102 results about "Edit distance" patented technology

Some embodiments generate prompts and submit them in queries to a language model trained on code to perform automated program repair. Some embodiments fix syntactic mistakes and semantic mistakes by combining multimodal prompts, iterative querying, test-case-based selection of few-shots, and program chunking. In some cases, edit distance is minimized between an initial flawed program and the automatically created improved version of that program. The initial flawed program is obtained from a programming student, or from a source code generator.

To provide a comparison method and a search method of strings to be optimized mainly focusing on Chinese search, based on a string search process of data information of a nuclear fuel management process.SOLUTION: These methods include steps of: splitting and preprocessing a target string and integrating a corresponding language library; calculating an editing distance and performing preliminary filtering using a character distance algorithm; and searching a Chinese string and verifying its collation degree using the Aho-Corasick algorithm, and caching frequently searched pattern strings after searching.SELECTED DRAWING: Figure 1

The invention belongs to the technical field of informatization management, and particularly relates to a data import method based on similarity calculation, automatic matching of file fields and database fields is realized by calculating the similarity of the file fields and the fields in a database, and the workload of manual configuration can be effectively reduced; a common editing distance calculation similarity algorithm is improved, the influence of the longest common subsequence on similarity calculation is considered, a similarity calculation formula is improved, and the matching success rate is increased; when the similarity is calculated, literal matching and semantic matching can be carried out by adopting a method of combining an improved editing distance similarity calculation algorithm and word semantic similarity calculation based on the Zhinetting, so that the matching success rate is effectively improved; besides automatic matching, a user configuration interface is provided, a user can manually adjust the mapping relation, and the flexibility and usability of the method are improved; user configuration content can be automatically recorded, and repeated configuration work of a user is avoided.

The invention provides a structured analysis and semantic template normalization method for an unmanned vehicle operation log. For the problems of unstructured unmanned vehicle log data, changeable formats, complex semantics and the like, logs are converted into semantic vectors by adopting a multi-language sentence vector coding model, and semantic grouping is performed by using a small-batch K-means clustering algorithm after dimension reduction is performed by adopting an incremental principal component analysis method; extracting a variable field from a clustering result by using a regular expression and generating a standardized template; and intelligent template combination is realized by calculating cosine similarity and editing distance between the templates. The method has the following three technical characteristics: 1) the template consistency is improved by adopting a strategy of clustering first and then normalizing; 2) introducing double similarity constraints to reduce template redundancy; and 3) retaining variable sequence and semantic information to support subsequent analysis. The method is suitable for log data analysis and intelligent operation and maintenance of various complex systems.

The invention discloses a rationality analysis method for acquisition and bidding evaluation information of power equipment, which comprises the following steps: constructing a regular expression to perform paragraph matching on preprocessed bidding document and bidding document texts, and extracting specific parameter contents; obtaining an absolute error and a relative error based on each bidding value and each bidding value, and comparing the absolute error and the relative error with a preset threshold to obtain a numerical deviation degree grade of the bidding file relative to the bidding file; the method comprises the following steps: splitting a bid invitation file and a bidding file into character sequences, creating a two-dimensional table, analyzing to obtain an editing distance value of each text, obtaining a comprehensive matching degree of the text through accurate matching and fuzzy matching based on a preset synonym library, and further obtaining a text deviation degree grade; constructing a deviation degree comprehensive evaluation rule matrix based on the numerical deviation degree grade and the text deviation degree grade of the bidding file relative to the bidding file, and obtaining a corresponding comprehensive evaluation result; according to the invention, the efficiency and accuracy of collection and bidding evaluation of the power equipment are improved, and personal errors and compliance risks are reduced.

The invention discloses a detection method and device, equipment, a medium and a product, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining target to-be-detected content; detecting the target to-be-detected content based on a preset plagiarism detector and / or a text detector, and outputting a statistical index and / or a content source corresponding to the target to-be-detected content; according to the statistical indexes and / or the content source, large language model generation content in the target content to be detected is determined, and the statistical indexes comprise at least one of a similarity measurement index, a syntax tree detection index, an editing distance index, a text style statistical index, a linguistic feature statistical index and a fact detection index; the method for detecting the large language model generated text is simple, convenient and accurate, the large language model generated content can be rapidly determined through the detected statistical indexes and / or content sources, and the detection effect of the large language model generated content is improved.

