171 results about "Word processing" patented technology

Techniques and systems are disclosed that perform content generation with integrated automated formatting using word processing applications that deploy large language models (LLMs). The techniques include receiving, a natural language (NL) query for a synthetic content for a document, identifying formatting rules of the document, and generating an augmented query that includes a representation of at least a portion of the NL query and a representation of the one or more formatting rules of the document. The techniques further include providing the augmented query to an LLM and updating the document with the synthetic content generated by the LLM in response to the augmented query.

Presented herein are interactive graphical user interfaces (GUIs) for providing automated, real-time interactive assistance to a human user creating electronic documents. In certain embodiments, the GUI includes an alphanumeric text processing region, an AI-generated text region, and a reference list region. The AI-generated text region contains one or more alternative blocks of text responsive to a user query and based on specific references (e.g., articles, books, or other written works). The references are listed in the reference list region of the GUI.

The invention discloses a large model-based government affair knowledge retrieval method, apparatus and device, and a medium, and relates to the technical field of natural language processing, and the method comprises the steps of performing information extraction on a to-be-processed government affair work order through a preset government affair large model to obtain target structured information; based on a word processing fusion algorithm constructed by a word frequency-inverse document frequency algorithm and a graph sorting algorithm, performing word processing on the target structured information to obtain work order keywords; converting the work order keyword into a target semantic vector; performing retrieval operation on a preset government affair knowledge base by utilizing a preset retrieval condition and the target semantic vector to obtain a target government affair knowledge document; the preset retrieval condition is a retrieval condition constructed based on time, space and the work order emergency degree. Therefore, the information accuracy and the processing efficiency can be improved through word processing; and in combination with retrieval conditions constructed based on time, space and work order emergency degrees, accurate retrieval can be provided, and the processing effect on various complex government affair work orders is improved.

The invention provides a cue word processing method, a cue word processing device, electronic equipment and a computer readable storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring corpus information, and generating a plurality of first cue words according to the corpus information; determining a plurality of second cue words based on the plurality of first cue words; processing the second cue words by using a pre-trained cue word classification model to obtain classification results corresponding to the second cue words; and according to the classification result of each second cue word, carrying out set division on the plurality of second cue words to generate a cue word set. According to the method and the device, the cue word can be accurately and effectively processed to obtain the cue word set.

The invention provides a CAD file summary information automatic extraction method and device, a medium and a product, and relates to the technical field of computer picture word process.The method comprises the steps that a target CAD file is analyzed through a CAD file analysis development tool, and first word information and geometric attribute information corresponding to the first word information are obtained; retrieving the first character information according to the retrieval keyword to obtain second character information; according to the corresponding geometric attribute information in the second text information, carrying out positioning adjustment display through a CAD file analysis development tool to obtain a picture file; calling a first large language model to perform character recognition and arrangement processing on the picture to form a text list; and calling the second large language model to obtain summary information of the target CAD file according to a preset prompt template and the text list. According to the scheme, when a large batch of CAD files are processed, the processing efficiency is remarkably improved.

The embodiment of the invention provides a cue word processing method and device.The cue word processing method comprises the steps that in the cue word generating process, labeled data and data labels are input into a large language model through an intelligent agent for labeling rule recognition, and labeling rules are obtained; and inputting the labeled data and the labeling cue word containing the labeling rule into a large language model for data labeling processing to obtain a data labeling result, comparing the data labeling result with a data label, and updating the labeling rule according to the comparison result to obtain a target labeling rule after reinforcement learning is completed, and generating a target annotation prompt word based on the target annotation rule.

Wake-word processing by a wearable electronic device could be carried out when the device is worn by a user and is in a device sleep state, the device including a linear microphone array having at least two microphones vertically spaced from each other, and the device also including a processor. And the example method could involve (i) the at least two microphones of the linear microphone array receiving an audio waveform representing a wake-word utterance, (ii) the processor making a determination, based at least on an angle of arrival of the audio waveform at the at least two microphones of the linear microphone array and / or an energy level of the audio waveform received at the at least two microphones of the linear array, of whether to accept the wake-word utterance or rather to reject the wake-word utterance, and (iii) the processor controlling operation of the device based on the determination.

