Hybrid Spelling Correction and Query Expansion for Relevance Document Searching

Soyusiawaty, Dewi; Wolley, Denny Hilmawan Rahmatullah

doi:10.14569/ijacsa.2021.0120838

Cited by 7 publications

(4 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…𝐸𝐷 (𝑚, 𝑛) = min 𝐸𝐷 (𝑚, 𝑛 − 1) + 1 (5) where ED is the edit distance between m and n, which means the minimum number of steps required to transform Q into the first n-1 characters of T plus the cost of adding a letter to T, which costs a single operation or (6):…”

Section: Dynamic Programming For Edit Distancementioning

confidence: 99%

See 1 more Smart Citation

Automatic spelling error detection and correction for Tigrigna information retrieval: a hybrid approach

Desta

Lehal²

2023

Bulletin EEI

View full text Add to dashboard Cite

This paper proposes a hybrid approach to design and implement query spelling error detection and correction (SEDC) for Tigrigna information retrieval (IR). Our approach, which is the main contribution to this work, is fast and robust to achieve better performance and also helps the users to easily insert their corrected queries to retrieve relevant information from the IR. This is achieved by combining the normalized measure of bigram overlap using the Jaccard coefficient (J.C) technique, a dynamic programming algorithm for edit distance, and probability of occurrence, which were used to make suggestions for the misspelt words. Our approach was evaluated on the SEDC subtasks separately. It achieved an F-measure of 98.85% on the spelling error detection subtask and an accuracy of 95.36% on the spelling error correction subtask. Thus, a comparison was conducted between our approach and the existing Tigrigna spell checker. It is found that our approach outperformed the existing spell checker and shows a 5.36% improvement in accuracy. This is by far the most promising result with regard to correcting the misspelt users’ queries and improving the overall performance of the IR.

show abstract

Section: Dynamic Programming For Edit Distancementioning

confidence: 99%

“…When users type quickly, they may add or delete letters in an unintentional manner. As mentioned in [6], typing an adjacent key on the keyboard by accident is very common, especially on mobile devices with small virtual keyboards. Besides the typographical errors, some of the errors come from the difficulty of spelling itself.…”

Section: Introductionmentioning

confidence: 99%

Automatic spelling error detection and correction for Tigrigna information retrieval: a hybrid approach

Desta

Lehal²

2023

Bulletin EEI

View full text Add to dashboard Cite

show abstract

“…The Levenshtein Distance Algorithm is an algorithm made by Vladimir Levenshtein in 1965 [16,17]. This algorithm is represented in the form of a matrix to calculate the number of differences in the string, entered by the user with the string contained in the database [18][19][20][21][22]. The final result of the algorithm value is in the lower right corner of the matrix [23][24][25].…”

Section: Introductionmentioning

confidence: 99%

Implementation of Levenshtein Distance Algorithm for Product Search Query Suggestions on Koro Pedang Edutourism E-Commerce

Halimah Tus Sadiah,

Lia Dahlia Iryani,

Tjut Awaliyah Zuraiyah

et al. 2024

ARASET

View full text Add to dashboard Cite

Users sometimes write queries that are inaccurate or typos in the product search contained in the Koro Pedang Educational Tourism e-commerce, so the system is not find product search results because the query entered in the system is incorrect. This can frustrate users because they cannot find the product they are looking for, so the users leave the website. According to these problems, it is necessary to suggest a query on the product search function. This is expected to assist users in finding the product they are looking for if there is an error in typing the query. This research purposes were to implement the Levenshtein Distance Algorithm for product search query suggestions on Koro Pedang Educational Tourism e-commerce. The stages of this research, namely the development of the search module, implementation of the Levenshtein Distance Algorithm and testing. The implementation of the Levenshtein Distance Algorithm in the search function for Koro Pedang Educational Tourism e-commerce products, a Suggestion Query is generated for Query typos in the search function with an accuracy value of 90%, Precision 95% and Recall 90.9%. This shows that the performance of the algorithm that has been applied to the search function for query suggestion is very good. The application of the Levenshtein Distance Algorithm gives a positive value to the usability of searching for e-commerce products for Koro Pedang Educational Tourism.

show abstract

“…Its importance stems from the complex nature of the Chinese language, where characters are logograms, making CSC a unique challenge compared to alphabetic languages. CSC finds applications in search optimization [1], relevant document retrieval [2], autocomplete services [3], and downstream NLP tasks like location extraction [4], OCR [5], ASR [6], parts-of-speech tagging [7], and text classification [8].…”

Section: Introductionmentioning

confidence: 99%

WOSES: A Word Segmentation-Enhanced Framework for Chinese Spelling Check

Li,

Tang,

Mao

et al. 2023

Preprint

View full text Add to dashboard Cite

Chinese spelling check (CSC) is a critical task in processing Chinese texts, impacting a variety of down-streaming tasks, such as information retrieval and search optimization applications. Traditional CSC methods have primarily focused on character-level phonetic and graphic information, often neglecting the crucial role of word segmentation. Our preliminary study reveals that typos frequently lead to incorrect word segmentation, and appropriately identifying word boundaries significantly enhances CSC accuracy. To address this gap, we introduce the Word Segmentation-Enhanced Speller (WOSES), a novel framework that uniquely incorporates word segmentation into the spelling correction process. WOSES is further developed into two distinct variants: the Hierarchical Word Speller (H-WSpeller) and the Word Speller (WSpeller). While both share a common architecture, they differ in their submodule interaction mechanisms, with H-WSpeller focusing on a layered approach and WSpeller employing parallel processing for enhanced efficiency. Our evaluation, conducted using benchmark datasets SIGHAN13, SIGHAN14, and SIGHAN15, demonstrates the effectiveness of our methods. On SIGHAN13 and SIGHAN15, WOSES variants outperform state-of-the-art models, and they achieve comparable results on SIGHAN14. Crucially, our analysis shows that incorporating word segmentation, whether derived from the original text or predicted, significantly boosts the CSC performance, underscoring the importance of this aspect in CSC tasks.

show abstract

Hybrid Spelling Correction and Query Expansion for Relevance Document Searching

Cited by 7 publications

References 15 publications

Automatic spelling error detection and correction for Tigrigna information retrieval: a hybrid approach

Automatic spelling error detection and correction for Tigrigna information retrieval: a hybrid approach

Implementation of Levenshtein Distance Algorithm for Product Search Query Suggestions on Koro Pedang Edutourism E-Commerce

WOSES: A Word Segmentation-Enhanced Framework for Chinese Spelling Check

Contact Info

Product

Resources

About