon detection of one or more trigger conditions corresponding to the proactive prompt.

The present disclosure provides an artificial intelligence-based alarm management based on structured time series data and unstructured contextual information where the conventional methods fails to do. Initially, the system receives a user query associated with an industry automation system and classifies into one of a) data request query b) an anomaly detection and diagnosis query c) a log request query and d) feedback for previous response using a query classifier. Simultaneously, the user query is converted into an associated timeseries query representation. Furthermore, a plurality of database queries is generated based on the timeseries query representation using query generator. Further, data pertaining to the user query is retrieved from an associated database based on database queries using Deep Learning models, thereby identifying a critical alarm, and to take appropriate actions if there is any anomaly. Finally, the retrieved data is displayed to the user in a user readable format.

In one example, a method performed by a processing system including at least one processor includes automatically inferring metadata about a structured query language database, constructing, based on the metadata, a prompt for generating a new structured query language query to be executed against the structured query language database, and providing the prompt to a language model for use in generating the new structured query language query.

The invention discloses a self-adaptive table structure generation and index optimization method for JSON data driven by a real-time mode, and relates to the field of computer information management application. According to the self-adaptive table structure generation and index optimization method for the JSON data driven by the real-time mode, efficient and automatic persistence from the JSON data to a relational database is achieved through a four-layer technical architecture including dynamic analysis, intelligent mapping, self-healing expansion and index optimization. Comprising a dynamic mode analysis engine which analyzes a JSON nested structure in real time and generates a mode feature tree with weight; the self-adaptive storage engine is used for dynamically generating or expanding a table structure according to the mode feature tree and executing data type mapping; the index optimizer is used for automatically creating / cleaning indexes based on query log analysis; and the uniform interface layer is compatible with multi-database dialects and shields bottom layer differences. Dynamic table structure generation, intelligent type inference and storage optimization, self-healing Schema extension, query-driven index self-optimization and cross-database compatibility are realized.

The invention provides a code vulnerability coping strategy generation method and related equipment. The method comprises the following steps: in response to a log recording instruction, generating a target log query statement according to the type of the log recording instruction; wherein the target log query statement is used for querying logs related to log4j vulnerability attack behavior characteristics; executing the target log query statement to obtain a target log query result; according to the target log query result, judging whether the log recording instruction contains a suspected log4j vulnerability attack execution behavior or not; if yes, link identification is carried out on the log4j vulnerability attack based on the target log query result, and a target attack link is obtained; and according to the target attack link, generating a target coping strategy for the log4j vulnerability attack. The network security defense capability can be improved.

The present disclosure relates to the technical field of big data governance and data asset operation management, and particularly relates to a data index evaluation device and a data index evaluation method. The method comprises: acquiring query logs, blood relation metadata, cost data and data quality evaluation results of each data index; determining health degree scores of each data index based on the query logs, the blood relation metadata, the cost data and the data quality evaluation results; determining unit cost values of each data index based on the cost data and the blood relation metadata; determining a screening index with a health degree score less than a score threshold and a unit cost value less than a value threshold based on the health degree scores and the unit cost values; wherein the screening index comprises any one of the data indexes; and generating evaluation information based on the screening index.

The invention discloses an internet-based medical insurance service platform comprehensive intelligent management system, which relates to the technical field of medical insurance intelligent management and comprises a data receiving module, a portrait construction module, a strategy adaptation module, a compliance verification module and a service navigation module. The system performs heterogeneous feature co-processing on the user identity authentication sequence and the medical service application text to generate a user service demand portrait; performing rule matrix matching through a multi-level policy adaptation engine, and outputting a grading policy adaptation result containing matching degree distribution; based on the distribution, bidirectional compliance verification is carried out on the expense settlement document image and the medicine directory query log, and a medical insurance settlement compliance report is formed; and generating and pushing a personalized service navigation strategy. According to the invention, deep integrated description of user service requirements and dynamic accurate compliance review based on quantitative evaluation are realized.

The invention discloses a real estate archive high-frequency stable user screening and demand label classification method. The method comprises the following steps: carrying out data preprocessing on real estate archive query logs, removing abnormal data and complementing missing information; screening high-frequency stable users through a multi-dimensional index, wherein the multi-dimensional index comprises monthly query frequency, a high-frequency time period proportion and a query interval fluctuation coefficient; based on the screened high-frequency stable user query behaviors, constructing a user demand feature library containing query purposes, regional attributes and archive types; comprehensively evaluating the demand coverage rate, the tag redundancy and the Jaccard similarity coefficient, and judging the matching degree of the existing archive tag and the user demand; and generating a dynamic demand label according to an evaluation result, and establishing a regular updating mechanism to adapt to user demand changes. The problems that in the prior art, high-frequency stable user screening accuracy is insufficient, and archive labels are disjointed with actual requirements of users are solved, and the real estate archive query efficiency and the user requirement satisfaction degree are effectively improved.

