.png)
An Analysis and Projection of AI in Education
NLP in Education
Industry Analysis
Natural Language Processing has been used for the past three decades in education with early models of computer-assisted linguistics systems (Dataversity, 2022). These early models of NLP provided feedback at the word and sentence level versus at the thematic level of the writing.
​
NLP is now used in many different ways in education. Popular applications include intelligent tutoring systems, plagiarism detection tools, grammar and writing feedback tools, translation and language learning apps, essay and test grading systems, and educational chatbots (Chowdhury & Chowdhury, 2019). It has been especially effective in personalizing educational instruction and feedback for students (Rezvin, 2023). Students can receive real-time feedback on their work and adjustments on their lessons based on their level of understanding (Pratim Ray, 2023).
​
With the advent of large language models like GPT from OpenAI, NLP in education is now omnipresent. Education software companies can use GPT's application programming interface (API) to integrate the large language model and format it for their applications (OpenAI, 2022). Some uses of large language model in education include (Holt, 2019):
​
-
Generating writing prompts and questions to help with reading comprehension
-
Providing personalized summaries of or paraphrasing texts based on a student's skill level
-
Generating practice problems and quizzes in all subjects
-
Producing solutions and step-by-step problem-solving guides
-
Grading student written work
-
Providing written feedback on student essays or tests
​
​
Industry Trends
There are many burgeoning trends in AI in education, and, more specifically, in the NLP in education space. It is important to note that NLP has been used widely in various learning and data analytics softwares in education for years. A major shift in knowledge about and access to NLP in education came with the public rollout of the NLP chatbot, ChatGPT, in November 2022 and subsequent chatbots and NLP tutors that students and teachers could use directly (OpenAI, 2022). This created a pivotal shift for NLP in education, as it made it accessible to both students and teachers, thus changing the way they view, understand, and use artificial intelligence.
​
Prominent NLP tools in education include intelligent tutoring systems, automated essay scorers, adaptive learning software, and educational chatbots. Intelligent tutoring software is used to provide personalized support for students across different disciplines by processing their input and generating support based on their skill level. Automated essay scorers use NLP to analyze coded trends and grammar structures in student writing and provide them with timely feedback. Adaptive learning software offers real-time educational experiences that are tailored to the students' skill level and their achievements of certain tasks. Educational chatbots, or general chatbots that are used for educational purposes, have gained a lot of recent traction due to their accessibility, ease of use, and fast generation of content.
​
According to a market research report published on Globe Newswire, the NLP segment of AI in education is expected to be the largest segment in the global market between 2023 and 2028 (GlobeNewswire, 2023). The adoption of generative AI tools has already been expansive across many different companies, and will only continue to grow. Chatbots will undoubtedly be leaders in the NLP in education space, as they have already proven to be versatile and have a strong value proposition in many different education disciplines and platforms. The next section profiles three key players that are already integrating developed chatbots (using OpenAI's GPT model) into their applications.
​
​
Major Players
Duolingo, Khan Academy, and Quizlet have all been well-established in the edtech space for over a decade. They swiftly adopted ChatGPT's API to incorporate generative AI into their models in order to maintain their relevance and value proposition as top learning models in edtech (Rosenberger, 2023).
​
Duolingo, a popular language learning app for all ages, created Duolingo Max using GPT-4 which has two AI-powered features (Duolingo, 2023). The features are called Explain My Answer and Roleplay. With Explain My Answer, users can engage in a conversational manner with a chatbot to help understand their answer to a question, whether they were right or wrong (Duolingo, 2023). This helps deepen the learning experience through personalization and answer explanation. The Roleplay feature allows users to practice real-world conversation skills with AI characters in the app (Duolingo, 2023). These are prime examples of how NLP can be used to create more holistic language learning experiences that are tailored specifically to the user. Duolingo is capitalizing on the NLP in education trend and leveraging their large user base to make advances in online language learning, making them a major player in the NLP in education industry.
Source
Major Players Cont'd.
Khan Academy, a nonprofit education organization with online tools, resources, and classes, also developed their own GPT-powered AI chatbot, Khanmigo, that serves as a tutor for students and a personal assistant for teachers (Khan Academy, 2023). Students can ask Khanmigo questions about specific problems and receive prompts that guide them to the answer without giving them the answer. This allows them to ask questions without reservation and get on-demand guidance while working on assignments. Teachers can use Khanmigo to help structure lesson plans and come up with activities (Khan Academy, 2023). What is noteworthy about Khanmigo is that it works across all subjects, so students can use it to help write an essay, understand a historical event, solve a math problem, or study for a science test.
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
Quizlet, an online tool for studying and learning, created a similar AI solution as Duolingo and Khan Academy. Theirs, called Q-Chat, is a GPT-powered chatbot that essentially allows users to have a study-buddy to ask questions to and practice with when studying (Quizlet, 2023). Q-Chat uses the socratic method of teaching to prompt students to use critical thinking and work through different strategies to answer their own questions (Quizlet, 2023). This is an important facet of Q-Chat; promoting critical thinking in the age of generative AI is at the forefront of many conversations in education.
​
​
​
​
​
All of these new AI tools use a subscription model and are currently in their beta phase; however, they have all proven to be pivotal in the space of personalized learning and using AI to create new learning environments for students that help them learn more efficiently, effectively, and deeply.
​
​
​
​
Source
Source
Industry Projections
What is NLP in education going to look like in 5 years? What about in 10?
​
According to a comprehensive report from Markets and Markets research, the global NLP in education market is projected to grow from $99 million in 2023 to $228 million in 2028 with a Compound Annual Growth Rate (CAGR) of 18.1% (2023). The overall AI in education market is projected to grow from $4.25 Billion in 2023 to $20.65 Billion in 2028 at a CAGR of 45.9% from 2023 to 2028 (GlobalNewswire, 2023). This is a large increase in market value in just five years, and, according to Global Newswire's report on AI in education, NLP in education will be a leader in growth, adoption, and revenue generation for the AI in education market (2023).
We are already seeing a rapid adoption of NLP in education in order to remain a competitive and desirable learning tool for students and teachers. Although AI has been used in education tools and for data analytics in education for multiple decades, the GPT model transformed how much educational experiences can be personalized to meet specific student needs. Additionally, chatbots offer a breakthrough for AI in education, as students and teachers are able to interact directly with AI and understand what AI is. Many edtech companies that are using the chatbot model are currently in their beta phase and already have a strong user base; at this rate, they will likely continue to grow over the next 5 years as the NLP models become more advanced.
​
Personalized, large language model chatbots in education are here to stay. By the turn of the decade, we will likely see most online education tools using some form of chatbot to assist with student learning. They will also continue to make advancements in how they interact with students, as more leaders in the space advocate for using AI to promote critical thinking in students rather than hinder it by supplying faster answers and generating content. This is an important consideration, as the current argument against the widespread dissemination of AI in education is the fear that it will supplement any real learning because students will be able to search for answers within seconds versus perusing search results for hours.
​
The industry will also grow in the foreseeable future as the market for teaching AI gains more traction. As more teachers, students, and professionals use AI in their everyday lives, the demand for learning about AI and creating AI will grow.