The invention discloses a character string fuzzy matching method and system, and relates to the field of character recognition, and the method comprises the steps: obtaining a plurality of characters of different font sizes and fonts, and constructing a font similarity matrix; in the replacement operation in the distance editing method, character similarity weighting is carried out on the cost of the replacement operation so as to construct a weighted distance editing method; a weighted editing distance is used for the two character strings to be matched, the obtained weighted editing distance is normalized, and the similarity of the two character strings to be matched is obtained; when the similarity is 1, the matching is successful; when the similarity is greater than a set similarity threshold value and less than 1, representing a suspicious matching result; when the similarity is smaller than a set similarity threshold value, the matching fails. The method can solve the problem that character errors exist in similar results recognized by the OCR technology.

The invention relates to a large language model double watermark embedding method, which comprises the following steps of: after generating the probability of each character according to an input cue word, circularly embedding a plurality of first characters by using a short watermark, and embedding a long watermark for a residual text to obtain a text embedded with double watermarks. According to the method, the robustness and the soundness of the watermarks are simultaneously ensured through double watermarks, the robustness of the watermarks is ensured through repeatedly embedded short watermarks, the soundness of the watermarks is ensured through long watermarks covering a full text, partial errors of the short watermarks are tolerated through editing distances, and partial errors of the long watermarks are corrected through error correction codes.

The embodiment of the invention provides a large-model-based natural language query analysis and database field matching method, which comprises the following steps of: performing intention analysis and entity recognition on natural language query of a user based on a large language model (LLM), and extracting a potential keyword set; for each extracted keyword, binding an entity type through an STAM mechanism, and constructing an entity-type mapping relationship; a column description vector library is constructed, efficient retrieval is achieved in combination with an approximate nearest neighbor ANN algorithm, and results of semantic vector matching, editing distance matching and traceability enhancement matching are combined through a multi-strategy fusion matching mechanism to form a final candidate column set; a database metadata interface is connected, information is analyzed, a basic structure is constructed, semantic enhancement field description is generated, and output is organized in a tetrad structure; and reversely deducing the affiliated table through column matching, judging the table structure value by combining the information density in the table and the connectivity with other tables, eliminating redundant tables, and optimizing database representation.

The invention discloses an equipment fault diagnosis method and device based on a dynamic knowledge graph and a large model fine tuning technology. The method comprises the following steps: firstly, identifying a core entity from multi-source heterogeneous equipment fault data through a named entity identification model for fine tuning of domain data and a relation extraction model for special fine tuning of a fault diagnosis domain corpus, mining deep semantic association, and injecting the deep semantic association into a graph database after cleaning to form an initial knowledge graph; receiving user natural language fault description, realizing term and standard entity linking through editing distance fuzzy matching and Sension-BERT semantic vector similarity calculation, and combining bidirectional retrieval and attention mechanism fusion to obtain an enhanced context; and finally, generating a structured diagnosis report containing thinking chain reasoning based on an enhanced context by utilizing a specialized fine-tuning fault diagnosis large language model. According to the method, the limitation of a traditional diagnosis method is effectively solved, high-precision and interpretable equipment fault diagnosis is realized, and the diagnosis efficiency and reliability are improved.

Detecting an anomalous configurator of a plurality of managed entities. A plurality of configuration trees, each representing configuration parameters at a corresponding managed entity, include annotations that identify a configurator that configured each parameter. For a particular configurator, a plurality of subtrees is generated from the plurality of configuration trees. A set of weighted edit distances are calculated from the plurality of subtrees, each representing a degree of difference between a different pair of subtrees. A distance matrix is populated with the set of weighted edit distances, and the distance matrix is used to identify anomalous subtree(s) within the plurality of subtrees for the particular configurator. In embodiments, a configuration corresponding to an anomalous subtree considered to have been anomalously applied by the particular configurator. Data that identifies at least one managed entity associated with the anomalous subtree, and the particular configurator is stored.

The invention discloses an entity word error correction method and device based on an editing distance, electronic equipment and a medium, and the method comprises the steps: obtaining a voice recognition text, inputting the voice recognition text into a pre-trained entity word detection model, and obtaining a to-be-corrected first entity word; determining a first syllable sequence of the first entity word; obtaining a plurality of second entity words according to a preset entity word list, and determining a second syllable sequence of each second entity word; determining an editing distance between the first entity word and the second entity word according to the first syllable sequence and the second syllable sequence, and selecting a target entity word from the second entity word according to the editing distance; and replacing the first entity word in the speech recognition text with the target entity word. According to the invention, entity word error correction is carried out on the basis of the speech recognition result by using the editing distance, the speech recognition efficiency and accuracy are improved, and the method can be widely applied to the technical field of speech recognition.

