114 results about "Table (information)" patented technology

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of financial science and technology, medical health and the like, and discloses a natural language driven table processing method, device, equipment and medium. And generating and executing an operation action by utilizing the reinforcement learning strategy network, generating a reward signal according to execution result data and user feedback information, storing the state vector representation, the target table operation action and the reward signal into an experience playback queue, and updating network parameters of the reinforcement learning strategy network based on historical data in the experience playback queue. According to the method, the state vector is constructed through the reinforcement learning strategy network in combination with the natural language instruction and the table context, and the operation strategy is continuously optimized according to the operation result and the user feedback, so that intelligent understanding and action planning of the spreadsheet operation intention are realized, and the data processing efficiency and the interaction intelligence level of non-professional users are improved.

According to the data backup method and device provided by the specification, description information corresponding to target data is obtained, and the description information comprises at least one of data change frequency, health state information of a storage medium stored in the target data or table association information; the table association information comprises a plurality of tables used for storing the target data and a dependency relationship among the plurality of tables, and a target backup strategy is intelligently generated through a pre-trained backup strategy model, so that backup of the target data according to the target backup strategy is more intelligent. The target backup strategy is intelligently obtained according to the data change frequency of the target data, the health state information of the storage medium and / or the table association information instead of a fixed backup strategy, so that the data can be effectively protected by performing data backup through the target backup strategy generated by the method in the specification.

The invention discloses a complex table semantic understanding method and device based on a large language model, and the method comprises the steps: firstly carrying out the preprocessing of a document through a table detection model and a table structure recognition model, and extracting the structure information of a table in the document; secondly, reconstructing the structure of a complex table through a large language model, and constructing two tree structures according to row and column information; and on the basis, semantic translation is performed on the reconstructed table through a preset rule, and a sentence which most conforms to logic is screened out through a large language model, so that deep understanding and accurate expression of complex table contents are realized. The invention provides a new method for processing complex table semantics, the table processing efficiency and accuracy are improved, and valuable references are provided for practical application scenes such as data mining, information retrieval and intelligent questions and answers.

The invention discloses a method for designing and training a two-dimensional table data multi-modal model. According to the method, firstly, a two-dimensional table data multi-modal model is designed, the two-dimensional table data multi-modal model comprises a table representation module, and the table representation module is composed of a plurality of serial bidirectional attention modules; the table representation word embedding set can interact with other related cells through processing of the bidirectional attention module, and global structured table semantics are captured; pre-training a table representation module, and training the table representation module by using a large amount of original table data to enable the table representation module to output accurate table representation information; and finally, carrying out alignment training on the table representation module and the large language model decoder, and carrying out joint supervision fine tuning by adopting the table representation module and the large language model decoder. According to the method, the table modality and the text modality can be jointly input into the large model to carry out table question-answering, data analysis and other generative tasks, and the effect and performance of two-dimensional table question-answering, code generation and other tasks are effectively improved.

The invention discloses an intelligent question and answer method and system for a table-containing document, and the method comprises the steps: firstly obtaining a document, carrying out the data preprocessing of the document, and obtaining a preprocessed table D and text data F; secondly, processing the table data through a large language model to respectively obtain table nodes and table entity nodes; and then the text data F is cut according to paragraphs, each text segment is embedded by embedding the large language model B, text nodes are generated, and a knowledge graph is constructed through recursion based on the text nodes, table nodes and table entity nodes. And finally, according to the knowledge graph, optimizing user query and maximizing retrieval information based on iterative retrieval. And constructing related texts and user queries into structured information, returning the structured information to the large model M, and generating answers. According to the method, documents in various forms can be better processed, and the precision is higher when intelligent question answering is carried out on the document containing the table.

