8 results about "Position Number" patented technology

The embodiment of the invention provides a distributed data deduplication method and product. The method comprises the steps of obtaining a data key of data to be subjected to deduplication; distributing the data to be subjected to duplicate removal to the ith sub-bucket according to the mixed hash value of the data key; mapping the mixed hash value according to a hash function attribute value configured for the ith sub-bucket to obtain k position numbers; reading numerical values of positions corresponding to the k position numbers from a bit array which is stored at the current moment and corresponds to the ith sub-bucket to obtain k numbers; if it is confirmed that the k numbers are all first numerical values, it is confirmed that the data to be subjected to duplicate removal belong to duplicate data; and if at least one of the k digits is confirmed to be a second numerical value, confirming that the data to be subjected to duplicate removal belongs to new data. According to the embodiment of the invention, the speed and accuracy of big data deduplication processing are remarkably improved, and the problem of single-node calculation overload (different buckets can be distributed to different calculation nodes) is effectively solved.

According to the bit weight amplitude weight quantizer and the quantization discretization method, the maximum amplitude of input continuous simulation information is discretely divided into a plurality of amplitude paragraphs of a content interval GS, the amplitude paragraphs are associated with a carry system, the amplitude of the input continuous simulation information is discretely divided into a plurality of paragraphs through a number of carry systems, and the amplitude of the input continuous simulation information is divided into a plurality of paragraphs. If a decimal system exists, the amplitude of the continuous analog information is discretized into ten paragraphs, the discretized paragraphs are converted into thermometer coding information through a comparator, and then the thermometer coding information is converted into bit weight information through XOR and XNOR logical operation; the value of the bit weight in the bit weight information is generated by comparing the binary states 0 and 1 on the node, one state on the unique node is compared with different states on other nodes to indicate that the node has the weight, and the position number of the node with the weight is used for indicating the bit weight of all the nodes in the bracket.

The application discloses a kind of intelligent inquiry method and device in refining field, the method includes: obtaining user asking data;Judge user asking data is real-time data asking data demand or index asking data demand;Judge is real-time data asking data demand data, execute: determine intent type;Generation word slot matching result;Determine position number information;Generate first SQL sentence;Execute first SQL sentence, output the first feedback result of user asking data;Judge is index asking data demand data, execute: complexity judgment is carried out to user asking data;According to complexity judgment result, generate second SQL sentence after execution, output the second feedback result corresponding to user asking data, can improve the accuracy, efficiency and generalization ability of intelligent inquiry in refining field, reduce cost, improve user experience.

This invention discloses an index structure for persistent storage, belonging to the database field. It includes a deletion flag bit vector and a first and second layer connected by a logical block index. The logical block index is an index constructed from logical blocks, where a logical block is k data blocks obtained by dividing a temporary array containing N data points arranged in a predetermined order. The first layer includes k+1 bit vectors, each containing N bits indicating whether a certain data point in the original array exists in a certain logical block. The second layer includes a sequence of position indicators with N position numbers, representing the position number of the corresponding data in the temporary array within the original array. The deletion flag bit vector includes N bits indicating whether the corresponding data in the original array has been removed. This invention avoids updating the original file's index structure while ensuring the index's validity.

The application discloses a communication data query method and terminal based on a distributed network, and after data in the distributed network is acquired, a data formula is matched for the data, the data is summarized by using the data formula to obtain abbreviated data, and a dictionary template corresponding to each data is established based on a dictionary index corresponding to the data, the abbreviated data, the data formula and a position number of the dictionary index in a dictionary data cluster, then the corresponding dictionary template can be found according to a data query request, and the data to be found can be restored according to the dictionary template, so that the efficiency of completely finding data in the distributed network is improved.

The invention discloses an intelligent question and answer method and device in the refinery field. The method comprises the steps of obtaining user question data; judging whether the user question data is a real-time data question number demand or an index question number demand; determining real-time data asking demand data, and executing the following steps: determining an intention type; generating a word slot matching result; determining bit number information; generating a first SQL statement; executing the first SQL statement, and outputting a first feedback result of the user question data; if the data is the index question number demand data, performing complexity judgment on the user question data; according to the complexity judgment result, executing the second SQL statement after generation, and outputting a second feedback result corresponding to the user question data, so that the accuracy, efficiency and generalization ability of intelligent question answering in the refinery field can be improved, the cost is reduced, and the user experience is improved.

The application discloses a data retrieval method and device, and relates to the field of databases.The method comprises the following steps: searching for a first secondary index item in a first secondary index page at a first position number, wherein the first secondary index item comprises time range information of a first primary index page and storage location information of the first primary index page, the time range indicated by the time range information comprises a target time point, obtaining the first primary index page based on the storage location information of the first primary index page, searching for a first primary index item in the first primary index page, wherein the first primary index item comprises the target time point, and obtaining data of the first position number at the target time point according to the first primary index item.The method is equivalent to reading multiple primary index pages by reading one secondary index page, the first primary index page is quickly located, and therefore, the data of the first position number at the target time point is quickly retrieved, and the data retrieval efficiency is improved.

The invention provides a large model reasoning control method and device, equipment and a medium, and relates to the field of artificial intelligence. Dividing a to-be-reasoned text input by a current user into text segments; generating a unique identifier for the text content of the text fragment to obtain a current identifier; and generating a current position number according to the position of the text fragment in the to-be-reasoned text. Extracting target historical reasoning information corresponding to a target historical identifier which is the same as the current identifier from a historical cache; the target historical reasoning information comprises a target historical position number, a target historical key and a target historical value, and the target historical key is generated according to the target historical position number and the target historical text content. Generating a target key according to the target historical text content and the current position number; and taking the historical value as a target value. And the specified large model performs reasoning operation on the to-be-reasoned text based on the target key and the target value to obtain a reasoning result. The reusable range of the cached historical reasoning information is improved, repeated calculation is reduced, and the reasoning efficiency is improved.