58 results about "Adnoun" patented technology

The invention relates to the technical field of data processing, in particular to a method and system for processing data based on a large language model.The method comprises the following steps that based on inter-sentence punctuation positioning and part-of-speech tagging, sentence blocks are divided to generate functional partitions, themes and relational words are extracted to judge semantic chain starting points, a trigger index sequence is constructed, and logic jump points and breakpoint positions are recognized; and mapping the label structure to the language model output analysis deviation, and generating a label mapping combination list. According to the method, semantic turning nodes can be captured by analyzing the semantic direction change trend, the relation between the semantic turning nodes and verb and noun combinations is judged, the position of a trigger point of an actual information transfer effect is extracted, and the break point area of a semantic path is recognized through the positioning of key word starting and stopping blocks and the logical judgment of noun group cross combination; the path integrity has a clear fracture identifier, a traceable semantic mapping structure path is established in a language model, and the accuracy, coherence and hierarchy clearness of semantic reconstruction are enhanced.

Systems and methods are described for obtaining accurate responses from large language models (LLMs) and chatbots, including for question and answering, exposition, and summarization. These systems and methods accomplish these objectives via use of noun phrase avoiding processes such as a noun phrase collision detection process, a query splitting process, and a topical splitting process as well as by use of formatted facts, formatted fact model correction interfaces (FF MCIs), bounded-scope deterministic (BSD) neural networks, processes and methods, and intelligent storage and retrieval (ISAR) systems and methods. These systems and methods avoid and bypass noun phrase collisions and correct for errors caused by noun phrase collisions so that hallucinations are eliminated from LLM responses.

A system and method for generating question-answer pairs is disclosed. The system and method can receive a document. A sentence and / or a further sentence in the document may be identified. A syntactic map for the sentence and / or the further sentence may be generated. Noun phrases and prepositional phrases may be identified based on the syntactic map. Sentence level questions may be generated based on phrases identified using natural language processing (NLP) techniques. Document level questions can also be generated based on syntactic maps generated and NLP techniques.

Systems and methods are described for obtaining accurate responses from large language models (LLMs) and chatbots, including for question and answering, exposition, and summarization. These systems and methods accomplish these objectives via use of noun phrase avoiding processes such as a noun phrase collision detection process, a query splitting process, and a topical splitting process as well as by use of formatted facts, formatted fact model correction interfaces (FF MCIs), bounded-scope deterministic (BSD) neural networks, processes and methods, and intelligent storage and retrieval (ISAR) systems and methods. These systems and methods avoid and bypass noun phrase collisions and correct for errors caused by noun phrase collisions so that hallucinations are eliminated from LLM responses.

Disclosed herein are methods and systems for generating and using a shared knowledge language. In one embodiment, a method comprises retrieving data generated as a result of at least one computing device accessing one or more applications from a set of applications; generating a computer model comprising a set of nodes where each node having an identifier corresponding to a series of nouns and verbs generated in accordance with a schema associated with a shared knowledge language; when the processor receives a request: identifying a node that corresponds to the request; retrieving a pair of related nodes within the nodal data structure the processor identifying whether the relevance threshold has been satisfied using each respective node's identifier corresponding to the schema.

The invention relates to the technical field of natural language processing, in particular to an AIGC content safety monitoring system and method based on dynamic reasoning and context awareness, and the system comprises a semantic graph construction module which is used for extracting entity nouns and predicate verbs in a text according to a received AIGC interaction text flow, generating a semantic concept node set, and sending the semantic concept node set to a database; and performing directed connection and hierarchical nesting on the concepts in the semantic concept node set according to a logic direction according to a subject-predicate-object dependency relationship rule. According to the method, the curvature value of the semantic track is calculated, and the similarity between the direction vector and the center of the sensitive semantic cluster is combined for double verification, so that sudden turning of an intention in a dialogue process or progressive induction to a sensitive field can be perceived, and abnormal mutation can be recognized through a curvature pulse form; therefore, hostile attack behaviors are accurately captured in real time in dynamic interaction, and the defense capability for context dependent attacks and implicit induction behaviors is improved.