The invention provides an automatic spreadsheet operation and reasoning method and system based on a large language model, and the method comprises the steps: S1, uploading a spreadsheet file, and extracting related information; s2, the table state, the table description and the system prompt of the information provider module are provided for the information provider module; s3, generating an SQL query statement, executing the SQL query statement, and generating a sub-table corresponding to the spreadsheet; s4, extracting forward data from the sub-table, and providing the forward data to the planner module in combination with a table state, table description and a system prompt of the planner module; s5, generating a Python code through a planner module, and executing the Python code in an interpreter to obtain an execution result; s6, judging and skipping according to an execution result until the code is successfully executed; and S7, repeating the steps S1 to S6 until the task is completed. The method can effectively improve the capability and efficiency of automatically solving spreadsheet operation and reasoning tasks.

The invention provides a form text boundary adaptive information extraction method and system in the technical field of computer information processing. The method comprises the following steps: S1, identifying a title frame from a form image; s2, calculating the aspect ratio of the title frame to determine the title arrangement direction of the table titles; s3, detecting a text detection box of the table image, and performing classification prediction of a text arrangement direction on the text detection box by combining a text direction model with a title arrangement direction; s4, controlling affine transformation of the table image through the text arrangement direction and the text detection box to obtain a corrected image; s5, identifying a table title in the corrected image, and matching a table type from the table type configuration file based on the table title; s6, extracting keywords from the corrected image based on the table type; and S7, processing the keyword to extract table information. The method has the advantages that the adaptability, accuracy and efficiency of table information extraction are greatly improved.

The invention discloses an enterprise stock equity atlas construction method based on multi-dimensional data verification and a storage medium, and relates to the technical field of knowledge atlas.The enterprise stock equity atlas construction method comprises the steps that a text is extracted from obtained document data, layout analysis and table structure reconstruction are conducted on the extracted text, and the text obtained after format conversion is input into a large language model; extracting the enterprise entity information and stock equity relationship in the text; constructing a preliminary equity atlas; detecting a loop in the preliminary equity atlas, performing benefit community identification on the equity atlas after contradiction resolution, and outputting a healthy equity atlas; enterprise entity information is extracted in real time, credibility verification is carried out, and the information is stored in a database. A set of complete automation technology chain is formed from data acquisition and OCR table reconstruction to LLM extraction, map construction, contradiction resolution, benefit community recognition and health map comparison, and the manual operation cost is greatly reduced.

The invention provides a question and answer method, a table question and answer model training method and device, equipment and a medium, and relates to the technical field of artificial intelligence. The method comprises the following steps: selecting a target table of which table information is matched with a natural language question from a plurality of tables to be selected, wherein the tables to be selected are in a coding format; generating a corresponding table question based on the natural language question and table information of the target table; and inputting the generated table question into a pre-trained table question and answer model to obtain a question and answer result corresponding to the natural language question. According to the technical scheme, the utilization rate of the table information in the table question and answer model can be increased, and the complex table and multi-table question and answer ability is improved.

The invention relates to the technical field of the power industry, and discloses a complex table structure identification method, system and device for the power industry and a medium, and the method comprises the steps: capturing key information, such as visual, text and layout features, of a target to-be-identified table image through a first feature extraction operation; and a data basis is provided for subsequent feature fusion and reasoning. The second feature fusion operation integrates different features, and the first stage strengthens the relation between vision and text features, so that the recognition is more accurate; in the second stage, layout features are fused, and the overall understanding of the table structure is enhanced. In the context aggregation step, a logic relation of table image elements is established, and recognition continuity and accuracy are improved. The line-column and cell relation between elements is analyzed through table image structure reasoning, and the recognition accuracy is ensured. According to the method, efficient and accurate identification of the complex table structure in the power industry is realized, and the automation degree and efficiency of table data processing are improved.

The invention discloses a table processing method and device, a medium, equipment and a product. The method comprises the steps of obtaining a table image of a first table; the table image is processed based on a first multi-modal big language model to obtain a first code, the first code is used for generating a second table corresponding to the first table, and the first multi-modal big language model is used for determining first table information of the first table according to the table image, the first code is generated according to the first table information, and the first table information comprises table structure information and / or cell content information of the first table; and generating the second table according to the first code. By adopting the scheme, the first code is output end to end according to the table image through the first multi-mode large language model, and the accuracy of table processing is improved.

