41results about How to "Improve error correction accuracy" patented technology

The invention discloses an error correction method based on a language model and word features. The method comprises the following steps: S1, acquiring a first sentence and inputting the first sentence into an error word detection system, and then detecting the first sentence and returning a suspect word by using the language model; S2, inputting that obtained suspect word into a candidate word recommendation system, selecting the candidate word by using at least two different similarity algorithms and outputting the candidate word; S3, replacing the suspect word in the first statement with the candidate word obtained in S2 to form a second statement, scoring the first statement and the second statement respectively, and selecting the statement with higher score for output. The system comprises a detection module, a recommendation module and a scoring module. Through the technical scheme, the error correction accuracy can be improved.

The invention provides an error correction method and device for search terms, wherein the method comprises the steps of recognizing wrong search terms; calculating weighted edit distance between the search terms and hot terms obtained in advance using weighted edit distance algorithm, wherein during the weighted edit distance calculating process, with respect to operations to transform the search terms into hot terms, respectively operation of inserting characters, operation of deleting characters, operation of replacing characters with similar form or pronunciation, operation of replacing characters without similar form or pronunciation and operation of switching characters, weightings of different numerical value are set; selecting hot terms of preset numbers to conduct error correction prompt according to the weighted edit distance and hotness of the hot terms. According to the error correction method for search terms , error correction accuracy rate for the wrong search terms can be reduced.

The invention provides equipment and a method for the error correction of query sequence based on the degree of error correction association, wherein the method comprises the steps of: acquiring a candidate error correction sequence corresponding to a query sequence input by a user through user equipment; selecting, from the candidate error correction sequence, the error correction sequence corresponding to the query sequence according to the degree of error correction association of the candidate error correction sequence; acquiring a query result corresponding to the error correction sequence according to the error correction sequence; and providing the query result for the user equipment. Compared with the prior art, the method completes reasonable selection for the corresponding error correction sequence through the degree of error correction association of the candidate error correction and provides the query result corresponding to the error correction sequence for user, which not only improves the error correction accuracy of search engine on the query sequence input by user, but also reinforces the interactivity between user and search engine so as to further improve the search experience of user.

The invention relates to a search engine-oriented error correction method and system of Chinese and English mixed querying. Based on an N-gram language model and a variety of error correction strategies, the method or system realizes error correction for the Chinese and English mixed querying with part of errors in a search engine.

The invention provides an artificial intelligence-based text error correction method and device and a computer readable medium. The method comprises the following steps of: obtaining an error corrected target fragment in an error correction text and an original fragment, in an original text, corresponding to the target fragment, wherein the target fragment is selected from a plurality of candidatefragments of the original fragment when error correction is carried out on the original text on the basis of a pre-trained fragment scoring model; obtaining feedback information of a target result fed back by a user on the basis of the error correction text; carrying out incremental training on the fragment scoring model according to the target fragment, the original fragment and the feedback information; and carrying out error correction on a subsequent original text on the basis of the trained fragment scoring model. According to the method, when the trained fragment scoring model is used for carrying out text error correction, the text error correction correctness can be effectively improved.

The invention discloses an error correction and detection method of an S mode ADS_B system. The error correction and detection method comprises the following steps of initializing parameters of the S mode ADS_B system, computing reference power value, extracting a received ADS_B information bit and analyzing confidence coefficient, and correcting and detecting errors of the S mode ADS_B system. The error correction and detection method has the advantages that according to the characteristics of ADS_B information data received by the ADS_B, a low confidence coefficient bit is regarded as a root of an ADS_B information error code through judging 'high' and 'low' confidence coefficient of each bit; on the basis that only five error bits can be corrected according to the existing method, a data stream is obtained through utilizing multiple XOR operations of error correction and CRC check codes, and is compared with a 24 bit data stream starting from a first low confidence coefficient bit, so that an error code larger than 5 bits but meeting a certain condition is further corrected; and according to the method adopted by the invention, the ADS_B information data which can be corrected is prevented from being lost, and the error correction and detection capacity of the S mode ADS_B system is improved.

