41 results about "Contextual similarity" patented technology

The embodiment of the invention discloses a context-based big language model dialogue processing method and device and electronic equipment, and can solve the problem of how to provide a dynamic management method based on context similarity analysis and intelligent screening to optimize big language model interaction. The method comprises the steps of obtaining current text information input by a user; coding the current text information and a plurality of pieces of pre-stored historical dialogue information to obtain a current text vector and a plurality of historical dialogue vectors; according to the semantic similarity between each historical dialogue vector and the current text vector, screening the plurality of historical dialogue vectors to obtain screened target historical dialogue information; splicing the current text information and the screened target historical dialogue information to obtain a target input text; the target input text is input into a target language model, target output response information is obtained, and the target language model comprises a large language model.

The invention belongs to but is not limited to the technical field of artificial intelligence, and particularly relates to a controllable content large model generation method and system based on multi-modal thinking chain reasoning, and the method comprises the steps: S1, improving the capability of multi-agent planning task dependence and execution sequence through employing a thinking chain method, enabling a complex task to be subdivided into a plurality of subtasks through the method, and enabling the subtasks to be subdivided into a plurality of subtasks; a resource transmission process between the sub-tasks is precisely planned and is executed in a chain mode; s2, complex tools are combined into sub-agents with different modal characteristics, so that the number of selectable sub-tasks during task planning is reduced, and the planning capability is improved; and S3, storing scene information of a previous dialogue by means of an external database, and when a new request is encountered, improving correlation and accuracy of multi-agent response by quoting historical interaction information and historical multi-modal resources which have context similarity with the current request, and realizing a more intelligent reaction mechanism.

Systems and techniques are provided for streamlining a proposal for a remote assistance system of an autonomous vehicle (AV) based on contextual similarities in remote assistance sessions that are associated with a set of AVs. An example process includes receiving a request for remote assistance from an AV in a scene and identifying a context associated with the AV. The context includes at least one of contextual factor(s) of the scene, AV characteristics, and task(s) associated with the request. The example process includes comparing the context associated with the AV with one or more contexts associated with one or more sessions associated with a set of AVs, determining a similarity between the context associated with the AV and the one or more contexts associated with the one or more sessions, and determining one or more proposals for the request based on the similarity.

The present disclosure enhances domain lookalike detection by integrating a phishing risk assessment score into a multi-layered evaluation framework. The method systematically generates lookalike domains through genetic algorithms. Registered domains undergo advanced phishing analysis, incorporating domain and URL checks, technical infrastructure assessments, content inspections, and reputation-based intelligence to calculate a dynamic phishing score. A comprehensive risk score is then determined by merging phishing likelihood with business attributes, graphical / contextual similarity metrics, and domain registration patterns. Domains are categorized into predefined risk levels including phishing, registered, preventative, company-owned, or watchlist, with specific action recommendations provided for each category. The system generates prioritized alerts for high-risk domains, offering customers actionable intelligence to mitigate threats. By combining phishing-specific indicators with contextual evaluations, this solution improves detection accuracy, reduces false positives, and enables organizations to respond effectively to domain-based threats in real time, addressing evolving cybersecurity challenges.

The invention belongs to the technical field of natural language processing, discloses a historical dialogue screening method and device for a dialogue system, and solves the problems that according to a historical dialogue screening scheme in the traditional technology, cross-time and cross-theme content may be omitted, dialogues without obvious keywords are difficult to process, the screening result deviates from user expectation, and the screening effect is poor. And the effect of screening dialogues lacking obvious intentions is poor. The method comprises the following steps: storing historical dialogue data between a user and a dialogue system based on the time and theme of a dialogue; when historical dialogue screening is carried out, firstly, the similarity between a current question and stored historical dialogue data is calculated, and weighted fusion of semantic similarity, keyword similarity, theme similarity and context similarity is included; and then based on a preset dialogue screening threshold, combining a similarity calculation result of the current question and the stored historical dialogue data, and screening out historical dialogue data most related to the current question from the stored historical dialogue data. The method is suitable for the dialogue system.

