Digital EducationOutlook 2021（Ⅲ）

Frontiers of smart education technology： opportunities and challenges

Smart data and digital technology in education

Digitalisation opens new possibilities for education. However， most uses of innovative technology have been to conserve existing educational practice and sometime enrich it， but rarely transform it. Might digital technology， and， notably， smart technologies based on artificial intelligence， learning analytics， robotics， and others， transform education in the same ways they are transforming the rest of society？ If so， how might this look like？

Key opportunities

Smart technologies can improve education systems and education delivery in different ways. They can enhance access to education， improve its quality for learners， and enhance its cost-efficiency for societies.

Effectiveness

In the classroom， applications that directly support student learning show early promise. Personalised learning aims to provide all students with the appropriate curriculum or task， and scaffold them within a task， based on a diagnosis of their knowledge and knowledge gaps. This is not only done at the academic level， focusing on the “what”， but increasingly takes into account how students learn and factors such as self-regulation， motivation or effort.

A second promise of learning effectiveness comes from classroom analytics that support teachers in providing more effective teaching. Many applications already show how a variety of solutions could support teachers in better using their time in class， for example， by suggesting when it is a good time to shift to the next teaching or learning activity， who would require their attention the most， how they could engage the whole class in collaborative learning activities.

At the organisational and system levels， smart technologies also hold promise in making education more effective. While this remains relatively rare， smart technologies can be integrated in most dimensions of school activities， providing administrators， teachers and learners with feedback to manage school resources as well as improve the effectiveness of teaching and learning.

Equity

Smart technologies can help education systems provide more equitable learning opportunities. In this respect， smart technologies are more ambivalent. On the one hand， they clearly do or could help reduce inequity both by increasing access to learning opportunities for all and improving learning effectiveness for those who need it the most. On the other hand， without a widespread and equitable availability of smart technologies， inequity could also rise.

There are at least two reasons why technology may have a negative effect on equity. The first， obvious reason lies in the difference in access to devices and connectivity by students from different groups， notably students from lower socio-economic backgrounds. These students may not have the devices， the connectivity or the resources that allow accessing and using smart technologies either at the school they attend or at home. A second reason is that， if technology （e.g. personalised learning） works the same for everyone， those who start with stronger prior knowledge can maintain their advantage or even make faster progress than those with less prior knowledge. This would widen rather than reduce the achievement gap.

There are also many reasons to believe that smart technologies can advance the equity agenda.

First， learning technology can expand access to learning opportunities. Educational platforms proposing open educational resources or massive open online course （MOOC） platforms are good examples. They allow learners to access learning materials with a quality that may be superior to what they can access locally.

As importantly， smart technologies can reduce inequity by facilitating the inclusion of students with special needs and by adapting learning to different learning styles. Technology has， for example， made it much easier to support the diagnosis of learning difficulties such as dysgraphia， and remedial digital responses have also been developed.

Second， solutions such as early warning systems are entirely focused on reducing inequity by helping students at risk of dropping out from high school （or university） to graduate - students who drop out typically come from disadvantaged and minority backgrounds. Some use of learning analytics within institutions， for example， to monitor student engagement or redesign study programmes， could also have the same effects， should the educational institution pay particular attention to inequity.

Third， the use of learning analytics as exemplifed by personalisation at the individual level， be it using intelligent tutoring systems or learning analytics to keep students engaged in learning， all hold promise in reducing inequity， notably by supporting students with less prior knowledge to learn at the right pace.

譯文

智能教育技術(shù)前沿：機遇與挑戰(zhàn)

教育中的智能數(shù)據(jù)和數(shù)字技術(shù)

數(shù)字化為教育帶來了新的可能性。然而，創(chuàng)新技術(shù)的大多數(shù)用途是為了保護現(xiàn)有的教育實踐，或是豐富現(xiàn)有的教育實踐，很少有技術(shù)改變現(xiàn)有的教育實踐。數(shù)字技術(shù)，尤其是基于人工智能、學(xué)習(xí)分析、機器人和其他技術(shù)的智能技術(shù)，是否會像改變社會其他部分一樣改變教育？如果是這樣，將會是什么樣子？

關(guān)鍵機會

智能技術(shù)能以不同的方式改善教育系統(tǒng)和教育服務(wù)。它們可以增加接受教育的機會，提高學(xué)習(xí)者的教育質(zhì)量，提高社會的成本效益。