The invention relates to the technical field of geographical name translation, and discloses a multi-language syllable segmentation and adaptive learning combined geographical name translation method, which comprises the following steps of: performing word segmentation and general name translation annotation on a source language geographical name address based on a general dictionary, converting an unannotated proper name part into an IPA phoneme sequence, and translating the IPA phoneme sequence into an IPA phoneme sequence; performing syllable division and translation matching on the IPA phoneme sequence of each word segment, matching a plurality of possible candidate transliteration results for each word segment, and then combining each candidate transliteration result of the special name word segment with a general name translation result to construct a place name address target language translation result candidate set; and performing adaptive feature learning and aggregation optimization on the candidate set by using a deep learning algorithm so as to comprehensively consider semantic information and context coherence of a multi-transliteration version and generate a final place name address target language translation. According to the application, the problems of inaccurate transliteration, rigidity, ambiguity and the like in the traditional place name and address translation method can be solved, and the translation efficiency and quality are improved.

The invention relates to the technical field of contract intelligent auditing, in particular to an international trade contract intelligent auditing system based on natural language processing, which comprises a data acquisition preprocessing module, a key vocabulary extraction module, a key vocabulary analysis screening module and a text similarity association module, the data acquisition preprocessing module carries out word segmentation and segmentation on a contract text and a regulation standard text, then the key word extraction module extracts a key word set in each text segment through a word frequency statistical method, and in order to prevent the situation that the key word set extracted by the key word extraction module is inaccurate, the key word set is extracted by the key word extraction module. The key vocabulary analyzing and screening module analyzes the paraphrases of the key vocabularies in the corresponding words, removes the key vocabularies which are not nouns in the key vocabulary set, provides a paraphrasing basis for the vocabulary same paraphrasing mapping unit, unifies same paraphrasing word vectors, and transmits the same paraphrasing word vectors to the key vocabulary same paraphrasing mapping unit; and the contract and the regulation text are deeply associated in terms of vocabulary semantics.

The present invention provides a method, device, and robot for generating knowledge for solving applied problems, comprising: obtaining input text corresponding to a target applied problem input by a user; segmenting the input text into sentences and assigning a corresponding sentence number to each acquired text sentence; obtaining the noun content associated with each text sentence to form noun content recognition knowledge; and matching noun content extension knowledge associated with the input text from a pre-constructed knowledge base based on the noun content associated with each text sentence and its corresponding sentence number. The method, device, and robot for generating knowledge for solving applied problems provided by the present invention can generate noun content extension knowledge that is not contained in the applied problem itself but is of great value in solving the problem, thereby more effectively assisting the development of various systems related to applied problems and improving the accuracy of problem solving in various systems to a certain extent.

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.

This invention relates to the field of information retrieval structure technology, specifically to a method and system for monitoring data during the investment promotion process in industrial parks. The method includes the following steps: extracting investment promotion text, identifying combinations of verbs and resource nouns, analyzing word order and semantic features, filtering action-driven segments, dividing structural sequences and identifying connection relationships, establishing semantic jump paths, reorganizing sentence order, and generating a data monitoring structure template for investment promotion in industrial parks. In this invention, by extracting verbs and resource nouns to construct sentence content, combining word order features and semantic tags to classify and filter segments, dividing structural sequences based on the position of resource nouns and merging content with consistent semantic direction, mapping connectors to jump paths to construct task behavior sequences, and realizing the logical organization and sequential reorganization of sentences, this generates a data expression method with traceability and structured features, improving the semantic recognition efficiency and data monitoring capabilities during the investment promotion process in industrial parks.

A method for anticipating actions includes receiving video data indicating an object in an environment, encoding a noun embedding using a noun encoder based on the video data, and encoding a verb embedding with a verb encoder based on the video data. The method also includes generating a verb anticipation associated with the object in the environment by processing the noun embedding using a verb decoder, generating a noun anticipation associated with the object in the environment by processing the verb embedding using a noun decoder, and generating an action anticipation by combining the verb anticipation and the noun anticipation.

The present application discloses a method, device, equipment and storage medium for eliminating hallucinations in large language models, which relates to the technical field of natural language processing and includes: constructing an association network among target nouns according to the association relationship among the target nouns, and obtaining the association paths corresponding to the target nouns according to the association network; determining a reference path from each association path, and obtaining the entropy difference value between the first target entropy corresponding to each association path and the second target entropy corresponding to the reference path; determining the hallucination risk type corresponding to the initial text to be analyzed based on the entropy difference value, and optimizing the initial text to be analyzed by using an optimization strategy corresponding to the hallucination risk type, so as to eliminate the hallucinations in the analysis result after analyzing the initial text to be analyzed by a preset large language model. By combining semantic analysis with probability statistics, it is ensured that no external knowledge base is relied on during the hallucination analysis process, and the generalization ability and reliability of the hallucination processing method are guaranteed.