The invention discloses a method and device for extracting concerned phrases in a text, a storage medium and equipment, and belongs to the technical field of computers. The method comprises the following steps: acquiring a reverse index and a multimode matching model which are constructed for concerned phrases; detecting an input text by utilizing the multi-mode matching model to obtain a plurality of fragments; collecting the plurality of fragments by using an inverted index to obtain a first fragment set; segmenting the first fragment set according to a preset editing distance to obtain multiple groups of second fragment sets; for each group of second fragment sets, screening candidate fragments according to editing distances between fragments in the second fragment sets and the concerned phrases; and combining the candidate fragments in each group of the second fragment sets into similar variants of the concerned phrases, and identifying the similar variants of which the editing distances to the concerned phrases are greater than a predetermined threshold value as the concerned phrases. According to the method, the similar variants of the concerned phrases in the text can be identified as the concerned phrases by utilizing the improved multi-mode detection model and the inverted index, so that the extraction accuracy is improved.

The invention discloses a static iteration construction method and system for a numerical expression type metamorphic relation, and relates to the technical field of software metamorphic tests.The static iteration construction method comprises the steps that numerical expression parameters are obtained to construct an editing distance matrix, and the similarity of numerical expression similarity analysis parameters is calculated; constructing a backtracking rule, and obtaining an editing operation for character string conversion by editing the distance matrix to obtain an initial metamorphic relation set; static iteration construction of the numerical expression type metamorphic relation is carried out, and a metamorphic relation set is output. According to the method, a tested program does not need to be executed through metamorphic relation compounding, calculation time consumption is reduced, the metamorphic relation construction cost is reduced, metamorphic relations which are meaningless to software testing are eliminated by utilizing the property of the numerical expression type metamorphic relation, the number of the metamorphic relations participating in compound calculation is reduced, and the metamorphic relation construction efficiency is improved.

The invention discloses a text-to-instruction system based on fuzzy pinyin matching and parameter robust analysis. The text-to-instruction system comprises an input preprocessing module, a text-to-instruction conversion module and a text-to-instruction conversion module, the fuzzy pinyin matching module is used for uniformly transferring the standardized text and a pre-stored instruction template into a silent pinyin sequence, and constructing a weighted editing distance matrix based on a preset pinyin character replacement cost; the parameter alignment search module is used for executing beam search on the weighted editing distance matrix and outputting a plurality of candidate matching paths with the minimum cost and a corresponding parameter fragment set; the parameter robust distribution module is used for performing legality and consistency checking on parameters in the candidate matching paths, completing parameter standardization and performing error correction in combination with a parameter candidate dictionary; the instruction output module is used for encoding the template identifier and the parameter key value pair obtained through robust distribution into a structured instruction and outputting the structured instruction; according to the method, text-to-instruction conversion can be completed in real time.

The invention discloses a text data enhancement method based on government affair text language disease error correction, which comprises the following steps: constructing a proper noun and terminology dictionary, and recording the positions and types of proper noun and terminology; analyzing a syntactic structure and literary and physical characteristics according to a proper noun and terminology dictionary, and constructing a language disease generation rule; the method further comprises the following steps: preprocessing the to-be-processed text data to obtain a preprocessed text; the preprocessed text is translated back, wherein the preprocessed text is translated into a text of a target language, and then the text of the target language is translated back to an original language; comparing the texts before and after the back translation, and screening out new texts of which semanteme is not damaged and which have language diseases as enhanced data; generating a negative sample on the basis of the enhanced data by using a language disease generation rule, and expanding the negative sample as an unprocessed negative sample; the editing distance of the unprocessed negative samples is calculated, and texts with the editing distance smaller than a threshold value are reserved to serve as the finally generated negative samples.

The invention relates to an OCR (Optical Character Recognition) medical record text error correction method, which comprises the following steps of: obtaining visual similarity and stroke similarity among characters of a Chinese character set hans to form a visual similarity matrix and a stroke similarity matrix; the method comprises the following steps: acquiring a to-be-recognized text, and setting a sliding window to execute the following steps on a text T of the to-be-recognized text: performing bidirectional maximum matching, recognizing the text T as a term if matching succeeds, otherwise, recognizing alternative terms, calculating a maximum difference rate, calculating a minimum editing distance between the text T and the alternative terms, and calculating a score and an offset rate; judging the text T as an alternative term based on the offset rate, otherwise, judging the text T as an undetected term part; and a forward set FTs and a reverse set BTs are obtained, and error correction is carried out. Compared with the prior art, the method has the advantages of improving the accuracy and availability of the medical record text data and the like.

