HINDIA: a deep-learning-based model for spell-checking of Hindi language

Singh, Shashank; Singh, Shailendra

doi:10.1007/s00521-020-05207-9

Cited by 14 publications

(13 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this work, different comparisons have been presented with other works that use other techniques, or are based on other assumptions. Because automatic spell-checking systems are currently of great interest in different languages [32][33][34][35], a challenge will be to test this approach on them.…”

Section: A Final Discussion About the Characteristics Of Ar2p-textmentioning

confidence: 99%

Automatic Spell-Checking System for Spanish Based on the Ar2p Neural Network Model

Puerto,

Aguilar,

Pinto

2024

Computers

View full text Add to dashboard Cite

Currently, approaches to correcting misspelled words have problems when the words are complex or massive. This is even more serious in the case of Spanish, where there are very few studies in this regard. So, proposing new approaches to word recognition and correction remains a research topic of interest. In particular, an interesting approach is to computationally simulate the brain process for recognizing misspelled words and their automatic correction. Thus, this article presents an automatic recognition and correction system of misspelled words in Spanish texts, for the detection of misspelled words, and their automatic amendments, based on the systematic theory of pattern recognition of the mind (PRTM). The main innovation of the research is the use of the PRTM theory in this context. Particularly, a corrective system of misspelled words in Spanish based on this theory, called Ar2p-Text, was designed and built. Ar2p-Text carries out a recursive process of analysis of words by a disaggregation/integration mechanism, using specialized hierarchical recognition modules that define formal strategies to determine if a word is well or poorly written. A comparative evaluation shows that the precision and coverage of our Ar2p-Text model are competitive with other spell-checkers. In the experiments, the system achieves better performance than the three other systems. In general, Ar2p-Text obtains an F-measure of 83%, above the 73% achieved by the other spell-checkers. Our hierarchical approach reuses a lot of information, allowing for the improvement of the text analysis processes in both quality and efficiency. Preliminary results show that the above will allow for future developments of technologies for the correction of words inspired by this hierarchical approach.

show abstract

Section: A Final Discussion About the Characteristics Of Ar2p-textmentioning

confidence: 99%

Automatic Spell-Checking System for Spanish Based on the Ar2p Neural Network Model

Puerto,

Aguilar,

Pinto

2024

Computers

View full text Add to dashboard Cite

show abstract

“…Using a sequence-to-sequence neural network, they obtained a 5% improvement in results compared to the previous model. Singh and Singh 46 corrected spelling errors in Hindi using deep learning. In this process, the misspelled words were automatically identified by the network and replaced by the closest word.…”

Section: Deep-learning-based Methodsmentioning

confidence: 99%

“…There are four different modes for the model to respond to the actual model. Since the model aims to identify and correct misspelled words, we consider misspelled word identification as positive like Singh and Singh study 46 . In our application, these states are: the real phrase has a misspelling and the network or rules have edited it correctly (TP), the real phrase has a misspelling but the network or the rules have not edited it correctly or it has not recognized incorrectly (FP), the real phrase is without spelling mistakes and the network or rules correctly recognized it without mistakes (TN), and the real phrase is without spelling mistakes but the network recognized it as having a spelling mistake (FN).…”

Section: Evaluation Measuresmentioning

confidence: 99%

Correcting spelling mistakes in Persian texts with rules and deep learning methods

Kasmaiee,

Homayounpour

2023

Sci Rep

View full text Add to dashboard Cite

This study aims to develop a system for automatically correcting spelling errors in Persian texts using two approaches: one that relies on rules and a common spelling mistake list and another that uses a deep neural network. The list of 700 common misspellings was compiled, and a database of 55,000 common Persian words was used to identify spelling errors in the rule-based approach. 112 rules were implemented for spelling correction, each providing suggested words for misspelled words. 2500 sentences were used for evaluation, with the word with the shortest Levenshtein distance selected for evaluation. In the deep learning approach, a deep encoder-decoder network that utilized long short-term memory (LSTM) with a word embedding layer was used as the base network, with FastText chosen as the word embedding layer. The base network was enhanced by adding convolutional and capsule layers. A database of 1.2 million sentences was created, with 800,000 for training, 200,000 for testing, and 200,000 for evaluation. The results showed that the network's performance with capsule and convolutional layers was similar to that of the base network. The network performed well in evaluation, achieving accuracy, precision, recall, F-measure, and bilingual evaluation understudy (Bleu) scores of 87%, 70%, 89%, 78%, and 84%, respectively.

show abstract

“…There are two main methods for checking non-word errors, one based on dictionary pairing and another based on N-Gram statistical analysis. The dictionary pairing approach is to pair the words to be checked with a pre-given list of dictionaries, and the main methods currently used for querying dictionaries are the binary tree, the finite-state automaton and the hashing [9]. The method based on N-Gram statistical analysis is an N-Gram partitioning of the words to be checked, where each N-Gram string in a word is matched with an N-Gram frequency table that is counted and computed from the correct corpus [10].…”

Section: Related Workmentioning

confidence: 99%

“…The minimum edit distance method [11] is to analyze the difference of strings morphological by the number of operations to transform the previous string into the later one by substitution, deletion, insertion and interchange, and calculating the minimum edit distance required for the transformation, the smaller the distance, the more likely it is to be the correct word. Similar skeleton keys are used to transform a word into a skeleton key that can represent the word by set rules, and then match strings with similar string morphology or similar pronunciation by skeleton keys [9]. The misspelling statistics dictionary-based approach uses a misspelling statistics dictionary that has been developed by manually counting a large number of texts in which non-word errors occur.…”

Section: Related Workmentioning

confidence: 99%

An Errors Correction Model for the Errors of Non-word and Real-word in English Composition

Huang¹,

Huang²

2022

Journal of Compurters

View full text Add to dashboard Cite

<p>In the procedure from composing English, it is inevitable to face the phenomenon of word writing errors. In recent years, English composition automatic correcting system has attracted much attention. However, the precision of the existing word errors correcting system is vague generalization. So as to move forward the accuracy of checking and correcting word errors, this paper designs a word errors correction model based on natural language processing technology. This model designs phoneme matching method based on an improved IDM algorithm, and combined with a non-word input errors correction method based on character distance. The accuracy of correcting non-word errors in this model reached 86.5%. The study also proposes a real-word errors correction method, which is implemented basing on the real-word confusion set and combining the binary statistical model and the GloVe word vector model, improving the real-word errors correction method based on feature annotation of the real-word confusion set, with an accuracy of 77.9%. </p> <p> </p>

show abstract

HINDIA: a deep-learning-based model for spell-checking of Hindi language

Cited by 14 publications

References 24 publications

Automatic Spell-Checking System for Spanish Based on the Ar2p Neural Network Model

Automatic Spell-Checking System for Spanish Based on the Ar2p Neural Network Model

Correcting spelling mistakes in Persian texts with rules and deep learning methods

An Errors Correction Model for the Errors of Non-word and Real-word in English Composition

Contact Info

Product

Resources

About