The invention aims to provide a method and equipment for correcting regional noun errors. Compared with the prior art, the method and the device have the advantages that the preset number of correct corpora containing the regional nouns are constructed, the regional nouns in each corpus are subjected to erroneous character construction based on the preset erroneous character construction rule, the constructed erroneous character corpus is determined, model training is performed based on the correct corpora and the erroneous character corpus, and the trained error correction model is determined; wherein the error correction model comprises region noun recognition and region noun error correction, and then region noun error correction is carried out on sentences containing region noun based on the trained error correction model. In this way, the effect of correcting regional nouns can be improved, and the error correction accuracy can be improved by assisting the text error correction function through the entity recognition function.

To accurately quantify the content characteristics of a document.SOLUTION: An information processing device 1 is connected to a service server 3, such as an SNS, via a communication line. The information processing device is equipped with means for collecting data regarding posts made by a specific user on an SNS service or the like from the server, means for performing natural language analysis processing on the posts and breaking them down into parts of speech, means for extracting multiple character strings corresponding to nouns among the parts of speech, and means for obtaining, as feature quantities that quantify the content characteristics of the posts, coordinates in an N-dimensional space where each of N categories classifies the extracted multiple character strings according to the meaning of the nouns, each axis corresponds to one category, and each axis is assigned with many character strings representing concrete subordinate concepts included in that category, arranged in order according to their semantic similarity, with each coordinate value pre-assigned.SELECTED DRAWING: Figure 1

This application relates to the field of speech semantics, and provides a question-answering method, device, equipment, and storage medium based on dual-modal feature fusion. The method includes: obtaining an input image and an input question sentence, extracting the target region features of the input image through a trained image detection model, and inputting the target region features into a trained convolutional neural network to obtain average image features. Performing part-of-speech tagging on the input question sentence to obtain a tagged question sentence. Extracting the word vectors of the tagged question sentence through a trained text detection model to obtain average text features. Fusing the average image features and the average text features to obtain fused features. Inputting the fused features into a long short-term memory neural network for decoding to obtain an answer text. The average image features do not contain background and have a relatively low noise content. The average text features obtained based on only the question words and relevant nouns are not affected by other words. The answer text obtained from the average image features and the average text features is also relatively accurate.

The present application relates to the field of place name translation technology and discloses a place name translation method that combines multilingual syllable segmentation with adaptive learning. The method performs word segmentation and common name translation annotation on the source language place name address based on a general dictionary, converts the unannotated proper noun portion into an IPA phoneme sequence, and then performs syllable segmentation and translation matching on the IPA phoneme sequence of each word segment. Multiple possible candidate transliteration results are matched for each word segment. The candidate transliteration results of the proper noun segment are then combined with the common name translation result to construct a candidate set of place name and address target language translation results. The candidate set is then adaptively learned and aggregated using a deep learning algorithm to comprehensively consider the semantic information and contextual coherence of multiple transliteration versions to generate the final place name and address target language translation. The present application can solve the problems of inaccurate, stiff, and ambiguous transliterations in traditional place name and address translation methods, thereby improving translation efficiency and quality.

This invention discloses a slow-fast diversity-aware prototype learning method for egocentric action recognition. This method utilizes a part-to-prototype (P2P) module and a slow-fast contextualization (SFC) mechanism. In P2P, partial tokens obtained from each frame by the video converter are further interacted with several initial prototypes through diversity-aware attention, refining / extracting the semantic information of the interacting objects from the partial level to the prototype level. In SFC, slow and fast awareness attention is introduced to interact with the original prototypes to finely supplement semantic information. Three streaming features are aggregated using slow and fast gates, ultimately yielding classification results for nouns and verbs.

This invention relates to a prompt-based machine translation method, belonging to the field of natural language processing technology. It solves the problems of inaccurate, omitted, and mistranslated nouns and proper nouns in existing machine translation models. By constructing a set of nouns and their translations from the text to be translated, the input text and adjustment matrix of the translation model are obtained. The translation model is then used to translate the input text, and the attention calculation of the model is adjusted using the adjustment matrix M, ultimately outputting the translation. Based on the input data containing noun translation prompts and the adjustment matrix, the accuracy of the noun translation by the translation model is guaranteed to a certain extent, solving the problems of omitted and mistranslated nouns, and improving the accuracy of noun translation in machine translation models.

