16 results about "Pair matching" patented technology

The present disclosure relates to systems, non-transitory computer-readable media, and methods that generate machine-learning training pairs across disparate modalities for a multi-modal machine learning model learning process. Indeed, in one or more implementations, the disclosed systems generate a first set of perturbation classification scores from a first data modality using a first classification model and a second set of perturbation classification scores from a second data modality using a second classification model. For instance, the disclosed systems compare the first set of perturbation classification scores with the second set of perturbation classification scores to determine a plurality of similarity measures. Moreover, in some instances, the disclosed systems identify pairs of data samples across the first data modality and the second data modality for a multi-modal machine learning model learning process using the pairs of data samples.

The present application discloses a text translation method and system based on multi-language entries. The method comprises: acquiring a multi-language document, and performing character string extraction on the multi-language document to obtain a target character string; performing character string segmentation processing on the target character string to obtain a plurality of entries, performing data extraction to obtain key values of the entries and content values of the entries, and performing key-value pair matching to obtain a plurality of key-value pairs; performing numerical value matching analysis on each key-value pair to obtain a numerical value matching analysis result, and performing correction processing on the plurality of key-value pairs to obtain a plurality of target key-value pairs; performing character string generation on the plurality of target key-value pairs to obtain a plurality of target key-value character strings, and performing format conversion on the plurality of target key-value character strings to obtain a document to be translated; and by means of a language translation model, performing document translation on the document to be translated, to obtain a target document.

The invention discloses a question and answer pair generation method and device, a storage medium and electronic equipment. The question and answer pair generation method provided by the specification comprises the following steps: acquiring a text material for generating a question and answer pair; inputting the text material into a large language model, and enabling the large language model to output a pseudo question and answer pair according to the text material; in response to a received real question input by a user, determining a target pseudo question-answer pair matched with the real question in each pseudo question-answer pair; embedding the target pseudo question and answer pair into a prompt template of the large language model, and inputting the real question into the large language model to obtain an optimized answer output by the large language model; determining the real question and the optimized answer as an undetermined question and answer pair, and performing quality evaluation on the undetermined question and answer pair; and in response to the fact that the to-be-determined question and answer pair passes the quality evaluation, determining the to-be-determined question and answer pair as an available question and answer pair.

The invention discloses an aspect emotion triple extraction method based on bidirectional interaction and reasoning strategy combination, a computer program product and a terminal, and belongs to the field of natural language processing, and the method comprises the following steps: carrying out vector splicing on context representation and word dependence information to obtain new feature representation; extracting candidate spans of aspects and viewpoint sets from the new feature representation; an aspect and viewpoint set is extracted by using a bidirectional interaction module; and a bidirectional matching module is adopted to match the input aspect and viewpoint set to obtain an aspect-viewpoint pair representation, and finally classification processing is carried out. A bidirectional interaction module is introduced, useful information of an aspect span and a viewpoint span is exchanged from two directions of aspects to viewpoints and viewpoints to aspects, and the semantic relationship between the aspects and the viewpoints is enhanced; a bidirectional matching module is introduced, and aspect-viewpoint pairs and viewpoint-aspect pairs are subjected to bidirectional matching by utilizing potential mutual indication relationships among different word pairs, so that the aspect-viewpoint pair matching accuracy is improved.

The invention belongs to the technical field of artificial intelligence, and provides a system operation process description generation method and device based on artificial intelligence. According to the technical scheme, key value pair matching is conducted on element names and display names; interaction information of a user and a service system is analyzed, analysis and feature extraction are carried out on the interaction information, key information and intention of user interaction are obtained, all service process information is stored in the constructed database, and the service process information comprises the interaction information, an analysis result, the key information and intention information; performing feature extraction, feature selection and feature coding on the business process information, and generating a sequence description of the business process by using the trained operation process description generation model; the key value pair matching result is performed according to the element name and the display name, and the sequence description of the business process is converted into the display result to be output, so that the problems of use cost and information security risk are solved, and accurate requirements are provided for RPA business process design.

