KidSpell: A Child-Oriented, Rule-Based, Phonetic Spellchecker

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 6937–6946 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 6937 KidSpell: A Child-Oriented, Rule-Based, Phonetic Spellchecker Brody Downs∗, Oghenemaro Anuyah∗, Aprajita Shukla∗, Jerry Alan Fails∗, Maria Soledad Pera∗, Katherine Wright‡, Casey Kennington∗ ∗Department of Computer Science & ‡Department of Literacy, Language and Culture Boise State University, USA {brodydowns, oghenemaroanuyah, aprajitashukla}@u.boisestate.edu, {jerryfails, solepera, katherinewright, caseykennington}@boisestate.edu Abstract For help with their spelling errors, children often turn to spellcheckers integrated in software applications like word processors and search engines. However, existing spellcheckers are usually tuned to the needs of traditional users (i.e., adults) and generally prove unsatisfactory for children. Motivated by this issue, we introduce KidSpell, an English spellchecker oriented to the spelling needs of children. KidSpell applies (i) an encoding strategy for mapping both misspelled words and spelling suggestions to their phonetic keys and (ii) a selection process that prioritizes candidate spelling suggestions that closely align with the misspelled word based on their respective keys. To assess the effectiveness of KidSpell, we compare the model’s performance against several popular, mainstream spellcheckers in a number of offline experiments using existing and novel datasets. The results of these experiments show that KidSpell outperforms existing spellcheckers, as it accurately prioritizes relevant spelling corrections when handling misspellings generated by children in both essay writing and online search tasks. As a byproduct of our study, we create two new datasets comprised of spelling errors generated by children from hand-written essays and web search inquiries, which we make available to the research community. Keywords: spelling correction, children, phonetic, child spelling datasets 1. Introduction Spelling is an essential literacy and life skill, the basics of which are taught to children during their first few years of school (Berkling, Kay, 2018). Cultivating spelling is con- sidered to be a good reflection of children’s understanding and learning of the alphabet, as well as a good predictor of their reading skills (Puranik et al., 2011). In teach- ing spelling skills to students, teachers traditionally center instruction based on children’s five spelling development stages: precommunicative, semiphonetic, phonetic, transi- tional, and correct (Gentry, 1982). However, due to chil- dren’s varying cognitive capabilities, some of them still struggle to spell as they progress in their education. To assist children in improving their learning abilities, teach- ers now incorporate the use of Assistive-Technology in the classroom (Simpson et al., 2009; Lawley, 2016), includ- ing spellcheckers that support spelling skill development. Spellcheckers are often embedded in software utilized in the classroom environment, e.g., word processors like Mi- crosoft Word (Kutuzov and Kuzmenko, 2015), as well as search engines like Google (Wang and Zhao, 2019). There are a number of spellcheckers, along with corpora, that can be leveraged in new developments (Kutuzov and Kuzmenko, 2015; Huang et al., 2013; Mitton, 2009). While these spellcheckers are tuned to the needs of tra- ditional users, they prove unsatisfactory for children. More- over, there is no empirical evidence on the applicability of these …

Content truncated to 500 words