The invention provides a system and method for digitally and intelligently recognizing a contract, and belongs to the technical field of character recognition, and the system comprises a data acquisition module which is used for obtaining an initial image of a target contract through image acquisition equipment, and preprocessing the initial image to obtain a standard image in an electronic copy format; the intelligent recognition module is used for recognizing the image text content of the standard image by using an OCR tool and converting the image text content into original text data; the text processing module is used for performing word processing on the original text data to obtain standard text data of the target contract; and the text analysis module is used for analyzing the standard text data by using a natural language recognition technology, extracting key contract data and integrating the key contract data into formatted contract data with a standard structure. Contract recognition efficiency and accuracy are improved, and a system recognition process is optimized.

The embodiment of the present invention provides a method and system for constructing a knowledge base in the field of power grid based on a vector database, which belongs to the field of emerging software and new information technology services. The method includes: obtaining a power grid data file to be stored; obtaining keywords in the power grid data file; using a preset vector database to process the keywords to obtain word vectors; determining similar word vectors of each word vector according to a preset similarity distance; word-processing the power grid data file; sequentially reading the power grid data file after word processing; processing the read word vectors or similar word vectors to obtain mapping values; counting according to the mapping values; using the counting result as the second feature of the power grid data file; clustering the power grid data file using the second feature and the keywords; index encoding the result of the clustering operation to obtain the power grid domain knowledge base.

The invention relates to the technical field of intelligent interface generation, in particular to an interactive interface generation method and system, computer equipment and a medium. The method comprises the following steps: receiving a multi-protocol interaction request, and extracting natural language context information carried in the multi-protocol interaction request through a protocol processing layer; according to the natural language context information, executing cue word processing to generate interface generation trigger judgment; when the interface generation trigger judgment indicates that an interface needs to be generated, generating an interface mode definition instance based on the natural language context information and a preset cue word template; and according to the interface mode definition instance, generating an interactive interface through a rendering output device. In this way, the technical problem that in an existing interface generation technology based on natural languages, it is difficult to collaboratively optimize the interface dynamic adaptation capacity and the system expansibility is solved, and the intelligent level, cross-platform compatibility and service response agility of interface generation are improved.

The present invention relates to the field of big data technology, and discloses an information recommendation method based on user portrait, comprising the following steps: receiving a human-computer dialogue chat record in a first scenario in real time through a log collection system Flume; performing desensitization processing on the chat record to obtain first data; performing stop word removal processing on the first data to obtain second data; extracting keywords of the second data through a preset word graph; obtaining a first tag set according to the keywords; removing duplicates from the first tag set data to obtain a second tag set; generating a user interest portrait based on the second tag set, and storing the user interest portrait in a database; receiving a recommendation instruction, and obtaining a user interest portrait according to the recommendation instruction; obtaining information to be recommended corresponding to the recommendation instruction according to the user interest portrait. The present invention also provides an information recommendation device, equipment and storage medium based on user portrait, and the information recommendation method based on user portrait provided by the present invention improves the accuracy of user portrait.

The invention provides a technology trend prediction method and device, a computer system and a medium, and aims to excavate key technology evolution trends in various technical fields and identify future hot topics. The method comprises the following steps: performing document matching on generalized topics of various technical fields based on a highly cited document database, and extracting documents related to the generalized topics; performing text preprocessing on the literature under the generalized theme, wherein the text preprocessing comprises field proper noun standardization, word segmentation, stop word processing and word form merging and text vectorization; adopting an LDA topic clustering model to perform detailed topic clustering on the pre-processed literature, generating a keyword set and keyword contribution degrees, and condensing the keyword set and the keyword contribution degrees into detailed topics; and training and predicting the popularity data of the detailed theme over the years based on a Chronos time sequence model, and identifying the detailed theme with high popularity in the future.

The invention relates to the technical field of natural language processing, can be applied to the fields of financial public opinion monitoring, compliance review, risk prevention and control and the like, and provides a multi-modal sensitive word processing method and device, a storage medium and computer equipment. Comprising the steps of obtaining to-be-published information and a pre-constructed knowledge base, wherein the information type of the to-be-published information comprises at least one of an image, a text, an audio and a video; performing sensitive word recognition on the to-be-published information based on a large language model and a sensitive word knowledge base to obtain a recognition result; when the sensitive content exists in the to-be-published information, determining replacement information corresponding to the sensitive content based on the enterprise-level knowledge base, the replacement information being used for replacing the sensitive content in the to-be-published information to obtain target published information; and when the sensitive content does not exist in the to-be-published information, determining the to-be-published information as target published information. According to the embodiment of the invention, automatic compliance review of cross-modal content is realized, the manual review cost is effectively reduced, and the accuracy and timeliness of sensitive information processing are improved.