A lane global car-following pair matching and extracting method of a video track dataset, the vehicle track dataset contains the track point information of each vehicle at each time on the collected road, and has an important role in the research of both macroscopic traffic phenomena and microscopic driving behaviors such as car-following and lane-changing. The method aims to establish a standardized car-following pair matching and extracting method, which can detect and classify the driving behaviors of each track in the track dataset, and can efficiently and quickly identify the car-following pairs in the car-following state within the length range of the entire lane, thereby laying a foundation for in-depth research on driving behaviors.

Systems and methods for entity item matching include, responsive to identifying a workflow, monitoring one or more first entity items of a first entity item type for updates to one or more first field-value pairs; determining, subsequent to processing a first monitored item of a plurality of monitored items, an update to a first field-value pair of a first entity item of the first entity item type; responsive to determining that the update matches a corresponding field-value pair of the first monitored item, matching the first entity item with the first monitored item by performing a lookup in the first monitored item for a field-value pair matching the update to the first field-value pair; and performing a first action specified in the workflow on the first monitored item using data corresponding to the update in the first field-value pair of the first entity item matched with the first monitored item.

The invention discloses an event venation diagram construction method and device, equipment, a medium and a product. The method comprises the following steps: acquiring an event data set of a target event, and respectively extracting an event tag set of each event data in the event data set through a natural language processing model; obtaining the support degree of each event tag in each event tag set, and obtaining at least one frequent tag set matched with the event data set according to the support degree of each event tag; according to the support degree of each event tag in each frequent tag set, obtaining at least one rule tag pair matched with each frequent tag set; and generating an event venation diagram matched with the target event based on each rule tag pair. By means of the technical scheme, construction of the event venation diagram based on the event labels can be achieved, the incidence relation of the event labels is defined, and the comprehensiveness and accuracy of the event venation diagram are improved.

The invention belongs to the field of natural language processing, and particularly relates to a precise data pushing method and system based on a working scene, and the method comprises the following steps: 1, creating the working scene; 2, assigning a work scene to ask questions; 3, preliminary question and answer pair matching; 4, understanding and supplementing problems; 5, deep retrieval and answer integration are carried out; 6, performing answer evaluation and feedback circulation; according to the method, a multi-dimensional working scene framework is constructed, the knowledge category is defined, and the carrying and utilization of working scene identifiers in the question and answer process are realized by collecting and integrating question and answer pair data only aiming at the scenes. The step ensures that the system can immediately identify the working environment behind the system when receiving the query, and further performs deep retrieval in a highly focused and highly targeted knowledge base, so as to output accurate answers highly matched with the query scene instead of wide and non-targeted information.

The application belongs to the technical field of question and answer matching, and provides a question and answer matching method and system based on question pair matching and question and answer pair matching. After obtaining a question, a preset question-question matching model is used to obtain question vector representation in the question and vector representation of questions in a preset question and answer library, and on this basis, the similarity between the question in the question and the question in the preset question and answer library is calculated to preliminarily screen the question and answer library. Then, a preset question-answer matching model is used to obtain question vector representation in the question and vector representation of answers in the screened question and answer library, and on this basis, the similarity between the question in the question and the answer in the screened question and answer library is calculated, and the question and answer pair with the highest similarity score is taken as the final matched question and answer pair of the obtained question. In the question and answer pair matching process, the two times of similarity calculation realize preliminary screening and final determination of the question and answer pair, improve the accuracy of the similarity calculation, and ensure the matching effect.

The embodiment of the invention provides a multi-mode retrieval method and device and electronic equipment. According to the embodiment of the invention, the method comprises the steps: obtaining a question-answer pair matched with a target question from cached question-answer pairs; then, a judgment model with a small model parameter quantity is introduced to be used for judging whether the answers in the obtained question and answer pairs are matched with the target questions or not, if yes, the target answers of the target questions can be directly determined based on the answers in the question and answer pairs, and if not, the user questions are further analyzed through an RAG large model to obtain the target answers. According to the method, the cached question and answer pairs are used for retrieval, the judgment model with the small model parameter quantity is combined for judgment, and the RAG large model with the large model parameter quantity is further introduced for analysis only under the condition that the retrieved question and answer pairs do not meet the requirements, so that the retrieval efficiency can be effectively improved, efficient retrieval is realized, and the retrieval efficiency is improved. And the calling frequency of the RAG large model can be reduced as much as possible, and resource consumption is reduced.

