7 results about "Relation (database)" patented technology

The application relates to the technical field of intelligent question and answer interaction, in particular to an intelligent question and answer interaction method fusing a private database and an enterprise knowledge graph, which comprises the following steps: in view of the influence of enterprise-specific terms and implied semantics on the connection of relations, semantic similarity, co-occurrence frequency characteristics and syntactic dependency degree characteristics in homologous text data are analyzed in depth to reflect the credibility of the three tuples formed between entity subwords; in view of the influence of the multi-source nature of enterprise data, the consistency of enterprise identification and the semantic equivalence characteristics between entity subwords are analyzed to analyze the confidence of the connection relationship between entity subwords in different source data; the completeness of the construction of the enterprise knowledge graph is improved, and the defects that the returned answers of the intelligent question and answer are not accurate enough and the fluency is low are remedied.

This application relates to a method and system for managing a large-scale model corpus knowledge graph. The method includes: collecting multi-source natural gas corpora to obtain a clean text set; extracting candidate entities using a large language model, and iteratively obtaining a list of valid entities through domain-strong rule validation; constructing relation extraction instructions to drive the large model to output candidate triples, and obtaining the final triple set through JSON repair and differential correction. An ontology hierarchical structure is constructed, mapping entities to triples to generate a knowledge graph, which is stored using a hybrid approach of relational databases, graph databases, and vector databases. Based on the graph, the method extracts the time-series and descriptive features of users' daily consumption, outputting user categories through k-shape clustering and fuzzy C-means clustering; performing STL decomposition on the time series of each user category, then predicting and reconstructing using a temporal convolutional network to obtain short-term consumption prediction values, which are fed back to the knowledge graph; and dynamically updating knowledge through vector similarity and ontology constraints when new corpora are added.

This invention discloses a metadata-driven adaptive Text-to-SQL method, relating to the field of natural language processing technology. It involves: initially screening pattern items based on database metadata to obtain a preliminary pattern subset; then, obtaining a simplified pattern subset through disambiguation verification of candidate columns; constructing a weighted set of join constraints based on foreign key relation metadata; determining an anchor table set based on the simplified pattern subset; calculating the reachability strength of each table; selecting and forming a set of joinable closure tables and determining the structurally joinable retrieval domain; establishing an inverted index for columns in the retrieval domain; allocating recall quotas based on the relevance scores of each column and the reachability strength of its respective table; and retrieving and filtering to generate a predicate evidence set. Finally, organizing database prompts based on the simplified pattern subset, relevant join constraints, and the predicate evidence set, and inputting them into an SQL generation model to generate executable SQL. This method adaptively constrains and enhances the Text-to-SQL generation process using metadata.

The application discloses a method, a terminal system and a readable storage medium capable of presenting internal and external correlation of legal provisions, and belongs to the cross discipline field of computer technology and jurisprudence. The method comprises the following steps: establishing data and a database, establishing data and a database formed by setting unique numbers for legal provisions as units; establishing a mapping relationship, converting the correlation between legal provisions into a connection on the data, establishing a mapping relationship between the connection on the data and the legal provisions, storing the connection on the data, establishing a mapping relationship between preset values (hereinafter referred to as user-selectable values) that can be selected by a user and the aforementioned connection on the data, so as to realize the calling of the connection on the data through the user-selectable values, and then realize the calling of the legal provisions having the mapping relationship with the connection on the data; presenting the legal provisions having the correlation, calling the legal provisions through the aforementioned correlation and displaying the legal provisions in a preset mode in the main and auxiliary presentation areas arranged horizontally, so as to present the legal provisions having the correlation, meanwhile, providing the user-selectable values that can be selected by the user, realizing the calling of the correlation having the mapping relationship with the user-selectable values, and then realizing the calling and display of the legal provisions having the mapping relationship with the correlation, so as to realize the display of the legal provisions having the internal and external correlation in the same window, and improve the efficiency.

In an example embodiment, a query plan is modified / generated to insert a new shared subplan relation between the operator-at-issue and its input (e.g., the child operator). The shared subplan relation has two references, one of which is connected to the operator-at-issue and one of which is connected to an aggregation operator. A new operator is then inserted as a parent of the aggregation operator. This new operator sums up incoming counts and updates the resultant sum as a stored variable (e.g., as a session variable).

The application discloses an association long-term memory method based on an entity relation graph, comprising the following steps: S1. stable entity and relation extraction; S2. constructing an entity relation graph and upserting to a graph database; S3. dynamic management of relation maximum impression value and edge distance; S4. relation conflict detection and intelligent correction; S5. batch construction of semantic vector data of entities and indexing to corresponding entities of the graph database; S6. triggering of association retrieval; S7. breadth-first association traversal based on BFS; S8. calculation of relation current impression value; S9. sorting and returning of association results. The application can solve the problems in the prior art, such as lack of association reasoning ability, loss of entity relation information, inability to express multiple relations, inability to quantify the importance of relations, lack of time decay mechanism, separation of vector retrieval and graph retrieval, and lack of relation conflict processing mechanism.