1179results about "Natural language translation" patented technology

Certain aspects of the disclosure provide a method for generating context-augmented responses. In aspects, the method includes receiving a job application; extracting, from the job application, a question; generating, using an embedding model, a vector representation of the question; identifying, within a vector database, a most similar vector to the vector representation of the question; extracting, from a question-response database, an encoded context group; retrieving, from a job applicant database, professional context of the job applicant; generating, based on the context group and the professional context, a prompt chain for causing a language model to generate a response; processing, using the language model, the prompt chain to generate the response; calculating, using a risk scoring model, a risk score for the generated response; and in response to the risk score being below a predetermined threshold, populating a portion of the job application with the response.

Techniques for knowledge discovery based on indirect inference of association are presented. A data management component (DMC) can determine and extract, in a structured format, entities, relationships between entities, and concepts relating thereto in documents, based on analysis of information in the documents and / or keywords relating to concepts, to generate an association inference model. Using artificial intelligence techniques, DMC can embed the entities and relationships to a common representation to generate and train a scoring model that can be used to evaluate and score similarity strength between entities, including entities that do not have a known relationship, and can predict or infer relationships, including indirect relationships, between entities or between concepts. In that regard, DMC or user can evaluate concept-level scores to determine a level of relationship between concepts. DMC can feedback information from the scoring model or evaluation to update the association inference model.

A text sequence recommendation method and system based on a large language model is disclosed, belonging to the technical field of recommendation algorithms. The method includes: a data preprocessing stage, a large language model pre-training stage, a sequence model fine-tuning stage and a matching stage. According to this disclosure, a large language model is introduced into a text sequence recommendation task, so that text can be better modeled by utilizing rich pre-training corpus of the large language model; meanwhile, sequence modeling is performed on the text, the capability of sequence recommendations modeling in a large model is activated, an ID-based recommendation paradigm in a traditional recommendation algorithm is eliminated, and recommendation task learning processing is better performed in a cold start scenario and a knowledge transfer scenario; and finally, a recommendation result is finally optimized by a sequence model.

Technology is disclosed for programmatically generating procedures based on historical conversation records and providing the procedures to users to resolve similar issues. In one implementation, an input instruction is generated to cause a language model to determine resolved historical conversation records from historical conversation records. A second input instruction is generated to cause a language model to generate procedure documents from the resolved historical conversation records. The second input instruction includes the resolved historical conversation records and a procedure generation prompt instructing the second language model to (1) extract one or more issues from each resolved historical conversation record, (2) generate a corresponding issue description of each issue, (3) generate a corresponding step-by-step procedure resolving each issue, and (4) generate a corresponding issue resolution description for each issue. The procedure documents are stored in a knowledge base to provide the step-by-step procedures to users to resolve similar issues.

A first large language model (LLM) instance may be instructed to request data while being prevented from performing calculations using the data. A second LLM instance may be instructed to provide a response to the request for data based on a known complete data set. The response may be translated into a machine-readable response in a format configured for processing by a calculation engine. The calculation engine may process the machine-readable response, thereby generating a calculation engine output. A mismatch between the calculation engine output and a known result obtained using the known complete data set may be identified, and the instruction to the first LLM may be modified in response.

A method and system for automated support sub-topic classification using large language models (LLMs). The method includes collecting user transcripts and LLM-generated summaries, performing unsupervised learning to identify common themes, generating sub-topics, and creating labeled datasets. A supervised learning model is trained to categorize user transcripts into the identified sub-topics. The system performs categorization of new user queries and generates appropriate action responses.

Techniques for prompt template optimization with language models are described. In some examples, a prompt template optimization request to optimize a generative artificial intelligence model prompt template is received, the prompt template optimization request including an initial prompt template and an indication of a selected function, the selected function to implement at least a portion of a prompt template optimization workflow. The prompt template optimization workflow is processed with the selected function, the prompt template optimization workflow including one or more iterations of generating, evaluating, and selecting prompt template variants based at least in part on the initial prompt template to yield a final prompt template. The final prompt template is output.