Techniques for correcting data drift of a language model are disclosed. A model is built, and this model is designed to solve a same task for which the language model has been trained. The model is applied to new input data. This application results in generation of a prediction comprising predicted label data. Context is stored in a context management structure (CMS). The context includes a prompt template, a prediction, and labeled input data used to train the language model. The data drift is determined to have occurred. This determination is performed by determining that the context is within a threshold level of similarity to a previously stored context. In response to determining that the data drift has occurred, an operation is performed to correct the data drift.

The invention discloses a method for analyzing software code similarity and an analysis system, and relates to the field of software engineering and machine learning. The method comprises the steps of establishing a code snippet database, extracting multi-dimensional features of code snippets, executing similarity analysis of the code snippets, and establishing a similarity analysis result; performing context backtracking analysis on the code snippets in the database by using an adaptive analysis network, and establishing a context similarity result; performing static and dynamic fusion analysis on the database, establishing a joint function similarity, and establishing a function similarity result; code version change analysis is executed, and a version similarity result is established; and performing code similarity retrieval management according to the four analysis results. The technical problem that the code similarity analysis efficiency and precision are low due to the fact that the code snippets are high in complexity in multi-dimensional feature analysis is solved, and the code similarity analysis accuracy and efficiency are improved in the mode that multi-dimensional analysis and self-adaptive backtracking analysis are combined.

The invention provides an alarm suppression method and system based on semantic coupling distance discrimination, and aims to introduce context similarity calculation aiming at the problem that existing traditional word similarity calculation often only considers semantic items of words and neglects meanings of the words in specific contexts, so that the context similarity calculation can better understand the specific meanings of the words in different contexts and improve the accuracy of alarm suppression. And the calculation accuracy is enhanced. Compared with a traditional method, the method has the advantages that by combining word semantic similarity and context similarity calculation and optimizing a similarity algorithm of coupling co-occurrence distance distinction degree, deep semantic association between alarm texts can be recognized more accurately, the accuracy of alarm clustering is remarkably improved, redundant alarm output is effectively suppressed, and the method is suitable for popularization and application. And operation and maintenance personnel are helped to quickly position and process batch alarm problems, so that the alarm processing efficiency and reliability of the system are improved.

The invention discloses a psychological health support dialogue generation method based on multi-modal input. The method comprises the steps that 1, text data are acquired and preprocessed; 2, calculating context similarity to obtain a high-quality example set; 3, extracting key visual clues from the image by utilizing a visual language model, wherein the key visual clues comprise facial expressions, actions and emotions; 4, integrating an external common knowledge base to understand the emotion and intention of the user; 5, generating a consultation strategy by using a large language model; and 6, dialogue generation in combination with the context. According to the method, the visual information, the dialogue context and the consultation strategy are comprehensively utilized, the performance of a large language model in the psychological health dialogue is enhanced, more detailed and similar supportive replies can be generated, the emotion recognition and response quality are remarkably improved, the emotion state of a user can be comprehensively understood and responded, and the user experience is improved. Therefore, the mental health support effect is effectively improved.

Artificial intelligence models are trained with training datasets of known input / output values. Test datasets are used to evaluate the trained artificial intelligence models. Context mismatch between the training dataset and the test dataset can slow down the development of artificial intelligence models. The described systems and methods can identify context similar datasets for the purpose of training and testing an artificial intelligence model. In one embodiment, a context similarity detector can ingest and combine a training dataset and a test dataset and generate a context similarity score for the two. If the score is above a threshold, the datasets are similar, and the relevant artificial intelligence model can be trained with one and tested with the other.

A method of finding online relevant conversing posts, comprises receiving, by a web server serving an online forum, a query post from an inquirer using the online forum, computing a contextual similarity score between each conversing post of a set of conversing posts with a query post, wherein the contextual similarity score is computed between the body of each of conversing posts and of the query post, wherein N1 conversing posts with a highest contextual similarity score are selected; computing a fine grained similarity score between the subject of the query post and of each of the N1 conversing posts, wherein N2 conversing posts with a highest fine grained similarity score are selected; and boosting the fine grained similarity score of the N2 conversing posts based on relevance metrics, wherein N3 highest ranked conversing posts are selected as a list of conversing posts most relevant to the query post.

