39results about How to "Improve Data Retrieval Efficiency" patented technology

The invention provides a classifying map method for industry information based on XML (extensive makeup language). The method utilizes XML to describe industry information classification standard and information unit data, and establishes a classification tree according thereto, dynamically extracts database inquiry code according to user needs, and utilizes a standard database procedure execution mechanism to improve the work efficiency of an industry information application system. The method comprises the following steps: (1) loading a description file and establishing an industry information classification tree; (2) driving traversal information classification tree to obtain classification node by information needs; (3) dynamically loading the database inquiry code when the node state indication is false, and setting the node state indication as true; (4) directly accessing the inquiry code corresponding to the node when the node state indication is true; (5) and conducting data inquiry procedure and acquiring data set. The work flow of the method is displayed as the above diagram. The method realizes the mapping to the industry information with large data amount and multiple classifications with the mode close to the need of industry user, thereby effectively increasing data retrieval efficiency and strengthening system expansibility.

The invention relates to the technical field of block chains, and discloses a block chain-based rapid data retrieval method, which comprises the steps of obtaining to-be-stored data, and performing encoding processing on the to-be-stored data by utilizing a pipeline-based data encoding scheme to obtain encoded data; storing the coded data into a block chain, and storing a copy of the to-be-storeddata into a block chain node by utilizing a time-based copy storage method; slicing the data stored in the blockchain by utilizing a data slicing mode based on the erasure code to obtain sliced data based on the erasure code; utilizing an index construction method based on a time sequence to construct a time sequence index; and utilizing the time sequence index of the data to quickly retrieve thedata. The invention further provides a rapid data retrieval system based on the block chain. According to the invention, data retrieval is realized.

The invention discloses a Solr-based Chinese search method for efficient full-text search by using Chinese. The Solr-based Chinese search method comprises the following steps: a SolrClound platform receives an inquire statement created by a Solrj client, the Solrj client is used for creating the inquire statement according to an inquire request of a user, and a plurality of Chinese index databases based on Chinese word segmentation are created in the SolrClound platform; the SolrCloud platform performs Chinese word segmentation on the inquire statement to obtain a plurality of Chinese words after Chinese word segmentation; the SolrClound platform separately creates corresponding search indexes according to the plurality of Chinese words; the SolrClound platform performs full-text search on the plurality of Chinese index databases in the SolrClound platform by using the search indexes which are separately corresponding to the plurality of Chinese words; and the SolrClound platform acquires a search result from the plurality of Chinese index databases and sends the search result to the Solrj client.

The invention relates to a millimeter-wave radar target speed ambiguity resolution method. The millimeter-wave radar target speed ambiguity resolution method comprises the steps of: S1, selecting a transmitting antenna, and adding a same linear frequency modulation signal with a time delay a before continuous linear frequency modulation signals are transmitted; S2, performing 2DFFT processing on the received signal according to the signal of the transmitting antenna which is received by the receiving antenna and is reflected by the target; S3, performing target detection according to a processing result of the 2DFFT and calculating a first fuzzy speed estimation value; and S4, calculating to obtain a second fuzzy speed estimation value according to the linear frequency modulation signal and the 2DFFT processing result of the linear frequency modulation signal with the time delay a. Through the scheme provided by the invention, the real speed of the target can be obtained.

The invention relates to a quick query method suitable for a research and development database. The method includes the steps that a storage server matches a retrieval label for the content in the database; a retrieval server conducts association correlation on the retrieval label; the retrieval server conducts retrieval hint on the retrieval label meeting the repeated appearance frequency requirement; the storage server stores a retrieval result and uploads the result to a publishing server; the publishing server synchronously sends the retrieval result to a cloud server and a publishing terminal. Therefore, quick matching of database content can be achieved by arranging an independent retrieval label. Label association can be achieved, and data retrieval efficiency is promoted. Retrieval word frequency prompt can be achieved, so that the retrieval efficiency is further improved conveniently. Retrieval data can be stored, and data safety is improved.

The invention provides a data retrieval method and device. The method comprises the steps of receiving a search request initiated by a terminal; determining a field type matched with the search request, wherein the field type is obtained by identifying query information contained in the search request based on a preset rule; according to the field type and a field marking type for pre-marking eachfield in the data table based on a preset rule, determining a field corresponding to the field marking type matched with the field type in the data table as a target field; and performing retrieval in the target field based on the query information. According to the technical scheme, the data retrieval efficiency can be improved.

The invention relates to a tape library based data protection system and method. The data protection system comprises a data backup server and a tape library, wherein the data backup server comprises a backup data storage module, a data archiving module and a position database, the backup data storage module is used for storing data backed up on a production server and a corresponding time point data set; the data archiving module is used for establishing a data archiving task, archiving data stored in the backup data storage module to the tape library according to the data archiving task by taking the time point data set as a unit and also recording the position of each data object in the tape library; the position database is used for storing positions of the data objects in the tape library. Compared with the prior art, the tape library based data protection system and method can effectively protect data, can reduce the cost by utilizing advantages of a tape and also facilitate management of data in the tape.

The invention provides a judicial assistant mechanism reputation evaluation method, system and device and a storage medium. The method comprises the following steps: obtaining historical reputation evaluation data of each judicial assistant mechanism; establishing a judicial assistant mechanism reputation evaluation model according to the historical reputation evaluation data; obtaining a static evaluation score of each judicial assistant mechanism based on a reputation evaluation model according to the historical reputation evaluation data; obtaining real-time reputation evaluation data of each judicial assistant mechanism in a set time period; obtaining a dynamic evaluation score of each judicial assistant mechanism based on a reputation evaluation model according to the real-time reputation evaluation data; and calculating a comprehensive evaluation score of each judicial assistant mechanism according to the static evaluation score and the dynamic evaluation score of each judicial assistant mechanism. The evaluation model is constructed by combining the collected historical data and big data analysis, and the judicial assistant evaluation method is given by collecting and processing the historical and real-time data, so that the efficiency of selecting the assistant by a judicial organ is improved.