The invention relates to a text error correction method and system, computer equipment and a computer readable storage medium, and the method comprises the steps: a data obtaining step: obtaining to-be-corrected text data; a negative sample construction step for creating a confusion word table and performing corpus replacement on the text to be corrected according to the confusion word table to generate a negative sample; a text error correction step used for taking the text data to be corrected and the negative sample data as training data, respectively training Chinese character features andpinyin features of the training data through a SoftMasked BERT pre-training model, splicing the Chinese character features and the pinyin features into a training result, and calculating cross entropy loss of the training result through a Softmax layer to obtain an error correction result; a model optimization step, wherein the SoftMasked BERT pre-training model is optimized through recursive prediction and word list filtering. Through the method, the text error correction accuracy is effectively improved, and the model effect and performance are improved.

The invention relates to a grammar error correction model training method and device, a grammar error correction method and device, computing equipment and a computer readable storage medium. The training method comprises the steps of performing data extension processing based on a first training set to obtain a second training set; obtaining a second source sample statement and a second target sample statement based on the second training set; inputting the second source sample statement into a grammar error correction model to generate an error correction sample statement; determining a lossvalue based on the error correction sample statement and the second target sample statement; and performing iterative training on the grammar error correction model based on the loss value until a training stop condition is met. By performing data enhancement processing on an existing training set, the purpose of automatically expanding the training set is achieved, and manual labor is effectively reduced.

The embodiment of the invention relates to the field of deep learning, and discloses a text model training method, a text error correction method, electronic equipment and a storage medium. Accordingto the method, text conversion is carried out on a voice conversation to obtain an error text, error correction is carried out on the error text to obtain a correct text, and word segmentation is carried out on the error text and the correct text according to a hot word list; and then, encoding the text sample after word segmentation to obtain a text training sample, and inputting the text training sample into a deep neural network for training to obtain a target text model. Compared with text error correction based on a rule model or a statistical model in the prior art, the word segmentationaccuracy is higher, and the error correction accuracy of the target text model obtained through model training is improved. Besides, error correction is carried out based on the established target text model, correct sentences can be obtained only by inputting wrong sentences in the error correction process, and the workload and cost in the error correction process are reduced.

The embodiment of the invention provides a text information processing method and device, a storage medium and electronic equipment, and the method comprises the steps: carrying out the word segmentation of a to-be-corrected text recognition result, and obtaining a plurality of text entries; inputting the to-be-corrected text recognition result into a search engine to obtain at least one search result; respectively matching each text entry in the plurality of text entries with the search result to obtain a matching result of the text entry in the search result; splicing the matching result corresponding to each text entry in the plurality of text entries to obtain a splicing result of each search result in the at least one search result, and forming a candidate result set by a set of the splicing results; and respectively matching the to-be-corrected text recognition result with each splicing result in the candidate result set, and determining a corrected text recognition result. Therefore, the accuracy of text recognition result error correction is improved.

The invention provides an ultra-wide-band complex-format vector modulation error correction method based on two-stage predistortion. A loop comprises a standard signal generation module, a first-stage predistortion error compensator, a second-stage predistortion error compensator, an origin offset compensator, a transmission channel and a vector modulator; the first-stage predistortion error compensator offsets an amplitude-phase frequency response fluctuating feature of a broadband vector modulator error reversely in a molded filter predistortion manner; the second-stage predistortion error compensator corrects a modulation error, with the broadband feature, after first-stage predistorion balancing in the whole work waveband, namely, the positive-angle error with I / Q gain imbalance and gain imbalance; and the origin offset compensator corrects a modulation error with a narrowband feature, namely carrier wave leakage, via an analog device. According to the invention, the vector modulation error with the broadband feature is corrected in a digital predistortion manner, and the broadband complex-format vector modulation error is corrected effectively.