The invention discloses a one-dimensional automatic processing method and system for a complex header of a multi-dimensional table, and the method comprises the steps: (1) extracting table data in a target document, and forming a set T; (2) for any table ti in the T, acquiring basic information of all header cells, and storing the basic information into LC; (3) calculating the hierarchy depth Li of the header, if Ligt; if so, executing the step (4); otherwise, executing the step (6); (4) taking each header cell in the LC as a node, gradually backtracking upwards from the Li layer to construct a header tree structure, merging text contents of all nodes layer by layer from a root node, and taking the merged text contents as a one-dimensional header of a column where the bottommost layer node is located; (5) removing redundant information in the one-dimensional table header to generate a normalized one-dimensional table; (6) generating a normalized one-dimensional table; and (7) circularly executing the steps (2)-(6) until processing of all table data is completed. The method is high in efficiency and high in accuracy.

The invention discloses a table retrieval method and device and electronic equipment. The method comprises the steps of obtaining a table retrieval request of a target object; in response to the table retrieval request, a target table retrieval library with the same category information as the to-be-retrieved table is determined, and the target table retrieval library comprises a plurality of tables with the same category information as the to-be-retrieved table; at least one candidate table is retrieved from a target table retrieval library, and a first matching degree between table description information of the candidate table and the table retrieval request meets a first matching degree threshold value; based on the table description information corresponding to the at least one candidate table, a target table is determined from the at least one candidate table, and the first similarity between the table description information corresponding to the target table and the table retrieval request meets a first similarity threshold value; and pushing the target table to the target object. According to the method and the device, the technical problem of inaccurate table retrieval in related technologies is solved.

The invention provides a table data processing method and related equipment. The method comprises the following steps: in response to a first operation on a first table, generating an isolation environment and a second table arranged in the isolation environment; the second table comprises table structure information and / or table data information of the first table; in response to a table configuration operation on the second table, table change information is generated; and in response to a deployment operation in the isolated environment, deploying at least part of the table change information to the first table. According to the table data processing method and the related equipment provided by the invention, modification of the table data in the isolation environment does not influence the table data in the production environment, so that the development and test cost of the table is reduced, and the safety of the table data is guaranteed.

The application discloses a method for automatically inferring semantics of a data cleaning script, comprising loading an original table to the data cleaning script and obtaining an input table and an output table; constructing a table change space, obtaining a table change characteristic set based on comparison between the input / output table and the table change space; taking data conversion operation types that all satisfy inevitable occurrence, impossible occurrence and possible occurrence as candidate data conversion operation types; converting and splicing multiple descriptive texts corresponding to function names, function parameters and table change characteristics to obtain a semantic vector, inputting the semantic vector into a twin convolutional neural network model to obtain candidate data conversion operation type ranking; constructing a parameter slot of each data conversion operation type, inferring the parameter slot of the candidate data conversion operation type by using data cleaning code, table content and table change information of a corresponding row, and obtaining the type and parameters of the data conversion operation based on the inference result. The method can simply and conveniently infer the type and parameters of the data conversion operation in the cleaning code.

The invention relates to a table processing method and electronic equipment. The method comprises the steps of obtaining a to-be-processed worksheet file; determining, through a large language model for table understanding, structured information of the worksheet file, the structured information comprising at least one of the following items of a worksheet for the worksheet file: information of a logical sub-table, row description information, or grouping information; and storing the structured information using a predefined format. In this way, the embodiment of the invention provides a method for understanding the structure of the table, and deep understanding of the internal semantic structure of the table is achieved.

Arrangements for partition handling for multi table scan semi join operations are provided. Partition information of a first logical relation may be determined. Partition information of a first physical operator may be determined based on the partition information of the first logical relation. The partition information of the first physical operator may be stored in a data structure which includes partition information of a fact table and of each reducer in a predefined sequence. Partition information of a second logical relation may be determined. The partition information of a second physical operator may be determined based on the partition information of the second logical relation and the partition information of the first physical operator. The partition information of the second physical operator may be stored in the data structure. Execution of the second physical operator may be performed using the stored partition information, skipping mappings between pairs of partitions.