The invention provides a character correction method and device and electronic equipment, and the method comprises the steps: obtaining a to-be-processed character, and enabling the to-be-processed character to be obtained through the conversion of voice data collected in real time; extracting to-be-processed nouns from to-be-processed characters by using the trained name recognition model; searching a target name of which the semantic similarity with the to-be-processed noun is greater than a semantic similarity threshold value; judging whether the phonetic similarity between the target name and the to-be-processed noun is greater than a phonetic similarity threshold value or not; and if yes, correcting the to-be-processed noun in the to-be-processed character as the target name. In the implementation process of the scheme, a dual verification mechanism combining semantic similarity and phonetic similarity is adopted, the semantic similarity ensures that candidate words conform to context logic, the phonetic similarity ensures that pronunciation or fonts of the candidate words are close to those of original words, the problems of excessive error correction and correction omission possibly caused by a single standard are avoided through combination of the semantic similarity and the phonetic similarity, and the accuracy of correction is improved. Therefore, the accuracy of converting voice into characters is obviously improved.

The invention relates to the technical field of intelligent document review, in particular to an intelligent document review method and system based on large model fusion causal reasoning, and the method comprises the steps: collecting documents, extracting formats and null lines, recognizing roles and actions, retrieving nouns after verbs, processing semantic repetition, outputting a table, checking causal connections, comparing action sequences, screening and joining, ranking verbs, and pulling through actions. And outputting causal fragments, checking consistency, separating objects, calibrating behaviors, outputting chain fracture, splitting adverbs, sorting relationships, and outputting conflict paths and review paths. According to the method and the device, by extracting the paragraph formats, roles and actions of the documents, the meticulousness of review is improved. And analyzing a causal conjunction and an action relationship, and identifying a causal relationship and a semantic conflict. The method comprises the steps of calibrating an action sequence, identifying missing parts, sorting multiple actions and role changes, optimizing an examination path, reducing manual intervention and improving efficiency.

The application belongs to the field of natural language processing, and particularly relates to a text key phrase extraction method, a storage medium and a device that integrate inter-sentence correlation, comprising: extracting a nominal phrase in a part-of-speech combination mode; filtering the nominal phrase by combining two Trie trees to obtain a candidate phrase set; calculating a global semantic similarity score of the candidate phrase; clustering each sentence to obtain a sentence cluster containing different semantic information; and sorting the candidate phrases in the sentence cluster according to the global semantic similarity score of the candidate phrase to obtain a key phrase set; the application integrates the inter-sentence relationship in the unsupervised key phrase extraction model based on embedding, improves the accuracy of the model, and constructs different Trie trees to calculate mutual information and left and right information entropy as a candidate phrase extraction method in the key phrase extraction model, thereby reducing the probability of incomplete semantic information appearing in the extracted candidate phrase set.

The invention discloses an object illusion relieving method and device for a multi-modal large language model, and the method comprises the steps: inputting a picture and a description instruction into a large visual language model for analysis, so as to generate a candidate description set corresponding to the picture; extracting an object noun of each candidate description in the candidate description set to obtain an object set; obtaining a conditional instruction, and calculating a conditional probability corresponding to each object noun in the object set according to the conditional instruction to obtain a confidence coefficient corresponding to each object noun; obtaining a final description corresponding to the picture according to the confidence corresponding to each object noun; therefore, by reducing excessive dependence of a large visual language model on language priori, the object illusion problem in an image description task is effectively relieved.

The present application provides a sentence rewriting method, apparatus, system, and storage medium. The sentence rewriting method comprises: obtaining at least one entity noun from a historical sentence based on a target entity recognition model; obtaining time information for each entity noun in the at least one entity noun; and inputting the sentence to be rewritten, the entity noun, and the time information into a target rewriting model to obtain a rewritten sentence output by the target rewriting model. Some embodiments of the present application can improve sentences to be rewritten based on entity nouns extracted from historical texts and the time at which the entity nouns were generated, effectively reducing rewriting time while improving rewriting efficiency and quality.