The invention relates to the technical field of medical science and technology, and particularly discloses a text error correction method and device and a storage medium. The method comprises the steps of obtaining a to-be-corrected text in the medical field; inputting the to-be-corrected text in the medical field into a network model corresponding to the medical field after fine adjustment, and determining to-be-corrected words in the to-be-corrected text; and performing error correction on the words to be corrected through the network model to obtain an error-corrected text. The method is beneficial to improving the text error correction precision.

The embodiment of the invention discloses an error correction method and an error correction device, which are used for reducing the bit error rate of an FFE and improving the equalization performance. The method comprises the steps of obtaining an output signal and an amplitude value of a forward equalizer FFE, wherein the amplitude value is a channel response amplitude value corresponding to anequivalent channel of the FFE; carrying out level judgment on the output signal according to the amplitude value to obtain a first judgment signal, wherein the first judgment signal comprises (2M-1) judgment symbols, and M is an integer not smaller than 2; carrying out (1 / (1+D)) decoding on the first decision signal to obtain a first decoded signal, determining the first decoded signal as a seconddecision signal, wherein the second decision signal comprises (M-1) decision symbols; if the absolute value of the second decision signal is greater than (M-1), determining that a burst error occursin the second decision signal; and performing error correction on a burst error in the second decision signal.

The invention discloses a text error correction method and system in a power grid field scheduling scene, and the method comprises the steps: carrying out the word segmentation and named entity recognition of sentences in a text, and obtaining a word segmentation set and a named entity set; performing entity linking on the named entity set and a knowledge base to determine real information of entities; performing word boundary correction on the word segmentation set by using the named entity identification set to update the word segmentation set; adding pinyin features to each word in the wordsegmentation set to generate a new word segmentation set; inputting the new word segmentation set into a probability model to obtain the probability of occurrence of each word, and determining suspicious wrongly written characters according to the probability; obtaining a candidate set of the word from a pinyin dictionary according to the pinyin of the suspicious wrongly written character; substituting the candidate words in the candidate set into sentences one by one, regenerating features, substituting the features into the probability model, and determining the optimal candidate accordingto the probability. Suspicious wrongly written characters are positioned in combination with the probability model, so that the text error correction efficiency can be improved.

The invention provides a data error-check method and a device, as well as a disk playing method and playing device. The error-check method is used for checking the error of data accessed from a disk, comprising the steps as follow: the error-check information of one data block relative to the accessed data is obtained; the type of data error of the data flow unit from the data block is determined according to the error-check information; and whether the data flow unit is used for playing the disc is judged according to the type of data error and the preset reference type of data error. The error-check method also further comprises the steps as follow: the data which is used in the disc playing referred to the data flow unit is abandoned if the data flow unit is not kept. The invention further checks errors from the data through the error-check information of the abandoned data block in the prior art so as to eliminate the data error which can interrupt the playing of the disc.

The invention discloses an error correction model training method, in the scheme, an initial error correction model can be trained through a search word before error correction and a search word after error correction in target training data, and in the training process, a predicted search word is determined through an attention module, a coverage rate module and a gating module in the model. The final error correction model is obtained after training, and the error correction model is trained in this way, so that the error correction accuracy of the model can be improved, and the accidental injury rate of the model can be reduced; the invention further provides a search word error correction method, according to the scheme, error correction can be conducted on the original search word by means of the trained final error correction model, and the situation that the search quality is reduced due to errors existing in the original search word is avoided; the invention further discloses an electronic device and a storage medium which can also achieve the above technical effects.

Embodiments of the invention disclose a text error correction method, device and equipment, and a storage medium, which are applied to the field of digital medical treatment, and the method comprisesthe steps of obtaining text features of initial text information, and determining a to-be-retrieved text according to the text features; determining target domain knowledge from a domain think librarybased on the to-be-retrieved text; and generating a to-be-corrected text sequence according to the to-be-retrieved text and the target domain knowledge, and performing error correction on the to-be-retrieved text based on the to-be-corrected text sequence to realize error correction of the initial text information. By adopting the embodiment of the invention, domain knowledge can be associated toperform error correction on the text, and the text error correction accuracy is improved.