The invention relates to the technical field of internet of things monitoring and data unification of logistics assets, in particular to a multi-system data unification logistics asset internet of things monitoring and processing system, which comprises an atlas construction unit used for collecting multi-source heterogeneous data streams in real time and aggregating to generate dynamic time sequence diagram snapshots; the heuristic quantization unit is used for taking the score as a pseudo tag; the time sequence diagram neural network unit is used for generating high-dimensional state vectors for the independent nodes; the model training unit is used for enabling the similarity of the high-dimensional state vector in a vector space to reproduce heuristic context similarity; the unified asset aggregation unit is used for resolving the high-dimensional state vectors and aggregating the high-dimensional state vectors into a unified asset view; the hidden event inference unit is used for generating a hidden event alarm and judging the state as a normal state when the abnormal inference score does not exceed an alarm threshold value; the technical problem that the business system identifier and the Internet of Things identifier can be aligned only through manual labeling is solved, and automatic association of heterogeneous data and construction of a unified asset view are achieved.

Described herein are methods, systems, and. computer-readable media, for classification. Techniques may retrieve content datasets, gather first sets of input data from the content datasets, and preprocess the first sets of input data. Techniques may next generate second sets of input data by embedding associated first metadata and second metadata. detennine a plurality of contextual similarities based on contextual information, and generate third sets of input data by grouping one or more sets of input data based on the determined plurality of contextual similarities. Techniques may further detennine, for each content dataset of the plurality of content datasets using one or more machine learning models, one or more second categories associated with the content dataset.

The present application relates to a kind of incomplete speech rewriting model construction method and device, method includes: based on span dependency and insertion dependency dependency modeling and using node link resolution mode obtains the dependency graph of incomplete speech rewriting text editing operation;Using GPT model, the context similarity feature and / or rewriting consistency characteristic of current incomplete speech sentence are calculated, the context similarity feature and / or rewriting consistency characteristic are used to enhance the interactive inference of incomplete speech rewriting;The dependency graph score feature is fused with the context similarity feature and / or rewriting consistency characteristic, and the final feature after the feature fusion is pushed to rewrite result based on. More abundant semantic features can be provided for parsing model, and the speech rewriting effect is improved.

The invention relates to the technical field of software analysis, and provides a code similarity detection method and device. The method comprises the following steps: determining context information of a query function according to a binary dependency graph of the query function; determining context information of each candidate function according to the binary dependency graph of each candidate function; determining context similarity between the query function and each candidate function according to the context information of the query function and the context information of each candidate function; according to the content of the query function and the content of each candidate function, determining the content similarity between the query function and each candidate function; selecting an initial similar function from the candidate functions according to the context similarity and the content similarity; and inputting the initial similarity function and the query function into the large language model to obtain a code similarity detection result output by the large language model. According to the code similarity detection method and device provided by the invention, the code similarity detection efficiency and accuracy are improved.

Described herein are methods, systems, and computer-readable media for the generation of classifications of content. Techniques may extract and clean information associated with a first instance of content associated with a first person. Techniques may next classify the cleaned information into a first set of categories and determine a second set of categories based on the cleaned information associated with the first and other instances of content and aggregate the cleaned information using the second set of categories into groups. Techniques further determine a third set of categories of information associated with a group of people including the first person to generate metadata for the information associated with the group of people. Techniques to generate metadata include using frequency data associated with the information based on the first set of categories, the second set of categories, and the third set of categories.

The invention provides a Text2SQL semantic caching method based on context and pattern matching, and relates to the field of natural language processing and database interaction.The method comprises the steps that user query is input into a semantic compression module, semantic significant keywords are extracted to obtain a candidate set, a vocabulary set is extracted and constructed through a large language model, and a database is obtained; obtaining a compressed query through a syntactic organizer; inputting the historical dialogue into a context encoder, and obtaining global context representation through encoding and two-stage attention; performing coarse-grained filtering on the compressed query to form a newest candidate set; performing fine-grained context matching on the global context representation and the newest candidate set to obtain context similarity; and comparing the context similarity with a preset threshold value, judging whether the cache is hit or not, if the cache is hit, interacting the hit SQL statement with a database, and returning a query result. According to the method and the device, the problems of low SQL multiplexing accuracy and high response delay are solved.