The invention discloses a similarity calculation method based on semantic editing fusion. The method comprises the following steps: calculating global semantic similarity of a source text and a target text by utilizing a semantic coding model; obtaining a first keyword list of the source text and a second keyword list of the target text through word segmentation processing, taking each first keyword in the first keyword list as a target word, and respectively forming a plurality of to-be-compared word pairs with a corresponding word in the second keyword list and a plurality of adjacent second keywords; based on the to-be-compared word pairs, calculating the semantic similarity between each group of target words and the corresponding words by utilizing a semantic coding model, and determining the editing distance between the source text and the target text; and determining the local semantic similarity of the source text and the target text according to the editing distance, and combining the global semantic similarity to obtain the final text matching similarity. According to the method, the problem that text matching only depends on character-level surface matching and neglects semantic association between word pairs is solved, and the text matching precision is improved.

The invention relates to the technical field of electronic design automation, and provides a PCB schematic diagram component information extraction method based on text recognition and a graph algorithm. The method aims at solving the problem that component models in a complex PCB schematic diagram are extracted inaccurately, and the method mainly comprises the steps that preprocessing of binaryzation, denoising and text area enhancement is conducted on the PCB schematic diagram, and an optimized image is generated; performing character direction detection and rotation processing on the obtained optimized image, inputting the fine-tuned OCR model to perform multi-scale character recognition, and outputting a recognition result containing character content, coordinates and confidence; constructing a graph model containing MCU position nodes and character block nodes, and establishing associated edges through a weight calculation module; performing global optimal matching on the graph model based on a Hungary algorithm, and screening associated pairs of the MCU and the character blocks; and performing editing distance calculation on a matching result and a preset MCU model template library, correcting character confusion errors through a multi-stage verification mechanism, and outputting final component information.

The invention provides a database operation and maintenance knowledge graph construction method and system for optimizing entity alignment, and the system comprises 1, a data preprocessing module which carries out the data cleaning of unstructured and semi-structured data; 2, a knowledge extraction module which extracts valuable information and knowledge from the cleaned data, including entities, relationships and attributes; the entity alignment module is used for aligning the same or similar entities in different data sources; and 4, a knowledge representation and storage module which performs standardized representation on the extracted and aligned knowledge and stores the knowledge in a knowledge base. According to the method, the context attention weight is introduced, and the semantic similarity between the words is calculated by calculating the attention weight of the context word vector of each word to the target word vector, so that the accuracy of similarity calculation is improved; and a hybrid editing distance algorithm is utilized to bring the semantic similarity among the words into an editing distance calculation process, so that the calculation method of the overall sentence similarity is optimized. Insertion, deletion and replacement operations of characters are considered, and word similarity calculation of a semantic level is also combined, so that sentence similarity calculation has a semantic understanding capability.

This application relates to the field of fintech and discloses a dynamic hybrid parsing and deterministic verification system and method for fuzzy bond instructions. The system constructs structured data through an instruction preprocessing module; a routing decision module calculates complexity scores based on structure, semantics, and business characteristics, and generates path selection instructions using a resource-aware dual-threshold strategy; a hybrid parsing module responds to instructions, employing a rule engine to handle low-complexity tasks and a large-scale model with enhanced retrieval to handle highly ambiguous tasks; a fallback verification module performs confidence-weighted integration and logical consistency checks, and repairs anomalies through issuer set operations and edit distance matching; a mapping generation module constructs standard query statements using an abstract syntax tree; and a closed-loop optimization module dynamically updates parameters based on user interaction feedback. This invention effectively balances computational efficiency and parsing accuracy, eliminates model illusion risks, and achieves adaptive evolution of the system.

This invention relates to the fields of audio signal processing and artificial intelligence, specifically to a lyrics alignment method and electronic device based on automatic speech recognition. The lyrics alignment method based on automatic speech recognition in this application uses a pre-trained automatic speech recognition model to recognize audio samples and generate recognized text. The edit distance sequence between the recognized text and the lyrics text is calculated, and spoken segments are determined by analyzing the distance change pattern through a sliding window. These spoken segments are then replaced with silent segments, effectively eliminating spoken segments and avoiding interference from non-lyric content in the alignment process, thus laying the foundation for subsequent alignment.

Methods and systems are provided for hallucination prevention for natural language insights. In embodiments described herein, a template-based insight with a set of facts is generated by a template-based insights engine. The set of facts are generated from a set of data and the template-based insight is generated based on a text template. A natural language insight is generated from the template-based insight using a language model. If a single fact of the template-based insight is missing from the natural language insight, the single missing fact is an integer and a remaining integer of the natural language insight is within a threshold edit distance of the integer, the hallucination of the natural language insight is corrected by replacing the remaining integer with the single missing fact.