We provide the system. [Solution] Information processing means for receiving old and new data, A text extraction means for extracting text data from old and new documents, A difference detection means that compares the text data of old and new documents and detects the differences, A table generation means that classifies the detected differences and generates a revised table, An output means for outputting the generated current revised table, A system that includes this.

The invention relates to the technical field of natural language processing, in particular to a pointer table field recall method, device, equipment and medium, in the method, human query statements and multiple pieces of table field information of the same table are spliced into a text sequence, and a lexical element vector of each lexical element in the text sequence is calculated by adopting self-attention; context information of the lexical elements is considered, lexical element vector expression information is richer, multiple lexical elements are calculated at a time, and efficiency is higher; determining a whole sentence semantic vector formed by each lexical element vector corresponding to the human query statement and each lexical element vector corresponding to each piece of table field information; and parallel cross attention calculation is performed on the whole sentence semantic vector and the lexical element vectors corresponding to the multiple pieces of table field information, the matching degree of each piece of table field information and the human query statement is determined, and the accuracy and efficiency of table field recall are improved.

The invention discloses a table biaxial position coding and task loss function combined large model table question and answer method. Collecting table, question and answer data, and performing structuralization and biaxial position coding processing on the table data to obtain a text word vector set and biaxial position codes; performing fusion processing to obtain a position code after biaxial fusion; after combination processing, sequentially inputting into a model for processing to obtain a table global representation vector; and establishing a loss function training model, and performing prediction processing on an actual application scene by using the trained model to obtain a table answer result. According to the method, through double-axis position coding, the two-way constraint relation between rows and columns is effectively captured, representation is made to better conform to the inherent structure of a table, through mask cell modeling, known data can be learned, missing information can be speculated, the incomplete data processing capacity in a real scene is improved, and the method is suitable for being popularized and applied. The accuracy and robustness in complex table data processing can be remarkably improved.

The application discloses a table information extraction method, system and device based on a compressed pre-training language model. The application models the table information extraction as a reading comprehension problem and inputs the pre-training language model, thereby avoiding error propagation and reducing the participation of rules. Meanwhile, the strong generalization ability and robustness of the pre-training language model can obtain relatively accurate cells in the table that can answer natural language questions, thereby completing the extraction of the associated information in the table.

The table reconstruction method, device, computer device and readable storage medium provided by the present invention, the method includes: obtaining a table image; detecting and recognizing the table image to obtain the text content and layout information corresponding to each of a plurality of text boxes, as well as the coordinates of row table lines and the coordinates of column table lines; wherein, the layout information includes position information, row numbers and column numbers; performing table reconstruction according to the text content, the layout information, and the coordinates of the row table lines and the coordinates of the column table lines. In the present invention, the row number and column number corresponding to each text content are obtained. Therefore, the table can be reconstructed by combining the row number and column number and the detected table lines. Compared with the prior art that only relies on the detected table lines to reconstruct the table, the present invention can accurately restore the actual structure of the table and improve the accuracy of reconstructing the table.

According to the training method for the table understanding ability of the large language model, effective information of the table can be compressed, meanwhile, row and column information capable of improving the effect is reserved, the effect can be improved at the same time, the effect improvement is embodied in the fact that the same table content is expressed in the form, the number of required tokens is generally small, and the method is easy to implement. The method has the advantages that the number of tokens participating in training and reasoning is smaller, computing power is utilized more efficiently, three major capabilities, namely table reading, numerical reasoning and data analysis, of table understanding are considered, the generation probability of correct answers is further increased, and generation of wrong answers caused by the probability is reduced. According to the method, the accuracy of table understanding is improved, and the information density after token conversion is effectively represented, optimized and improved through the table data, so that the efficiency of table data training and reasoning is improved.

The invention discloses a table structure extraction method based on a deep cascade network, and the method comprises the steps: obtaining a table image, extracting the visual features of the table image, coding the visual features to obtain enhanced features, and obtaining target query information according to the enhanced features; according to target query information, decoding the enhanced features to obtain a prediction result of a text region bounding box; decoding the prediction result of the text region bounding box to obtain a prediction result of a cell region bounding box; and performing logical position allocation according to a prediction result of the cell region bounding box to obtain table structure content.