This disclosure relates to the field of artificial intelligence technology, and particularly to a question-answer pair matching model training method, question-answer pair matching method, and apparatus. The method includes: acquiring question-answer pair training data, wherein the question-answer pair training data includes question training data and answer training data; inputting the question-answer pair training data into a question-answer pair matching model, predicting the matching degree between the question training data and the answer data, and outputting a predicted matching probability, wherein the predicted matching probability represents the similarity between the question training data and the answer data; calculating a target loss function based on the predicted matching probability; and adjusting the model parameters of the question-answer pair matching model based on the target loss function to obtain a target question-answer pair matching model. This method can improve the accuracy of the answer data output by the model for question data.

The application provides a method for instruction fusion using a superscalar processor and related equipment. A decoding unit obtains first instruction control signals of multiple instructions to be executed, and sends the first instruction control signals to a fusion decoding unit. The fusion decoding unit determines, for any one of at least one instruction pair, that a first instruction in the any one instruction pair is an I-type instruction, a second instruction in the any one instruction pair is an I-type instruction or an R-type instruction, and judges whether there is an instruction pair matching the any one instruction pair in a plurality of preset instruction pairs. If there is, instruction fusion is performed under the condition that a source operand of the first instruction is equal to a destination operand of the first instruction and the destination operand of the first instruction is equal to at least one source operand of the second instruction, to obtain a fusion instruction. A fusion instruction execution unit executes the fusion instruction according to an operation logic of the fusion instruction.

The present disclosure relates to systems, non-transitory computer-readable media, and methods that generate machine-learning training pairs across disparate modalities for a multi-modal machine learning model learning process. Indeed, in one or more implementations, the disclosed systems generate a first set of perturbation classification scores from a first data modality using a first classification model and a second set of perturbation classification scores from a second data modality using a second classification model. For instance, the disclosed systems compare the first set of perturbation classification scores with the second set of perturbation classification scores to determine a plurality of similarity measures. Moreover, in some instances, the disclosed systems identify pairs of data samples across the first data modality and the second data modality for a multi-modal machine learning model learning process using the pairs of data samples.

The invention relates to a signal identification method based on a signal fingerprint hash function, and belongs to the field of radio signal identification. The method comprises the following steps: segmenting continuous signals sampled by a receiver into frames with the same length, setting an overlapping ratio between adjacent frames, and capturing transient communication characteristics; adding frequency disturbance and burst pulse interference to the continuous signals; adopting a hybrid convolution kernel to extract different features of the spectrum data; mapping the high-dimensional feature vector into a low-dimensional binary hash sequence as a signal fingerprint; hash sequences of the same or similar patterns are identified through communication pair matching. According to the method, on the basis of the hybrid convolution kernels, through differentiated convolution kernel design, multiple convolution kernels are mixed in one convolution operation, the advantages of multi-scale convolution are utilized, different features of spectrum data are extracted, and finally, multi-modal features are converted into low-dimensional binary Hash sequences to serve as signal fingerprints, so that efficient signal recognition is achieved.

The invention relates to the technical field of table recognition, in particular to a multi-model collaboration-based table detection structured output method and device, equipment and a storage medium, and the method comprises the steps of obtaining a table picture containing a target table in response to a detection instruction for the target table; performing target detection on the image by using a pre-trained deep learning table detection model, and obtaining position coordinates of a table according to a detection result; performing target detection on the image by using a pre-trained deep learning character detection model, and obtaining position coordinates and character types of characters according to a detection result; and calling different character recognition models for character content recognition according to character categories, processing the table position coordinates, the character position coordinates and the character content, outputting table character content and table coordinates, and sending the output into a large language model to obtain a final text key value pair matching result. According to the invention, manual intervention and misrecognition are reduced, and the efficiency and accuracy of automatic processing are improved.