The invention relates to a network model generation method and device, a computer readable storage medium and a logistics system. The network model generation method comprises the steps of S1, obtaining transportation data of a historical order; step S2, screening out key features based on the transportation data, the key features including departure place, destination and transportation associated information; and step S3, based on the key characteristics, generating a corresponding transportation relationship between the nodes of the network model and the nodes. Through the scheme provided by the invention, the order searching speed can be improved.

The invention provides an operation behavior recording method and system applied to an operation behavior auditing system. The system comprises a violation rule definition module, a violation rule account binding module, an account auditing policy definition module, an operation behavior content data acquisition module, an account violation rule acquisition module, a violation judgment module and an auditing data recording module. The operation behavior recording method and system have the advantages that (1) an operation behavior is judged from the fine grit to whether the content of a behavior is violated or not, and it is defined that different auditing policies are adopted when a violation behavior and a non violation behavior occur, so that compared with a conventional auditing operation system, on the premise of meeting auditing requirements, the recording quantity of auditing data can be remarkably reduced and the occupation of storage space is reduced; and (2) violation data and non violation data are stored by partitioning, so that the retrieval can be carried out more efficiently and the data retrieval efficiency can be improved.

The invention discloses a high-load service storage and retrieval method for non-relational data storage. The storage process is to receive a request for creating/updating non-relational data, construct metadata information for the non-relational data, put the metadata information into a buffer and store the metadata information permanently. Gets the index key information, builds the index information, caches the index information, and stores it permanently. The retrieval process receives the data retrieval request, obtains the metadata information from the cache, and returns the metadata information. Searches for index information based on the metadata information and returns the index information. For a large amount of data, the invention can realize efficient storage and retrieval, andis simple and reliable, easy to operate and maintain, and low in cost.

The invention provides a multi-dimensional automatic vehicle identification method, which comprises the following steps that: S1, multi-dimensional roadside sensing equipment acquires multi-dimensional data of a vehicle, and sends the multi-dimensional data and time and position information for acquiring the data to a sensing data processing platform; S2, after the sensing data processing platform gathers the multi-dimensional data sent by the multi-dimensional roadside sensing equipment at different point locations within a period of time, performing feature evaluation on each piece of multi-dimensional data through a data dimension feature evaluation unit; and S3, on the basis of each multi-dimensional data feature evaluation, a multi-dimensional data association construction unit realizing the construction of an individual vehicle multi-dimensional data association relationship through observation and comparison of individual vehicle data diversity, and giving the confidence of the association relationship through an association confidence calculation unit, thereby realizing one-vehicle one-gear taking the vehicle as the center.

The invention relates to a noise reduction method for a medium-voltage uninterruptible power supply device, which belongs to the technical field of noise reduction and comprises the following specific steps of: (1) installing monitoring equipment and detecting a synchronization rate; (2) information of each group of uninterruptible power devices is collected and classified; (3) collecting noise data of each group of devices for analysis; (4) performing noise reduction feedback on each group of devices; by constructing the classification storage library, the missing data can be subjected to supplementary feedback, the integrity of data collection is ensured, meanwhile, the retrieval data set is constructed, the data retrieval efficiency of workers is improved, redundancy screening can be performed on noise data generated by each group of uninterruptible power devices, the accuracy of the collected noise data is improved, and the data retrieval efficiency is improved. The accuracy of the fed-back noise decibel data of each group of uninterruptible power devices is ensured, and a worker can check and analyze the noise decibel data conveniently.

The invention provides a hash mapping storage method for second flow thickness control. According to the method, second flow thickness control calculation is realized, a data structure of a hash tableis used, inlet thickness measurement values needed in the second flow calculation process are stored in the hash table with the fixed length, the advancing length of a strip serves as a keyword, a hash function is used for calculating a storage index, the thickness value of an inlet entering a roll gap at present is retrieved, and the method participates in calculation of a second flow controller. According to the method, a traditional first-in first-out queue data structure is replaced by a hash table data structure, when the currently used entrance thickness value is retrieved, only one-time calculation is needed, the shifting operation needed in the first-in first-out queue is omitted, the retrieval efficiency is high, the calculation time is short, and the method can be used for entrance thickness calculation in second flow thickness control.

The invention relates to the technical field of transaction message storage and reading, and discloses a method for storing a transaction message based on a MongoDB database, a gateway layer is mainly used for receiving a transaction request sent by a client, generating a unique serial number from the request, temporarily storing the serial number of the request to a context environment of the transaction request of the gateway layer, obtaining a request message and a response message by the context environment, and the storage mode of the gateway layer and the MongoDB adopts an asynchronous thread form. According to the method, the request message and the response message in the transaction message are converted into the MongoDB document format in an asynchronous mode, the MongoDB document format is stored in the MongoDB database through the asynchronous thread, the execution process and the execution efficiency of the original transaction are not affected in the storage process, the stored data volume can support more than a hundred million level record number through a sub-library sub-table mode, each request message and each response message can be quickly retrieved and associated through different retrieval conditions, so that the data retrieval efficiency and the message checking accuracy of problem backtracking are remarkably improved.