The invention relates to the technical field of intelligent gateways, in particular to a method and system for automatically expanding intelligent router plug-ins through LLM. The method comprises the steps that a user describes a plug-in function requirement of the intelligent router in a natural language through voice or text input equipment; and the cloud preprocessing service carries out character conversion processing on the voice data, carries out contradiction check and invalid information filtering on natural language requirements, extracts structured requirement information and inputs the structured requirement information to the cloud big language model. According to the intelligent router plug-in development method and system, the problems that in traditional intelligent router plug-in development, the manual intervention degree is high, the process is long and compatibility is insufficient are effectively solved by constructing an automatic system with the cloud large language model and the multi-service module cooperating. Through cooperation of multiple technical modules, a full-automatic process covering demand input, code generation, engineering integration, test verification and terminal deployment is constructed, and the intelligent router plug-in development efficiency and quality are remarkably improved.

The invention provides a cue word processing method, device and equipment, belongs to the technical field of artificial intelligence, and solves the problems that manual design of cue words is low in efficiency, lacks data verification and cannot be updated and iterated. The method comprises the steps of obtaining a first cue word and task rule data; processing the first cue word and the task rule data to obtain a second cue word; acquiring first induction strategy data; determining a third cue word and an error result set according to the first induction strategy data and the second cue word; and determining a target cue word according to the third cue word and the error result set. According to the scheme, automatic generation and closed-loop self-evolution of the cue word are realized, and the task suitability and stability are improved.

The invention discloses a method for extracting and processing text data of a PDF (Portable Document Format) document, which comprises the following steps: word processing: dividing the data content of a word object in the PDF document into three attributes of characters, paper coordinates and printing size, and carrying out differentiated processing according to the three attributes so as to obtain standard data of the word in an article; extracting character objects from the PDF document, extracting characters with coordinates in a table range region, and storing the characters into a character object set in sequence; table processing is carried out, data content of a table in the PDF document is divided into table structure objects, paper coordinates and printing sizes, and the table structure objects are divided into cell characters, cell sizes and row and column positions; extracting table objects in each page, and storing the table objects in a table object set in sequence; the problems that in the prior art, tables and characters in bidding document PDF documents are not accurately extracted, and the tables cannot be accurately recognized during page crossing are solved, common announcements, papers and scheme type PDF documents can be processed, and accurate article structure data can be restored.

The invention relates to the technical field of character processing, in particular to a Chinese character stroke extraction method based on domain self-adaption, which comprises the following steps of: expanding stroke data corresponding to Chinese characters in a standard character library; encoding the stroke order priori knowledge of each Chinese character in the standard character library to obtain a stroke splicing graph with stroke order information, and taking the stroke splicing graph as a stroke label of the standard character library; taking each Chinese character in the standard character library and the corresponding stroke spliced graph as source domain data, taking the unlabeled handwritten Chinese character as target domain data, and performing field adaptive training on a pre-constructed stroke extraction model by utilizing the source domain data and the target domain data; and inputting the handwritten Chinese character and the coded stroke splicing graph into the trained stroke extraction model, and outputting a multi-channel stroke image corresponding to the number of strokes of the handwritten Chinese character. According to the method, the accuracy and robustness of handwritten Chinese character stroke extraction can be improved.

The embodiment of the invention discloses a conference information processing system and a terminal, and belongs to the technical field of computers. The system comprises a terminal, a server and an edge device, the terminal is used for carrying out voice-to-word processing on the audio data of the conference to obtain a transliteration text corresponding to the audio data; the terminal is also used for determining target equipment from the server and the edge equipment based on the confidentiality level of the conference; sending the transliteration text to the target device; the server is used for processing the transferred text under the condition of receiving the transferred text to obtain a conference summary of the conference; sending the conference summary to the terminal, and displaying the conference summary by the terminal; the edge device is used for processing the transferred text under the condition of receiving the transferred text to obtain a conference summary of the conference; and sending the conference summary to the terminal, and displaying the conference summary by the terminal. According to the system, the conference information processing process is intelligently and efficiently realized through cooperation of the multi-terminal equipment, and manual arrangement of a user on the transferred text is avoided.