Embodiments of the present disclosure relate to a method, an apparatus, an electronic device, and a medium for text translation. The method includes determining a keyword set associated with a chapter-level monolingual corpus in a target language, where the keyword set includes a plurality of entity words and a plurality of pronouns, and masking the chapter-level monolingual corpus based on the keyword set. The method further includes generating a chapter-level text translation model based on the masked chapter-level monolingual corpus. According to the embodiments of the present disclosure, it is possible to enable translations of the same or associated words to have contextual consistency throughout a text, and to explicit a noun indicated by a pronoun, and further to supplement a missing pronoun, thereby improving accuracy of the text translation model.

The application relates to the technical field of natural language processing, and discloses a Chinese numeral quantifier correction method and system, computer equipment and a storage medium. The application constructs a dynamically expandable collocation word table and a numeral dictionary; performs sentence segmentation, word segmentation and part-of-speech tagging on input text, and identifies a three-element structure composed of numerals, quantifiers and nouns; excludes misjudgments on inherent expressions through a fixed phrase filtering mechanism; performs legality verification on the structure based on the collocation word table, and generates processing information for semantic correction if the verification fails; when collocation abnormalities are confirmed and the dictionary is insufficient, a large language model is triggered to replace the quantifier and adapt to the context based on the processing information, so that a final corrected sentence is generated. Through multi-layer cooperation of rule matching, word table verification and large model verification, the application effectively reduces the mis-correction rate while ensuring high-precision correction, and improves the accuracy and practicality of automatic correction of Chinese quantifiers.

The present disclosure provides a new word screening method and device and electronic equipment, relates to the technical field of data processing, and is used for solving the problem of how to quickly and effectively identify new words in the field of text processing and information mining. The method comprises the following steps: obtaining a text to be analyzed; performing text processing on the text to be analyzed to determine at least one text segment; determining a statistical parameter of each text segment according to the total number of text segments and the total number of each text segment; in the case that the statistical parameter meets a preset condition and the characters contained in the text segment meeting the preset condition are weak words, determining that the characters contained in the text segment meeting the preset condition are new words; wherein the weak words include words that have been used in the target field but are not nouns.

The application relates to an artificial intelligence technology and discloses a question expansion method and device, equipment and a storage medium. The method comprises the following steps: extracting an inquiry mode word and an entity noun in a to-be-expanded question; extracting a standard question containing the inquiry mode word and a synonymous question with the same meaning as the standard question in a question and answer library, and marking the inquiry mode word and the inquiry mode word contained in the synonymous question as a keyword; performing word segmentation and part-of-speech tagging on the standard question and the synonymous question to obtain a sentence structure; extracting a front adjacent substantive word and a rear adjacent substantive word of the keyword in the standard question and the synonymous question according to the sentence structure; extracting the inquiry mode word in the keyword, and extracting the front adjacent substantive word and the rear adjacent substantive word; and according to the keyword, the front adjacent substantive word, the rear adjacent substantive word and the proper noun, composing a preset number of expansion questions in a preset grammar format. The application can automatically generate expansion questions according to an input question.

The embodiment of the present disclosure provides a kind of aerospace field-oriented open vocabulary target detection method and device, it is related to target detection technical field.The method comprises: the image to be detected and the description text corresponding to the image to be detected are input into the detection model pre-trained, obtain at least one mark frame and the target prediction category of the object in each mark frame of detection model output, target prediction category is the entity noun included in description text.Detection model is obtained according to pre-constructed sample dataset training, sample dataset includes self-built dataset;For each sample image in self-built dataset, the real category of sample object is used as training label, real category includes entity noun and extension word, by jointly as training label with key word and extension word, it can directly incorporate domain expertise into model training, enhance the semantic understanding ability of detection model in aerospace field, effectively improve the generalization ability and accuracy of detection model identification unseen class.

Systems and methods are described for obtaining accurate responses from large language models (LLMs) and chatbots, including for question and answering, exposition, and summarization. These systems and methods accomplish these objectives via use of noun phrase avoiding processes such as a noun phrase collision detection process, a query splitting process, and a topical splitting process as well as by use of formatted facts, formatted fact model correction interfaces (FF MCIs), bounded-scope deterministic (BSD) neural networks, processes and methods, and intelligent storage and retrieval (ISAR) systems and methods. These systems and methods avoid and bypass noun phrase collisions and correct for errors caused by noun phrase collisions so that hallucinations are eliminated from LLM responses.