The application relates to the technical field of image processing, and provides an image processing method, a computer program product, a storage medium and an electronic device. The image processing method comprises the following steps: obtaining the self feature of a target in a to-be-processed image and the self feature of the target in a reference image; fusing the self feature of the to-be-processed target with the self feature of at least one other target in the to-be-processed image and / or the reference image to obtain the context feature of the to-be-processed target, and fusing the self feature of the reference target with the self feature of at least one other target in the to-be-processed image and / or the reference image to obtain the context feature of the reference target; calculating the context similarity of the to-be-processed target and the reference target according to the context feature of the to-be-processed target and the context feature of the reference target; and determining the final similarity of the to-be-processed target and the reference target according to the context similarity. Since the context information of the target is fully considered when the similarity is calculated, the finally calculated similarity has high precision.

The application discloses a mental health support dialogue generation method based on multi-modal input, including: 1. Obtain text data and preprocess; 2. Calculate the context similarity to obtain a high-quality example set; 3. Use a visual language model to extract key visual clues from images, including facial expressions, actions and emotions; 4. Integrate an external common sense knowledge base to understand the user's emotions and intentions; 5. Generate counseling strategies by using a large language model; 6. Context-based dialogue generation. The application enhances the performance of large language models in mental health dialogue by comprehensively utilizing visual information, dialogue context and counseling strategies, enabling the generation of more detailed and empathetic supportive replies, thereby significantly improving emotional recognition and response quality, and more comprehensively understanding and responding to the emotional state of users to effectively improve the effect of mental health support.

Systems and techniques are provided for streamlining a proposal for a remote assistance system of an autonomous vehicle (AV) based on contextual similarities in remote assistance sessions that are associated with a set of AVs. An example process includes receiving a request for remote assistance from an AV in a scene and identifying a context associated with the AV. The context includes at least one of contextual factor(s) of the scene, AV characteristics, and task(s) associated with the request. The example process includes comparing the context associated with the AV with one or more contexts associated with one or more sessions associated with a set of AVs, determining a similarity between the context associated with the AV and the one or more contexts associated with the one or more sessions, and determining one or more proposals for the request based on the similarity.

The invention discloses a non-reference problem quality evaluation method and system based on context similarity calculation, and the method comprises the steps: carrying out the feature extraction in each layer through a pre-training language model by using a trainable adaptive weight according to an input problem and context, connecting the weighted feature representations, and carrying out the feature extraction; generating a three-dimensional semantic representation of the question and the context; obtaining a correlation matrix between the context and the problem, generating a mask matrix by setting a threshold value, sparsifying the correlation matrix by using the mask matrix, and determining the content related to the problem in the context; the question features and content features related to the question in the context are fused from different levels; and calculating a similarity score between the fused problem features and content features related to the problem in the context to obtain an evaluation result. According to the method, the content related to the problem is determined from the context, the correlation degree of the generated problem and the context is evaluated, and the evaluation limitation caused by the fact that an existing method depends on manual annotation is solved.

A real-time fraud detection system is disclosed that models financial transactions as evolving memory graphs using biologically inspired dendritic architectures. The system employs graph neural networks with edge decay and reinforcement mechanisms to retain behavioral context over time, enabling dynamic anomaly detection. An attention-based recall module reactivates latent fraud patterns based on contextual similarity. Optionally, quantum walk engines and entangled memory states are integrated to accelerate pattern discovery in high-dimensional data. The invention supports few-shot meta-learning, real-time inference, and deployment on neuromorphic or quantum-classical platforms, offering adaptive, scalable, and interpretable fraud detection across dynamic financial ecosystems.

The present invention makes it possible to calculate an index with high accuracy, even for an area with limited data. This quality estimation system includes: a context data generation unit that generates a plurality of pieces of context data from measurement data; a first index calculation unit that generates first index data for each context and for each area, using the context data; an identification unit that identifies a second context similar to a first context; a second index calculation unit that defines an area for which the first index data is not generated, among a plurality of areas, as an unmeasured area, and generates second index data for the first context by using the context data of the first context for an area other than the unmeasured area, and using the context data of the second context for the unmeasured area; and an output unit that outputs the second index data.