The invention provides a cue word processing method and device, equipment and a storage medium, and relates to the field of artificial intelligence, in particular to the fields of large models, big data, intelligent search and the like. According to the specific implementation scheme, multiple pieces of sample data are obtained, wherein the sample data comprise search problems belonging to a target field and target categories of the search problems; based on the multiple pieces of sample data, determining classification standard information conforming to each preset category in the target field; based on the plurality of candidate categories of each search question and the classification standard information of the candidate categories, splicing to obtain an initial cue word of each search question; the initial prompt word is used for indicating a category which the search problem conforms to is determined from a plurality of candidate categories; the plurality of candidate categories are at least partial categories in a plurality of preset categories; detecting the initial cue word of each search question to obtain a total detection result which is used for evaluating the cue effect of the initial cue word of each search question so as to adjust the initial cue word based on the total detection result.

Introduced here are computer programs and associated computer-implemented techniques for facilitating the creation of a master transcription (or simply “transcript”) that more accurately reflects underlying audio by comparing multiple independently generated transcripts. The master transcript may be used to record and / or produce various forms of media content, as further discussed below. Thus, the technology described herein may be used to facilitate editing of text content, audio content, or video content. These computer programs may be supported by a media production platform that is able to generate the interfaces through which individuals (also referred to as “users”) can create, edit, or view media content. For example, a computer program may be embodied as a word processor that allows individuals to edit voice-based audio content by editing a master transcript, and vice versa.

Techniques and systems are disclosed that perform content generation with integrated automated formatting using word processing applications that deploy large language models (LLMs). The techniques include receiving, a natural language (NL) query for a synthetic content for a document, identifying formatting rules of the document, and generating an augmented query that includes a representation of at least a portion of the NL query and a representation of the one or more formatting rules of the document. The techniques further include providing the augmented query to an LLM and updating the document with the synthetic content generated by the LLM in response to the augmented query.

The invention provides a large model calculation method and device oriented to a low GPU resource card group calculation system. The method comprises the steps that parameters of a part of a large language model are unloaded into a CPU memory in the GPU resource card group calculation system; obtaining a reasoning request, wherein the reasoning request comprises a batch processing size and a sequence length; in the cue word processing stage, whether the sequence length is larger than two times of the dimension of a hidden layer of the large language model or not is judged, and if yes, the calculation intensity of all attention layers and all multi-layer perceptron of the large language model is calculated; judging whether the calculation intensity of all the attention layers is greater than that of the multi-layer perceptron or not, if yes, unloading all the attention layers in a GPU video memory and unloading all the multi-layer perceptron in a CPU memory, and if not, unloading all the multi-layer perceptron in the GPU video memory and unloading all the attention layers in the CPU memory; and using the unloaded GPU resource card group computing system to run the large language model for reasoning to obtain a reasoning result of the reasoning request.

The invention discloses a conference summary transcription method and system based on multilevel semantic analysis, relates to the technical field of word processing, and solves the technical problem that conference summary generated in the prior art cannot ensure the overall accuracy and the front and back logic smoothness. The method comprises the following steps: 1, collecting a preliminary conference summary document, and verifying the identity of a spokesman in a conference process; 2, dividing the conference summary document into sub-documents according to the discussion theme and the agenda of the conference, and giving corresponding sub-themes for the sub-documents; 3, processing each sub-document in sequence, and evaluating the accuracy of the sub-documents after processing is completed; 4, after each sub-document is processed, adjacent speaks between the sub-documents serve as transition speaks, and the overall accuracy of the corrected conference summary document is further evaluated by evaluating the association strength of the transition speaks; according to the method, the conference summary is evaluated and adjusted from coarse to fine and then from fine to coarse, and multi-level analysis ensures that the content of the conference summary is logically smooth before and after.

The present application provides a method and device for processing polyphones, and a model training method and device, wherein the polyphone processing method includes receiving a Chinese sentence containing polyphones, and determining a first position feature of the polyphones in the Chinese sentence; segmenting the Chinese sentence, and determining the part-of-speech feature of the polyphones based on the words obtained by the word segmentation and the part-of-speech of the words; obtaining a word containing the polyphone, and determining a second position feature of the polyphone in the word containing the polyphone; inputting the first position feature, the part-of-speech feature, and the second position feature into a polyphone disambiguation model to obtain the pronunciation corresponding to the polyphone.