Techniques are disclosed for a conversational assistant system that employs a machine learning system to act as an intermediary between natural language questions of a user and one or more Application Programming Interfaces (APIs) implemented by a network analysis system. In one example, the conversational assistant system receives API definition files including sets of function definitions and corresponding function descriptions and a natural language question from a user device. The conversational assistant system processes the API definition files and the natural language question to obtain function definitions invoking API endpoints of the APIs. The conversational assistant system sequentially invokes, based on the function definitions, the API endpoints to obtain responses from the APIs implemented by the network analysis system. The conversational assistant system generates, based at least in part on the responses from the APIs, a natural language answer for output at an output device.

To improve convenience of management of a facility that is managed based on data from a plurality of sensors or cameras.SOLUTION: In an information processing system includes a server 1 that executes predetermined processing using input data, and one or more input cells 2 that provide at least part of the input data to the server 1, the input cells 2 acquire information on the positions of the input cells. The server 1 manages input cell position information and a model in a three-dimensional map of a facility in which the input cells 2 are arranged. The input cells 2 execute re-learning based on information on the three-dimensional map and information on the positions of the input cells 2 managed by the server 1.SELECTED DRAWING: Figure 1

Embodiments described herein provide a method of generating a vision-language task output to a text instruction relating to an input image, the method comprising receiving, via a data interface, the input image and the text instruction comprising an instruction relating to the image. The method further includes encoding, via an image encoder, the image into a first image representation. The method further includes generating, by a multimodal encoder, a second image representation based on cross-attending the first image representation to the text instruction. The method further includes generating, by a neural network based language model, a vision-language task output in response to the text instruction based on an input combining the second image representation and the text instruction.

This application discloses a method and system for multilingual enhancement based on a large language model using model ensemble. The system includes: a pre-trained multilingual translation model, a semantic representation mapping module, and a large language model. The multilingual translation model is used for multilingual semantic modeling and language generation, including a multilingual encoder module and a multilingual decoder module. The semantic representation mapping module is used to transform the latent space representations of different models into an interactive unified semantic space based on a cross-model representation mapping mechanism. The output of the multilingual encoder is mapped to the unified semantic representation space of the large language model, and the mapped semantics are input into the large language model to perform language-independent instruction understanding. The intermediate semantic representation output by the large language model is mapped and transformed to a cross-attention representation space, generating the final output text under the target language distribution. The system of this application outperforms existing technologies in terms of efficiency, stability, and generation quality in multilingual capability extension.

The system according to this embodiment aims to support learning exchanges that promote intercultural understanding and broaden international perspectives. [Solution] The system according to the embodiment comprises a reception unit, a selection unit, a pairing unit, a support unit, and a translation unit. The reception unit receives input from the user regarding their interests and learning objectives. The selection unit analyzes the information received by the reception unit and selects the most suitable learning partner. The pairing unit pairs the user with the learning partner selected by the selection unit. The support unit assists the learning interaction with the learning partner paired by the pairing unit. The translation unit provides real-time translation in the learning interaction supported by the support unit.

A data processing system implements receiving, from a client device, first textual content inserted into a user interface element of a first user interface on the client device, the first textual content comprising a natural language description of a first project for which a first workspace is to be created, analyzing the first textual content to obtain keywords in the first textual content using an NLP model trained to receive the textual content and to output the keywords, conducting a search for candidate collateral items associated with each of the keywords using a first search engine, causing the client device to present the candidate collateral items on the first user interface, receiving, from the client device, a first user input selecting one or more of the collateral items from among the candidate collateral items, and causing the client device to present a second workspace user interface representing the first workspace.

Provide a system. 【Solution means】 Means for receiving unstructured data including unclear points, Means for analyzing the received unstructured data and extracting relevant information, Means for searching past materials based on the extracted relevant information, Means for generating a plurality of information sets based on the search results, Means for assigning reference information to each generated information set, Means for presenting a plurality of information sets to the user, Means for displaying an information set on a device for controlling an industrial automation machine, A system including the above.

The present disclosure provides a problem positioning method and device, equipment and storage medium, relates to the technical field of computers, in particular to problem positioning technology. The specific implementation scheme is: obtaining a natural language input, the natural language input is used to describe an abnormal problem of application runtime; the natural language input is analyzed to obtain an analysis result, the analysis result contains problem semantic information corresponding to the abnormal problem; based on the analysis result, an operation and maintenance knowledge segment corresponding to the problem semantic information is retrieved from a pre-constructed knowledge base; based on the analysis result and the retrieved operation and maintenance knowledge segment, target log data is obtained; based on the retrieved operation and maintenance knowledge segment and the target log data, a problem positioning result is generated. The method realizes the automatic processing of the natural language description problem, and improves the efficiency and accuracy of operation and maintenance problem positioning through the collaborative analysis of the knowledge base and the log data.