The application discloses a flink task log processing method based on K8S. The flink task log processing method comprises the following steps: associating a task identifier with a flink task role; listening to state information of a K8S container corresponding to the flink task, and storing the task identifier of the flink task role; collecting and storing log information of the flink task in the K8S container; and querying log information of the flink task role associated with the task identifier. The application also discloses a flink task log processing method based on K8S, a flink task log processing device based on K8S, a computer device and a computer readable storage medium. The flink task log information collection, storage and display of K8S are realized, so that a user can complete flink task submission, state updating and log information checking through a unified flink big data computing platform, the running problem of the flink task of K8S is conveniently located, and the problem locating efficiency is improved.

The application discloses a database slow query log processing method and device, equipment and a storage medium. The method comprises the following steps: in response to obtaining a to-be-executed structured query language (SQL) statement, executing the SQL statement, and monitoring a latch associated with the SQL statement to obtain time consumption information of the latch; determining whether the SQL statement is a slow query statement according to the execution time of the SQL statement; and in the case that the SQL statement is a slow query statement, outputting a slow query log of the SQL statement based on the time consumption information of the latch. The technical scheme of the embodiment of the application can output a slow query log containing latch time consumption information for a single SQL statement, facilitate the investigation of the cause of slow query, and ensure the high-performance operation of a database.

The application relates to a data query method, system, device and medium based on vector retrieval acceleration, comprising the following steps: collecting historical query log information generated by real users in a target application scenario, and constructing a query distribution model; based on the constructed query distribution model, adopting a multi-granularity data division strategy to divide original vector data to obtain fine-granularity data blocks; receiving a new query request initiated by a user to generate a query vector, and dynamically selecting nodes and data blocks participating in current query calculation according to the feature distribution of the query vector and the real-time load state of each node; and combining local Top-K results returned by each participating node, and performing accurate distance reordering to obtain a query result. The application can be widely applied to the fields of computer systems and big data processing technologies.

The application discloses an AI prediction-based graph structure optimization strategy generation method and device, equipment and a medium, relates to the technical field of data processing, and comprises the following steps: filtering out a key path and a key node of a graph structure through path access information recorded in a historical query log and node association information of the graph structure, and intelligently predicting an optimization strategy suitable for the current graph structure based on the topology of the key path and the key node and query characteristics through a reinforcement learning reward mechanism, so that directional iterative optimization of the graph structure is realized, and the execution efficiency of backtracking query is improved from the root.

The application discloses a kind of multi-dimensional data management method and system based on deep learning, it is related to data storage technical field, method includes: obtaining historical compression record and pre-processing, extraction each item compression feature and compression effect index, form historical compression sample set;Predictive model is constructed and is trained based on historical compression sample set, obtain the trained predictive model;Data dimension information and historical query log of target equipment are collected, the importance and query frequency of each data dimension are obtained;The co-occurrence frequency of any two dimensions is counted and its stability is calculated, and a global weighted co-occurrence matrix is established;The candidate compression method set of the target equipment is preset, the matching score of any pairing combination between each data dimension and each candidate compression method is calculated;The global compression graph of the target equipment is constructed, and the global optimal compression strategy is solved.The application can realize the efficient storage management of massive multi-dimensional data, and solve the problem of slow multi-dimensional collaborative query response.

The invention discloses a dynamic index generation method and system based on archive metadata, and relates to the technical field of archive information management.The method comprises the steps that a user interface is provided, and a user is supported to define a metadata model and archive field query conditions according to archive categories and business requirements; field query importance is evaluated in combination with query frequency and result influence degree so as to set index priorities, and a future query mode is predicted by utilizing historical query logs and machine learning so as to provide a basis for dynamic index generation; according to a field data type, a query condition and a performance requirement, automatically selecting or recommending an adaptive index type and optimizing configuration, and automatically executing a statement in a database to generate an index; performing performance test on the index, deploying and monitoring use and query performance after verification is passed, visually analyzing performance data, generating optimization suggestions, dynamically adjusting the index, automatically identifying field change to trigger update, and continuously optimizing configuration. The workload of file operation and maintenance can be reduced, and the file data query efficiency is improved.

An online system performs an inference task in conjunction with the model serving system and / or interface system to generate relevant product images for query auto-completion and query suggestion to help users better navigate their search experience. The online system generates a collection of query suggestions using search query log mining. For each query suggestion in the collection of query suggestions, the online system retrieves one or more catalog images that depict the query suggestion from a product catalog. The online system constructs a prompt to a text-to-image model including the query suggestion, and a request to generate one or more query images based on the query suggestion. The online system receives the query images from the text-to-image model and ranks the catalog and query images to identify an image to display to the user in association with the query suggestion.

