Current practices and future direction of artificial intelligence in mathematics education: A systematic review

Current practices and future direction of artificial intelligence in mathematics education: A systematic review Awang Yusop Danaee

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Current practices and future direction of artificial intelligence in mathematics education: A systematic review 2025 International Electronic Journal of Mathematics Education 10.29333/iejme/16006 awang2025-current-practices-and-future-direc awang2025-current-practices-and-future-direc.txt, the project's staged copy 8f1d74875c0a8ce670e1a5361f4786126b67fa1f open-access education

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them existence Quotation withheld because this record is not cleared for redistribution. artificial intelligence in mathematics education Quotation withheld because this record is not cleared for redistribution. systematic-review systematic-review Quotation withheld because this record is not cleared for redistribution. mechanisms learning analytics turns large volumes of student data into identification of struggle points, which lets teachers intervene in a targeted way and so improves outcomes Quotation withheld because this record is not cleared for redistribution. mechanisms tool type is matched to learner maturity: chatbots require the capacity to handle complex interactive systems and so suit older students, while AR/VR carries engagement for younger ones Quotation withheld because this record is not cleared for redistribution. The five-year (2019-2024) window may miss longer-term developments in the field. what this limits: the review's temporal coverage: THIS TIMEFRAME MAY NOT CAPTURE LONG-TERM DEVELOPMENTS Quotation withheld because this record is not cleared for redistribution. Only English-language publications were included, excluding multilingual research. what this limits: the linguistic coverage: THIS STUDY FOCUSED ON ENGLISH-LANGUAGE ARTICLES DUE TO PRACTICAL CONSTRAINTS Quotation withheld because this record is not cleared for redistribution. The review's map of the field is a map of the market: it covers only tools that are already commercialised and market-ready. what this limits: this review as a survey of AI in mathematics education: it is scoped to commercialised, market-ready tools, so non-commercial and research-only systems are outside it by design Quotation withheld because this record is not cleared for redistribution. The literature searched excludes non-English publications and every non-journal form, so the picture also excludes conference and book literature. Quote is from the criteria table. what this limits: the coverage of this review: the criteria table excludes non-English publications, book series, books, book chapters and conference proceedings Quotation withheld because this record is not cleared for redistribution. The evidence base is 32 articles, reached after 24 were deleted by the exclusion criteria and 25 removed as redundant - so the synthesis rests on a set whose size is set as much by the filters as by the literature. Quotation withheld because this record is not cleared for redistribution. 81 remaining articles screened on the complete text, with 24 deleted against the stated exclusion criteria eliminates: articles that meet the search string but not the question - screening on the full text rather than the abstract is what catches them · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. commercialized AI tools commercialized AI tools Quotation withheld because this record is not cleared for redistribution. systematic review systematic review Quotation withheld because this record is not cleared for redistribution. market-ready market-ready Quotation withheld because this record is not cleared for redistribution. AIME AIME Quotation withheld because this record is not cleared for redistribution. AI in mathematics education - AS THIS REVIEW MAPS IT: commercialised, market-ready tools, reported as a deliberate bridge from research to practice rather than as a survey of the research Quotation withheld because this record is not cleared for redistribution. the gap this review addresses - stated as a gap in ACCESSIBILITY rather than in knowledge: the synthesis missing is of tools educators can readily adopt Quotation withheld because this record is not cleared for redistribution. the gap this review fills - stated as a gap in ACCESSIBILITY rather than in knowledge: a shortage of syntheses of tools 'that educators can readily adopt', which makes adoptability the criterion of relevance Quotation withheld because this record is not cleared for redistribution. extends → Hwang (2022) Quotation withheld because this record is not cleared for redistribution. extends → Mohamed et al. (2022) Quotation withheld because this record is not cleared for redistribution. extends → Niño-Rojas et al. (2024) Quotation withheld because this record is not cleared for redistribution. frequency count frequency count Quotation withheld because this record is not cleared for redistribution. Most studies of AI in mathematics education were conducted at the middle/high school level. Quotation withheld because this record is not cleared for redistribution. The AI tools used in mathematics education fall into eight named types: adaptive learning systems, intelligent tutoring systems, chatbots, augmented reality, virtual reality, serious games, learning analytics and mathematics problem solvers. Quotation withheld because this record is not cleared for redistribution. Chatbots were the most frequently studied AI tool category in the 2019-2024 window, with virtual reality tools second. Quotation withheld because this record is not cleared for redistribution. Chatbot studies appear only at university and adult-learner level, never in school-level mathematics classrooms. Quotation withheld because this record is not cleared for redistribution. AI-driven serious games and mathematics problem-solver applications are the two least-researched categories in the corpus. Quotation withheld because this record is not cleared for redistribution. Most commercialized tools identified are pedagogical rather than generative, seventeen of them being teaching-and-learning platforms such as GeoGebra, Khan Academy and Squirrel AI. Quotation withheld because this record is not cleared for redistribution. Augmented and virtual reality technologies raise students' motivation and problem-solving skills in mathematics. Quotation withheld because this record is not cleared for redistribution. The reviewed literature converges on a call for longitudinal work measuring long-term effects of AI in mathematics education on learning outcomes, motivation and educational equity. Quotation withheld because this record is not cleared for redistribution. artificial intelligence (AI) artificial intelligence (AI) Quotation withheld because this record is not cleared for redistribution. systematic review systematic review Quotation withheld because this record is not cleared for redistribution. Generative AI Tools Generative AI Tools Quotation withheld because this record is not cleared for redistribution. Pedagogical AI Tools Pedagogical AI Tools Quotation withheld because this record is not cleared for redistribution. publication period publication period Quotation withheld because this record is not cleared for redistribution. literature type literature type Quotation withheld because this record is not cleared for redistribution. search category search category Quotation withheld because this record is not cleared for redistribution. N = 32 — number of primary studies included in the final synthesis Quotation withheld because this record is not cleared for redistribution. 36 out of 117 — articles removed at title-and-abstract screening as irrelevant Quotation withheld because this record is not cleared for redistribution. 43.8%, N=14 — share of studies conducted at middle/high school level, against 31.2% (N=10) at university/adult level and 25% (N=8) at primary level Quotation withheld because this record is not cleared for redistribution. seven categories — number of AI tool categories the 32 studies were classified into Quotation withheld because this record is not cleared for redistribution. N=35 — total category frequency count, higher than the 32 studies because some studies used more than one tool type Quotation withheld because this record is not cleared for redistribution. twenty-one (N=21) — commercially available AI mathematics-education tools found, out of 28 identified tools Quotation withheld because this record is not cleared for redistribution. N=17 — commercialized tools classified as Pedagogical AI Tools Quotation withheld because this record is not cleared for redistribution. seven (N=7) — studies from China, the highest of any country in the corpus Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.29333/iejme/16006.