We provide the system. [Solution] A means for acquiring location information to determine the current location of a user within the facility, A means for obtaining a floor map of a facility from the internet based on the aforementioned location information, A means for analyzing the acquired floor map and generating a route to the destination, A means of providing guidance to the user based on the generated route information, A means of determining the user's current location from their camera image and updating route guidance, A system that includes this.

The present application is applicable to the technical field of artificial intelligence. Provided are a method for training a driving risk recognition model, and a driving risk recognition method and apparatus. The method comprises: acquiring data to be processed of a plurality of drivers; determining corresponding events on the basis of the data to be processed; textualizing identifiers of the drivers and the events corresponding to the drivers to obtain textualization results; and on the basis of the textualization results, training a driving risk recognition model to be trained to obtain a trained driving risk recognition model. The driving risk recognition model to be trained is a large language model, and the trained driving risk recognition model is used for outputting a driving risk. The model trained by means of the above method is more accurate.

Herein disclosed is a method performed by at least one processor. The method includes receiving request to create generated content, creating the generated content as a function of the requested theme using generative artificial intelligence, determining if the generated content is approved based on a content approval machine learning model configured to generate a content approval output based on the generated content and at least one preference parameter defining approval criteria for the generated content, and providing access to the generated content in response to the content approval output indicating approval. The request includes a requested theme directed to a target characteristic of the generated content. The content approval machine learning model includes a neural network. The content approval output may include a content approval score, and the content approval output may indicate approval when the content approval score satisfies an approval threshold.

The system according to this embodiment aims to enable children to access accurate and in-depth knowledge. [Solution] The system according to the embodiment comprises an acquisition unit, an analysis unit, a translation unit, a dialogue unit, and a suggestion unit. The acquisition unit acquires highly reliable information. The analysis unit analyzes the information acquired by the acquisition unit. The translation unit translates and summarizes the information analyzed by the analysis unit for children. The dialogue unit provides an interactive learning experience. The suggestion unit analyzes the learning history and suggests new topics.

Systems and methods for configuring a smart contract are disclosed. A system may receive unstructured data and identifying, based on the unstructured data, an intent to perform a transfer of data. The system may then identify, based on the unstructured data, parameters associated with the transfer of data. The parameters may include an amount of data to be transferred, a transferee, and at least one condition associated with the transfer of data. The system may then retrieve identifier data from a storage module, the identifier data including a first unique identifier. The system may then send, to a Large Language Model (LLM) via a first prompt engine module and an LLM Application Programming Interface (API), a prompt based on the unstructured data and the identifier data. The system may then receive LLM output from the LLM, and based on the LLM output, configure a smart contract.

We provide the system. [Solution] A generation mechanism means for automatically generating documents based on transaction conditions received from an input device, An analytical mechanism means for detecting hazards and defects contained in the aforementioned document, A display mechanism means for presenting detected hazards and defects to the user, A dialogue mechanism for receiving questions regarding legal issues and generating appropriate responses, An interface means for displaying transaction-related information on a terminal device, A system that includes this.

The invention relates to translating memory references. Apparatus, systems, and techniques translate program instructions that use reference to access objects stored in memory to program instructions that use register identifiers to access objects stored in registers. In at least one embodiment, a compiler translates one or more references to one or more corresponding register identifiers.

A computing system executing a game maker model receives a user query including a description of the 3D object, generates an input prompt based on the user query, encodes the input prompt into embeddings, inputs the embeddings into a control network to generate latent features, inputs the latent features and the embeddings into a diffusion model to generate a 2D image, and generates and outputs the 3D object based on the 2D image.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for table detection using text streams. One of the methods includes detecting, in a text stream and using column identification data, text for one or more cells in a table; creating, using the text for at least some of the one or more cells in the table, a data structure for the cell a) that associates two or more values from the table and b) for use by a downstream system as part of a natural language analysis process of data from the text stream; and storing, in memory, the data structure.

The application discloses an input method, an input device, an electronic device, a medium and a program product, and belongs to the technical field of electronic products. The input method comprises the following steps: receiving a character input and a voice input of a user; and displaying first input information based on character information corresponding to the character input and voice information corresponding to the voice input.