The embodiment of the invention provides a geographic entity information processing method and device, a storage medium and a program product. In the embodiment of the invention, for at least two pieces of geographic entity description information, a geographic semantic feature encoder, a distance feature encoder and a geographic context feature encoder are respectively utilized to encode names and address semantics, spatial positions and surrounding geographic environments; obtaining a geographic semantic similarity, a spatial distance feature and a geographic context similarity; and inputting the geographic semantic similarity, the spatial distance feature and the geographic context similarity into a classification head network, and outputting a classification result of whether the at least two pieces of geographic entity description information point to the same geographic entity or not. Therefore, by introducing the model to learn the incidence relation between different features, full utilization of multi-dimensional features such as geographic semantics, spatial positions and surrounding geographic environments is realized, feature fusion and accurate judgment are realized in the classification head network, and the accuracy and robustness of geographic entity matching are improved.

Systems and methods are provided for determining calls and contacts associated with calls that are contextually similar based on content of the calls. Data associated with a call include one or more utterances made by speakers and a set of values indicating relevance between content of the call and topic categories of the call. The disclosed technology generates a topic vector associated with a call and / or respective speakers of the call. The topic vector includes a multi-dimensional vector where each dimension corresponds to a topic category. The disclosed technology determines calls that are contextually similar by comparing angular distances between topic vectors. A search query receiver receives a search query that queries contacts and calls that are contextually similar to a given call and / or a speaker. The disclosed technology identifies calls with topic vectors that are within a predetermined angular distance.

The application provides a non-structured content similarity determination method and device and electronic equipment, which normalizes extracted content obtained from non-structured content to obtain normalized content; sets a label for the normalized content based on a specified feature extracted from the normalized content, performs semantic similarity matching on the normalized content with the set label to obtain a first matching result; performs relationship calculation of metric fusion on the normalized content with the set label based on context correlation of the normalized content to obtain relationship metric information; performs context similarity matching according to the relationship metric information to obtain a second matching result; and performs content similarity matching on the non-structured content according to the normalized content, the label set for the normalized content and the relationship metric information to obtain a third matching result. This method comprehensively utilizes multi-dimensional methods of content analysis to perform similarity matching on non-structured content, so that each dimension method can complement each other, and the accuracy of analysis is improved.

This application provides a method for protecting prompt words in large models based on differential privacy. By constructing a perturbation mapping table, differential privacy perturbation is applied only to the first token encountered in a multi-turn dialogue, and the mapping relationship is stored. Subsequent repeated tokens directly reuse the historical perturbation results, thus avoiding the accumulation of privacy budget with the increase of dialogue turns. Simultaneously, an attention-based context-aware utility function is combined to integrate the static embedding similarity of tokens with dynamic context similarity to maintain semantic consistency across turns. Finally, a two-stage bucketing index mechanism is adopted, first bucketing candidate tokens according to utility, and then fine-sampling within the selected buckets, effectively mitigating the long-tail effect under large vocabulary and improving the stability of perturbation selection. This application achieves a balance between privacy protection strength and semantic usability in multi-turn dialogues, and only requires perturbing the input prompt words without accessing the model's internal architecture, making it suitable for black-box large language model inference scenarios.

The invention discloses a causal class activation mapping-based region perception open set identification method, which comprises the following steps of: performing closed set training through a convolutional neural network by taking a training image of a known class as input, and generating class activation mapping reflecting causal contribution of a local region to a classification result based on region ablation analysis; the method comprises the following steps: in a training stage, extracting a region perception feature vector of a training image by utilizing a saliency map, and constructing a region prototype corresponding to a known category; in a test stage, region perception feature vectors are extracted from a test image, a region consistency score is calculated with the region prototype constructed in the training stage, and fusion judgment is carried out in combination with a global uncertainty score based on an energy model, so that known category identification and unknown category rejection are realized. According to the method, causal class activation mapping is introduced into an open set identification judgment process, so that the capability of distinguishing context similar unknown classes can be enhanced, and a regional interpretable basis is provided.

The present subject matter discloses techniques to identify linkable items and accordingly generating recommendations indicating such linkages. One or more work data objects, each linked with a corresponding work item, may be obtained. Further, one or more support data objects, each linked with a corresponding support item, may be obtained. The one or more work data objects and the one or more support data items may be encoded into work vectors and support vectors, respectively. A similarity score may then be estimated based on a quantitative comparative assessment of the work and support vectors. The similarity score may indicate a contextual similarity between each of the one or more work data objects and the one or more support data objects. Based on the similarity score, the one or more support items may be linked with the one or more work items.