The embodiment of the invention provides an information error correction prevention method and device and an electronic device, wherein the method comprises the steps of after identifying that the inputted information needs to be corrected, determining the error correction candidate information corresponding to the inputted information, wherein the error correction candidate information comprisesan error correction candidate item and an error correction score of the error correction candidate item; determining a complete probability of a statement corresponding to the inputted information; determining a correction score of the error correction candidate item according to the complete probability and the error correction score; after the correction score meets a preset condition, displaying the error correction candidate item, so that the invalid error correction is effectively avoided, and the error correction accuracy is improved.

The invention discloses a data error correction method and a device on a flash storage medium and relates to the field of data storage. The data error correction method and the device on the flash storage medium are used for improving error correcting accuracy of data on the flash storage medium. Metadata which correspond to user data are written into a physical main sector and a physical backup sector. When the user data is read, the metadata in the physical main sector and the physical backup sector can conduct automatic error correction processing, and the metadata obtained by the automatic error correction processing conduct error correction to the user data. The data error correction method and the device on the flash storage medium can improve error correction accuracy of the data on the flash storage medium.

The invention relates to the technical field of artificial intelligence, and discloses a text error correction method and device, equipment and a storage medium, and the method comprises the steps: obtaining a to-be-corrected text sequence; inputting the to-be-corrected text sequence into the Bert model, identifying wrongly written characters in the to-be-corrected text sequence to obtain a wrongly written character set, and correcting the wrongly written character set based on a preset candidate character set to obtain a corrected target text sequence; sequentially extracting first target characters after error correction from the target text sequence, obtaining second target characters before error correction corresponding to the first target characters, and calculating font similarity and character similarity between the first target characters and the second target characters; and calculating an error correction judgment factor according to the font similarity and the character similarity, comparing the error correction judgment factor with a preset threshold value, and determining an error correction result of the text sequence to be subjected to error correction according to a comparison result. By means of the mode, the problem that class overfitting is caused due to the fact that a Bert model is too rough in word similarity judgment can be solved.

The invention provides a voice recognition method, a terminal and a computer readable storage medium. The method comprises the following steps: acquiring voice data, performing voice recognition on the voice data to obtain a plurality of candidate character strings, aligning the plurality of candidate character strings to obtain a plurality of aligned character strings, and processing the plurality of aligned character strings by using a trained error correction model to obtain a target character string corresponding to the voice data. According to the technical scheme provided by the invention, the accuracy of the speech recognition result can be improved.

The embodiment of the invention provides a method and a device for achieving grammar error correction by using a machine, and a training method and training equipment for a grammar error correction model. The method for achieving grammar error correction by using the machine comprises the following steps of: encoding an input text of a first language into an intermediate representation; decoding the intermediate representation into a text of a second language; and generating a corrected text of the first language subjected to grammar-correction based on the intermediate representation and the text of the second language. According to the method for achieving grammar error correction by using the machine, not only can possible small grammar errors be ignored, but also information loss in a back translation process can be avoided by combining intermediate representation, so that the accuracy of grammar error correction can be remarkably improved, and better experience is brought to a user.

The invention discloses a Chinese error correction method based on mutual information. The method comprises the following steps: obtaining a target short sentence to be corrected, performing word segmentation processing on the target short sentence, obtaining a word segmentation sequence, and determining a pinyin combination sequence corresponding to the word segmentation sequence , each pinyin combination in the pinyin combination sequence corresponds to each participle in the word segmentation sequence one by one, and based on the homophone words mapped by each pinyin combination in the pinyin combination sequence, the error correction word sequence set is obtained, and the error correction word sequence set is calculated. The mutual information of each error correction word sequence, according to the size of the mutual information, determines the error correction result of the target short sentence. By applying the technical solution provided by the embodiment of the present invention, the error correction of the target short sentence can be performed according to the frequency of words and the frequency of word collocation, and the accuracy rate of Chinese error correction can be improved. The invention also discloses a mutual information-based Chinese error correction device, a mutual information-based Chinese error correction device and a computer-readable storage medium with corresponding technical effects.