The invention discloses an SQL annotation generation method, a computer readable medium and electronic equipment, and the method comprises the steps: obtaining a table list in an initial statement, and loading a table structure corresponding to the table list; the table list refers to a list of table structures contained in the initial statement; expanding the initial statement based on the table structure to obtain a fully expanded field of the initial statement; performing grammar analysis on the fully-expanded field, and determining a sub-query list in the fully-expanded field; the sub-query is a query clause in a fully expanded field; according to the structure and the type of the sub-query, optimizing a field corresponding to the sub-query to obtain a statement to be annotated; adding corresponding annotation information to a target position of the to-be-annotated statement based on a table structure and a sub-query contained in the to-be-annotated statement; therefore, the readability of the code statement can be improved and the logic of the code statement is clearer by splitting the multilayer nested SQL initial statement, intelligently identifying the table and the sub-query to which the field in the statement belongs and adding the corresponding annotation information to the target position.

This invention relates to a table-based question-and-answer method and apparatus. The table-based question-and-answer method includes: structuring a table in a technical standard document to obtain table information; receiving an input user question and processing it through a preset model to obtain an SQL statement; and querying the answer corresponding to the user question based on the SQL statement and the table information.

This application relates to the technical fields of data processing and artificial intelligence, and discloses a method and apparatus for evaluating and processing retrieved tabular documents. The method includes: parsing the tabular document to obtain at least its table structure information, data content, and metadata information; performing multi-dimensional evaluation based on the table structure information, data content, and metadata information to obtain at least two of the following: table standardization evaluation result, metadata integrity evaluation result, data clarity evaluation result, and contextual relevance evaluation result; and weighting at least two of these results to obtain a comprehensive evaluation result for the tabular document. This method allows for a comprehensive evaluation of tabular documents, providing users with clear optimization guidance.

The invention discloses a complex table structure understanding and information extraction method, system and device and a medium, belongs to the cross technical field of artificial intelligence, computer vision and natural language processing, and aims to solve the technical problem of how to realize accurate understanding and information extraction of a complex table structure. According to the technical scheme, the method comprises the following steps: multi-modal feature extraction: representing table data as a combination of a visual feature vector, a text semantic vector and a layout feature vector, namely, a multi-modal table vector = the visual feature vector, the text semantic vector and the layout feature vector; multi-modal fusion: generating a multi-modal table vector based on the multi-modal large model architecture, performing efficient processing and fusion on the multi-modal table vector, and fully fusing multi-dimensional information of a visual feature vector, a text semantic vector and a layout feature vector; structure perception decoding and semantic enhancement reconstruction are carried out; an adaptive output is generated.

The invention provides a code generation method and system based on a table structure, a server and a medium, and relates to the technical field of computers.The method comprises the steps that table structure analysis is conducted on a data table in a preset data source, and table structure meta-information is obtained; performing field semantic recognition on the table structure meta-information according to a preset rule base, and determining a front-end control type and a back-end verification rule; according to a preset generation strategy, table structure meta-information, a front-end control type and a back-end verification rule, calling a code template to execute rendering processing, and generating an original code; and carrying out front-end and rear-end interface binding on the original code to generate a target code. The project development efficiency and the code quality can be effectively improved.

This invention relates to the field of computer vision technology and discloses a method, apparatus, device, and readable storage medium for table recognition. The method includes: acquiring an image to be recognized; inputting the image to be recognized into a preset table detection model to identify a table image from the image; inputting the table image into the preset table recognition model to determine the endpoint coordinates of each cell in the table image and the connection relationships between the endpoints; and generating a target table based on the endpoint coordinates and connection relationships. By implementing this invention, table recognition is achieved through endpoint extraction and endpoint connection relationship extraction, eliminating reliance on table line extraction, ensuring the accuracy of table structure recognition, and improving the accuracy of table information recognition.