The application discloses a to-do task generation method and device, and relates to the technical field of speech recognition. The method comprises the following steps: segmenting a recording audio signal and extracting features of the recording audio signal to obtain recording audio segment features; grouping the recording audio segment features based on similarities in a similarity matrix, and taking the recording audio segment features in each recording audio segment feature group as a cluster; calculating the similarity between the clusters based on the similarity matrix; clustering based on the similarity between the clusters to obtain a clustering tree; assigning a speaker label based on the clustering tree to obtain a target recording audio signal segment, and then performing speech-to-text processing on the target recording audio signal segment to obtain text content corresponding to the target recording audio signal segment; dynamically constructing a prompt word based on a keyword table, a historical task template and the text content; taking the text generation type as a task type corresponding to a preloaded model; and generating a to-do task by using a task generation model according to the prompt word. In this way, the efficiency of to-do task generation is improved.

The invention provides a wake-up word processing method, a wake-up word processing device and a computer storage medium. The wake-up word processing method comprises the following steps: in response to a user registration instruction, inputting user registration information into an end-to-end acoustic model to obtain a registration lexical element sequence; mapping the registration lexical element sequence into a registration embedding vector; positioning a related core network of the end-to-end acoustic model by using the registration lexical element sequence; generating an acoustic encoder model using the related core network, the model calculation amount of the acoustic encoder model being less than the model calculation amount of the end-to-end acoustic model; and issuing the registration embedded vector and the acoustic encoder model to an end side device for wake-up word recognition. Through the wake-up word processing method, the cloud side cuts the acoustic model according to the user registration information and then issues the cut acoustic model to the end side device, so that the wake-up word detection efficiency and recognition precision of the end side device are improved.

The invention discloses a method and system for integrating retrieval and enhancing generation ability, and the method comprises the steps: S1, building a basic SQL set, and enabling each basic SQL to correspond to a preset business intention and to comprise a placeholder condition field; s2, establishing a mapping relation between fields and noun types; s3, performing intention analysis by using a large language model, and matching a basic SQL; s4, performing query condition reasoning, extracting specific nouns, and executing expansion word generation and compound word processing; s5, generating a complete SQL according to rule driving; s6, executing the SQL and dynamically correcting a query result to enable the query result to meet a context length requirement; and S7, inputting the question and result combination into the large language model to generate an answer. The method does not need a vector database, adapts to a small parameter model, is low in cost and high in real-time performance, and can be widely applied to various traditional service systems needing to integrate the AI dialogue ability.

The invention relates to the technical field of crossing of artificial intelligence and software engineering, and discloses an AI generation code recognition method based on a multi-modal feature fusion model, which comprises the following steps of: obtaining global features of a method source code sample to be recognized, and performing word processing on the whole source code by adopting a sub-word level word processing strategy; distinguishing the sub-words based on the sub-word types to realize sub-word level separation modeling of the code stream and the annotation stream; carrying out fusion through bidirectional cross attention to obtain window-level local semantic representation; splicing the window-level text local representation and the method-level global feature to obtain a window-level fusion vector; inputting the window-level fusion vector into a classifier to obtain a window-level probability set; when the maximum probability in the window-level probability set is larger than a threshold value during prediction, the method is judged to be AI code generation, and otherwise, the method is judged to be human compiled code. According to the method, the generalization ability of the model can be improved, and the detection ability of local AI generated fragments and the engineering application feasibility are improved.

The present application provides a method and device for detecting outliers in text data, relating to the field of text processing technology. The method can obtain a first benchmark sentence element and a benchmark threshold based on the sentence elements in the to-be-processed sentence set of target text data; obtain a preliminary sentence set based on the first benchmark sentence element, the benchmark threshold, and the to-be-processed sentence set; obtain an intersection sentence set corresponding to the preliminary sentence set based on the benchmark threshold and the preliminary sentence elements contained in the preliminary sentence set; obtain a fusion topic feature value based on the intersection sentence elements contained in the intersection sentence set; determine the outlier sentence elements of the target text data based on the first similarity value between the topic of the edge sentence elements contained in the edge sentence set and the fusion topic feature value; the edge sentence set is determined based on the elements in the preliminary sentence set excluding the intersection sentence set and the first benchmark sentence element. The method can detect outliers in text data and improve the overall quality of text data.

The mobile application enables users to make changes to texts located in the text fields; these reproductions show how to use the application; the user clicks on the “Scribens” icon in the foreground to display the main menu containing functionalities such as correction, rephrasing, translation, etc; the reproductions show the interface in English and the interface in French; the protection of the mobile application covers the menu opening method, the floating "Scribens" icon, the menu, and the graphic layout of the elements in the menu.