有效性

在課堂上，直接支持學(xué)生學(xué)習(xí)的應(yīng)用程序率先顯示出前景。個性化學(xué)習(xí)旨在為所有學(xué)生提供適當?shù)恼n程或任務(wù)，并根據(jù)對學(xué)生知識水平和知識差距的診斷，在任務(wù)中為他們提供支持。這不僅要在學(xué)術(shù)層面進行應(yīng)用，重點是應(yīng)用“什么”，而且要越來越多地考慮學(xué)生的學(xué)習(xí)方式以及自我調(diào)節(jié)、動機或努力等因素。

學(xué)習(xí)效率的第二個前景是課堂分析，提供更有效的教學(xué)以支持教師。許多應(yīng)用程序已經(jīng)表明，多種解決方案可以幫助教師更好地利用課堂時間，例如，給出轉(zhuǎn)移到下一個教學(xué)或?qū)W習(xí)活動好時機的建議，誰最需要他們的注意力，他們?nèi)绾巫屓鄥⑴c協(xié)作學(xué)習(xí)活動。

在組織和系統(tǒng)層面，智能技術(shù)也有望使教育更加有效。雖然這仍然相對少見，但智能技術(shù)可以整合學(xué)校活動的大多數(shù)方面，為管理者、教師和學(xué)習(xí)者提供反饋，管理學(xué)校資源，提高教學(xué)和學(xué)習(xí)的效率。

公平性

智能技術(shù)可以幫助教育系統(tǒng)提供更公平的學(xué)習(xí)機會。但在這方面，智能技術(shù)更加矛盾。一方面，它們增加了所有人獲得學(xué)習(xí)機會的途徑，提高了有強烈需求的人的學(xué)習(xí)效率，或多或少地減少了不平等。另一方面，如果智能技術(shù)得不到廣泛和公平的使用，不平等問題也可能加劇。

技術(shù)可能對公平性產(chǎn)生負面影響的原因至少有兩個。第一個明顯的原因在于不同群體的學(xué)生，尤其是社會經(jīng)濟背景較差的學(xué)生，在使用設(shè)備和網(wǎng)絡(luò)連接方面存在差異。這些學(xué)生可能沒有設(shè)備、網(wǎng)絡(luò)連接或資源，無法在他們就讀的學(xué)校或家中訪問和使用智能技術(shù)。第二個原因是，如果技術(shù)（如個性化學(xué)習(xí)）對每個人都適用，那些事先具備較強知識的人可以保持優(yōu)勢，甚至比那些具備較少知識的人進步更快。這將擴大而不是縮小成就差距。

但還有很多理由相信智能技術(shù)可以推進公平事項。

第一，學(xué)習(xí)技術(shù)能擴大獲得學(xué)習(xí)機會的途徑。開放教育資源的教育平臺或大規(guī)模開放在線課程（MOOC）平臺就是很好的例子。它們允許學(xué)習(xí)者訪問學(xué)習(xí)材料，其質(zhì)量可能優(yōu)于他們在當?shù)乜梢垣@得的資源。

同樣重要的是，智能技術(shù)能促進有特殊需要的學(xué)生的參與，學(xué)習(xí)適應(yīng)不同的學(xué)習(xí)風格，從而減少不平等。例如，技術(shù)使得診斷書寫困難等學(xué)習(xí)障礙變得更加容易，并且還開發(fā)了補救性數(shù)字響應(yīng)。

第二，早期預(yù)警系統(tǒng)等解決方案完全著眼于通過幫助高中（或大學(xué)）到研究生階段輟學(xué)的學(xué)生減少不平等現(xiàn)象（輟學(xué)學(xué)生通常來自弱勢群體或有少數(shù)民族背景）。如果教育機構(gòu)特別關(guān)注不平等問題，那么在機構(gòu)內(nèi)部使用學(xué)習(xí)分析技術(shù)來監(jiān)控學(xué)生參與度或重新設(shè)計學(xué)習(xí)計劃，也可能產(chǎn)生同樣的效果。

第三，學(xué)習(xí)分析的使用，如個人層面的個性化，無論是使用智能輔導(dǎo)系統(tǒng)還是學(xué)習(xí)分析來讓學(xué)生參與學(xué)習(xí)，都有助于減少不平等，特別是支持先前知識較少的學(xué)生以正確的速度學(xué)習(xí)。