The invention discloses a method, a system and a terminal for generating a dynamic report through visual configuration, and the method comprises the following steps: metadata configuration: defining a report name and a code through a visual interface, and establishing a mapping relation between a query condition attribute name and a display name; engine binding is executed, wherein an SQL mode and a storage process mode are included; the SQL mode comprises the step of automatically replacing an SQL script containing an attribute name placeholder with an actual parameter value; the storage process mode comprises the steps of performing standardization processing on an interface attribute name and a storage process parameter name, and matching parameters by adopting an editing distance algorithm allowing 1-2 character differences; and multi-table rendering: independently rendering the master table and the slave table according to the bound query result serial number, and configuring a column display rule and an exclusive script. According to the method, the development efficiency and the parameter matching accuracy are improved, and complex structures such as listing are supported; configuration takes effect in real time, and the operation and maintenance cost of restarting service in a traditional scheme can be avoided.

The invention discloses a memory multi-thread fuzzy retrieval method. The method comprises the following steps: respectively calculating an editing distance similarity value and a common sequence similarity value of an entry and a user retrieval word by adopting a multi-thread parallel operation mode; and calculating a mean value of the editing distance similarity value and the common sequence similarity value corresponding to each entry and the search word, sorting the mean values, and selecting the first N character strings from the mean values as fuzzy search results. According to the method, under the condition of not depending on a big data retrieval component, a reliable result can be efficiently searched from million-level or even ten million-level entries in an extremely short time according to the retrieval word input by the user.

The application discloses a low-complexity forward-backward decoding method based on weighted edit distance, and the method comprises the following steps: b A code word sequence with a length of N L symbols is generated by an encoder of an LDPC code d A mark code w is uniformly inserted into the code word sequence d to generate a sending code word with a length of N c and output the sending code word x After the sending code word x passes through an insertion / deletion-substitution channel, a receiving sequence with a length of N y is generated; a watermark decoder decodes the receiving sequence y by using a low-complexity forward-backward decoding method and outputs a likelihood ratio sequence l ; and an LDPC decoder decodes the likelihood ratio sequence l and outputs the application stores the calculation result of an intermediate metric value in a lookup table, reduces the number of repeated calculations of the intermediate metric value, reduces the calculation complexity of the decoding algorithm and improves the decoding speed.

The invention relates to a code error feedback method combining static analysis and dynamic analysis, and mainly solves the problems of insufficient error feedback and low efficiency in an Online Judge system: reconstructing codes of correct and error versions, including alignment and renaming of variables, and rearranging statements on the premise of not influencing semantics; static and dynamic multi-level alignment is carried out, in the static alignment, an editing script with the minimum cost is calculated through a heuristic editing distance algorithm, an alignment block and a difference list are obtained, in the dynamic alignment, compiling execution is carried out on source codes of two versions, test cases which do not pass operation are dynamically printed through gdb, and execution track information of the test cases is dynamically printed through gdb; error positioning is achieved, data flow backtracking analysis and scoring are carried out on error positions, and error feedback is generated. The method has the characteristics of high execution efficiency, high accuracy and easier understanding of generated feedback, automatic feedback of error codes can be realized, and the learning efficiency of students in programming tasks is improved.

This invention provides a Pinyin error correction method that simplifies the noise channel model used in most current Pinyin error correction algorithms by employing a real-time frequency counting method, effectively improving the efficiency and lightweighting of the algorithm. Furthermore, this invention uses a direct character substitution method instead of the traditional edit distance calculation method, avoiding frequent edit distance calculations in the Pinyin error correction algorithm. The dictionary database itself is localized to the individual user, resulting in a highly personalized, targeted, and small-scale dictionary that provides accurate candidate words. Compared to existing algorithms, this invention offers significant improvements in error checking rate, candidate word accuracy, and execution efficiency.

The invention provides a program error positioning and explanation generation method and system based on reinforcement learning, and belongs to the field of AI auxiliary programming education and natural language processing, and the method comprises the steps: S1, based on student submission records of a real teaching scene, screening out error codes and correct code pairing samples through an editing distance, and constructing a training set; s2, sampling and filtering the training set by using a pre-training language model, and constructing a sampling rejection data set to supervise and finely adjust the pre-training language model to obtain a base model; s3, constructing a multi-signal fusion reinforcement learning reward function for performing reinforcement learning training on the base model to obtain a trained error positioning and explanation generation model; and S4, inputting the question description and the error code into the trained error positioning and interpretation generation model, and generating structured feedback. According to the method, the error positioning accuracy is remarkably improved, the false drop rate is effectively reduced, and a feasible technical path is provided for intelligent programming education.