The application discloses a kind of cross data source query dynamic optimization method and system, the method includes: in the query log retrieval SQL sentence in key word join, further identify and handle the join query sentence identified, generate the training set according to join condition and predicate information extracted from the join query sentence;Using the training set, the model is trained, and the predicate selectivity estimation model is obtained, and the predicate selectivity estimation is carried out for the given data table and predicate information;After receiving new query sentence, it is judged whether the new query sentence is cross-source join query statement, parse and extract the data table and predicate information of the new query statement, and input to the predicate selectivity estimation model to obtain the selectivity estimation parameter of all data tables;According to the selectivity estimation parameter, the data table corresponding to the new query statement is connected and sorted, and the optimal connection order is obtained, which greatly improves the query efficiency of cross data source join query.

The application provides an index scheme recommendation method and device, electronic equipment and storage medium. The method comprises: obtaining a query log corresponding to a target query statement and a plurality of data columns, the data columns including a result column to be queried by the target query statement and a condition column as a query condition; extracting a column query feature corresponding to each data column from the query log; determining a target data column among the plurality of data columns based on the column query feature corresponding to each data column and a preset screening rule; creating a plurality of candidate indexes based on the target data column; inputting the plurality of candidate indexes into a preset index recommendation algorithm to output a corresponding index scheme by the preset index recommendation algorithm. Thus, the index scheme is intelligently recommended, thereby improving the efficiency of recommending the index scheme and saving labor costs.

The invention discloses a table retrieval method and device for a database cold boot scene and electronic equipment. The method comprises the following steps: S1, preprocessing structure information; s2, semantic enhancement and vector representation generation; s3, constructing a reverse index; s4, performing semantic recall; s5, carrying out keyword constraint and filtering; and S6, performing multi-score fusion sorting. The method is suitable for performing accurate, stable and interpretable retrieval on new or semantic sparse data tables in an enterprise database under the condition of lack of historical query logs, manual annotation or statistical information access. According to the method, data table filtering, semantic enhancement description data table generation and table-level semantic vector generation, reverse index construction, semantic similarity score calculation and semantic similarity score calculation are carried out on the whole. The processes of keyword matching score calculation, coverage rate score query, accurate matching score calculation, filtered candidate data table generation, fusion sorting score calculation and sorting are executed in sequence, and all the steps form a complete and closed-loop retrieval processing link.

The invention discloses a multi-dimensional data management method and system based on deep learning, and relates to the technical field of data storage, and the method comprises the steps: obtaining historical compression records, carrying out the preprocessing, extracting all compression features and compression effect indexes, and forming a historical compression sample set; constructing a prediction model and training based on the historical compressed sample set to obtain a trained prediction model; collecting data dimension information and historical query logs of target equipment, and obtaining importance and query frequency of each data dimension; counting the co-occurrence frequency of any two dimensions and calculating the stability of the co-occurrence frequency, and establishing a global weighted co-occurrence matrix; presetting a candidate compression method set of the target equipment, and calculating a matching score of any pairing combination between each data dimension and each candidate compression method; and constructing a global compression graph of the target equipment and solving a global optimal compression strategy. According to the method, efficient storage management of mass multi-dimensional data can be realized, and the problem of slow response of multi-dimensional collaborative query is solved.

The invention relates to the technical field of block chains, in particular to a counterfeit domain name detection method and system based on the block chain technology, and the method comprises the following steps: data collection, block chain construction, intelligent contract execution and application execution feedback. Data collection: collecting multi-source data such as DNS query logs, certificate transparency CT logs, user report information and third-party threat intelligence, including network flow collection, certificate log collection, user report collection and threat intelligence collection, and block chain construction: adopting a decentralized block chain architecture; based on the decentralized architecture of the block chain and the cross-mechanism cooperation capability of the smart contract, data sharing and cooperation processing between different mechanisms are promoted, and the security protection level of the whole network space is improved. Through application of a zero-knowledge proof technology, privacy protection of legality of domain name query by a user is realized, and leakage and abuse of user access records are avoided.

The invention relates to the field of data analysis, in particular to an enterprise business management data analysis method based on data driving, which comprises the following steps: acquiring and preprocessing query logs and data block meta-information to obtain a standardized data set; analyzing an overlapping relation between the query range and the cluster coverage range to obtain a range overlapping reward factor; analyzing the query range size and the data distribution matching relationship to obtain a range size and data distribution matching degree reward factor; obtaining a corrected distance measure by combining the range overlapping reward factor and the range size and data distribution matching degree reward factor with the Euclidean distance; a data pre-fetching strategy and a service decision support basis are obtained through analysis based on a clustering result of modified distance measurement, so that the problems of insufficient analysis efficiency and inaccurate decision support caused by the fact that an existing method cannot comprehensively consider a query range overlapping degree, a range size and a data distribution matching degree are solved.