The invention relates to a correction system and method mainly used for a large-span gantry crane and particularly relates to an automatic correction system and method for traveling of a gantry crane. The automatic correction system comprises an encoder group, a programmable controller, a traveling motor for driving traveling mechanisms of rigid legs and flexible legs of the gantry crane and a frequency converter for controlling the traveling motor, wherein the encoder group is connected with the programmable controller, and the frequency converter is connected with the traveling motor. Therefore, the automatic correction system has the advantages that (1) the automation degree is high, and automatic correction can be realized without need of artificial intervention; (2) the error correction accuracy is high, the deviation degree can be accurately detected by utilizing the encoder group, the traveling motor is driven by the frequency converter to carry out accurate adjustment, simultaneously, the slipping phenomenon of the low-speed ends of traveling wheel groups can be found in time, and corresponding adjustment can be made; and (3) the application scope is wide, the deformation of a traveling track of a cart and the whole deflection of the cart can be judged and error operation can be prevented.

The invention discloses a data processing method and system, a storage medium and calculation equipment. The method comprises the steps that an original address text is obtained, and errors exist in the original address text; performing word segmentation processing on the original address text by utilizing the trained sequence labeling model to obtain an original first granularity geographic region and an original second granularity geographic region; processing the original first granularity geographic region by using a trained text generation model to obtain a target first granularity geographic region, the target first granularity geographic region being correct; and generating a target address text based on the target first granularity geographic region and the original second granularity geographic region. According to the data processing method and device, the technical problem that the processing accuracy is low if the address base does not contain the corresponding clean address text due to the fact that a data processing method in the related technology is achieved by searching the clean address text in the address base is solved.

The invention discloses a pinyin-based dialogue system text error correction system and a pinyin-based dialogue system text error correction method. According to the embodiment of the invention, afterabnormal information is detected in a dialogue system text based on pinyin, the abnormal information is matched with a set information list by adopting a set text fuzzy matching strategy (TFM), and then matched information is obtained; and similarity calculation is performed on the abnormal information and the obtained matched information through adoption of the set Pinyin-based text similarity algorithm (PTS) to obtain information with the highest similarity, and the abnormal information is corrected. According to the embodiments of the invention, the error correction accuracy is improved onthe basis of reducing the error correction cost.

The invention provides a text processing method and device. The text processing method comprises the steps of obtaining a text to be corrected; processing the to-be-corrected text according to a preset text error detection rule to obtain an initial error correction text; and inputting the initial error correction text into a text error correction model for processing, and determining a target correction text corresponding to the to-be-corrected text according to a processing result. The error correction accuracy is improved in a rule and model combined error correction mode, the error correction coverage range is wider, and the problem that rule maintenance is difficult due to the fact that rule error correction is independently adopted in the prior art is solved.

The invention discloses a human resource data processing system and method based on AI deep learning, and the method comprises the steps: obtaining the text information of human resource data before error correction in response to an error correction request of a client; identifying the type of the human resource data based on the text information before error correction; based on a preset mapping relation, selecting a preset text error correction subdivision model corresponding to the type; inputting the text information before error correction to the preset text error correction subdivision model to obtain target text information after error correction, and returning the target text information after error correction to the client. According to the method, the data association degree of the training data set for training the preset text error correction subdivision model is improved, so that the error correction effect of performing error correction on the text information before error correction of the human resource data is improved, and the error correction accuracy of performing error correction on the text information before error correction of the human resource data is improved.