The invention relates to the technical field of databases, in particular to a database operation and maintenance method and system based on machine learning. The method comprises the steps of obtaining a real-time query log of a database; analyzing the real-time query log, and determining a query type; analyzing a query type, and determining a type distribution condition; inputting the type distribution condition into a preset analysis model, and determining a real-time change entropy value of the query type; comparing the real-time change entropy value with a type mutation threshold value; and if the real-time change entropy value exceeds the type mutation threshold value, updating the original index to obtain the latest index. Through real-time monitoring and analysis of a real-time query log, a data basis is provided, and the trend of a current query type is identified. The real-time change entropy value is calculated through the preset analysis model, the change degree of the query type is quantified, and whether the query type is remarkably changed or not is judged by comparing the real-time change entropy value with the type mutation threshold value, so that whether the original index is dynamically adjusted or not is decided, and the query performance is optimized.

The present disclosure provides a data processing method, which can be applied to the technical field of big data and the technical field of finance. The method comprises: in response to a user query request, obtaining a reference field by analyzing the user query request, the reference field comprising at least a target query period field and a target institution field to which the user belongs; based on the target query period field and the target institution field, reading target historical query log data from a database, the target historical query log data being generated before the target query period and being associated with the target institution; associating the target historical query log data with a plurality of predefined query semantic word groups to obtain a target query semantic word group; and associating the target query semantic word group with data labels of a plurality of labeled data to determine a plurality of target data from the plurality of labeled data. The present disclosure also provides a data processing device, equipment, storage medium and program product.

The invention discloses a method and device for automatically optimizing ordering keys of an OLAP database, and relates to the technical field of database query, and the method comprises the following steps: obtaining a query log, table metadata and a cluster state of the OLAP database; performing feature analysis on the query log, the table metadata and the cluster state to obtain a plurality of large-class features of the field; each large-class feature comprises a plurality of small-class features; performing quantitative scoring on each field according to each large-class feature and each small-class feature of the field, obtaining a dynamic score of each field, performing sorting, and selecting a plurality of fields with the maximum scores as candidate sorting keys; and determining whether to adjust the current sorting key to the candidate sorting key according to the dynamic score of the candidate sorting key and the dynamic score of the current sorting key. According to the method and the device, the sorting key can be automatically changed under the condition that the field with high query frequency is changed or the field is newly added due to service change, so that the query efficiency of the OLAP data system is improved.

The invention provides a log data searching method and device, electronic equipment and a storage medium, and the method comprises the steps: when log data is received, carrying out the aggregation of span data contained in the log data, and obtaining the core meta-information of the span data in the log data; the log data corresponds to a unique tracking identifier; fragmenting the log data to obtain multiple pieces of fragmented data; establishing a secondary index table, and writing the fragmented data, the core element information corresponding to the fragmented data and the position information of the storage position of the fragmented data into the secondary index table corresponding to the tracking identifier; and when a log data query request is received, determining a target secondary index table corresponding to the tracking identifier of the to-be-queried log data, and searching the to-be-queried log data according to the target secondary index table. According to the method, the data needing to be acquired can be quickly queried only by searching the secondary index table, so that the search time is greatly shortened, and the search difficulty is reduced.

The invention relates to the technical field of databases, and discloses a database table optimization method and device, equipment and a storage medium, and the method comprises the following steps: extracting a performance influence query statement from a query log of an original database table; analyzing the performance influence query statement, and determining a target query field and a target time range; and generating a target database table based on the target query field, the target time range and the original database table. According to the method, the performance influence query statement is extracted from the query log, the target query field and the target time range of the performance influence query statement are analyzed, then the optimized target database table is generated, redundant data are prevented from participating in calculation, query delay is reduced, and query statement execution efficiency is improved.

The invention discloses a self-adaptive index optimization system based on timing context driving, and relates to the field of computer databases. The context extraction unit is used for initializing a lock-free circular queue by applying a sliding window accumulation mechanism, calculating feature data in combination with a time sequence data stream and a query log, performing normalization processing on the feature data and constructing a context feature vector; the self-adaptive decision-making unit is used for taking the context feature vector and an algorithm library as input and outputting an optimal algorithm and a corresponding weight through an attention network; and the index reconstruction execution unit is used for executing a dynamic adjustment mechanism of the index structure in a linkage manner according to the output optimal algorithm and the corresponding weight, evaluating and updating the current state based on a dynamic adjustment result, and feeding back an updating result to the context extraction unit. According to the method, high-precision spatial-temporal feature modeling can be realized, the problem of mismatching of a strategy and a load is solved, and index expansion and delay peak are avoided.