Abstract: This writing piece was a group contribution, my group came up with an app. Our app will target high school and college students by adjusting to their unique learning style and teaching them.
Engineering Proposal: SmartClass
Saqib Rahman, Bhavesh Shah, Luiza Carvalho DeGodoy, Sandy Qiu
Professor Julianne Davidow
ENGL 201007 – Writing for Engineers
CUNY The City College of New York
Table Of Contents
Introduction………………………………………………………………………….3-6
Technical Description of SmartClass…………………………………………………..7-8
Logistics…………………………………………………………………………………………9-10
Other Innovations…………………………………………………………………………..11-13
Conclusion……………………………………………………………………………14
References…………………………………………………………………………15-17
Introduction
Background – Education
Education is the core to developing young adults into members of society. Students need to get an education to live “morally, creatively, and productively in a democratic society” (ASCD 2012). Currently, there are a few commonly used teaching methods. These include “storytelling, discussion, teaching, training, and directed research” (Worldeducationcom 2016). However, as seen repeatedly, these methods are not always effective. Every student has a different learning style, but it is very difficult for classroom teachers to cater to every student’s specific style. We believe that artificial intelligence can be used as a teaching tool to help every student achieve their maximum potential.
Background – Artificial Intelligence
Artificial intelligence refers to when machines or computer systems perform tasks that normally require human intelligence. It is categorized into 3 parts: computer vision, machine learning, and natural language processing. Computer vision “captures and analyzes visual information using a camera, analog-to-digital conversion and digital signal processing” (Rouse, 2010). Essentially, this technology can be used to process visuals to understand what they depicts. Computer vision is utilized in games, ecommerce, searching, and more. Natural language processing is the processing of human language by machines. With this technology, computers can translate human language into computer language in order to process and understand what the human is asking. The final form of artificial intelligence is machine learning. Machine learning allows computers to learn by themself without any human-intervention. For the scope of our application, we plan to mainly rely on machine learning in order to learn and adapt to every user’s particular learning style.
History
Going all the way back to the hunter-gatherers, education started with children learning “through self-directed play and exploration” (Gray, 2008). Overtime, as the industry became more automated, people believed children should physically be at a place where all their time would be dedicated to learning; hence, the first school in America was created in 1635. To further help students learn without relying on teachers alone, the first textbooks, made by William Holmes McGuffey in 1836, were introduced to schools around the country. With textbook, students were able to learn outside of the school and no longer needed to rely on teachers. However, learning was still difficult for a lot of students due to the lack of personal attention in classes over 30 students. To alleviate this struggle, more inventions were introduced in the 20th century. One prominent invention was the “1971 invention of electronic books, or eBooks” (Miller, 2011) by Michael Hurt. This product allowed students to access an electronically printed book rather than a printed book, and later added additional functionality such as taking notes, accessing the internet, and reading pdfs. Another invention made to help students learn on their own was the invention of the Apple I, the first computer made by Apple. Invented on “June 29, 1975” (Thompson 2016) by Steve Wozniak, this allowed students to play learning games, which helped in making learning more engaging and interactive. In the 21st century, technology started becoming even more advanced. One of the most used educational platforms designed for school use was the Blackboard, which was “founded in 1997 by two education advisors, Matthew Pittinsky and Michael Chasen, as a consulting firm to provide technical standards for online learning applications,” (Bradford, 2007). This was predominately used by college students to submit their homework, learn, and connect to their professors. Then we have YouTube. Created in 2005, this website is used by billions of people worldwide. It is a very useful tool for students to learn because of its video-sharing feature which allows for whoever to post any videos they’ll like, so students can access the educational videos and learn by themselves. With Youtube also came Khan Academy, an online educational platform invented in 2007, that provides short lessons in the form of videos. With a broad range of subjects, Khan Academy is one of the most used tool by students when they learn on their own. There are many tools still being developed for students to get the education they need outside of the traditional classroom.
Problem and Thesis
The primary problem with our current education system is that students don’t get personalized attention to meet their learning needs. Each student has a different learning style and different needs, which makes it difficult for teachers to accommodate their curriculum to each and every single one of them. The lack of a personalized curriculum is leading to many students falling behind. To counter this issue, we propose the inclusion of artificial intelligence in education. By creating an app that uses AI, it will be able to learn about a user’s learning style. Our AI algorithm can then curate a lesson plan specifically tailored to a user’s pace and learning style. Through our app, students will be able to seek the help they need in a faster and simpler manner. Ultimately, we hope that our app will revolutionize education to the point where it is something that students want, rather than something they are forced to just put up with.
SmartClass App Technical Description
- User Authentication
When a user first opens our app, he would be prompted to either log into his account or create an account. If a user chooses to create an account, they will be prompted with a pop-up. This pop-up would ask them to authorize us to access their Apple Pay account, Each month, we will deduct a certain fee from their Apple Pay account as our monthly charge for using this application. Users can chose to cancel their plan at any time.
- Subject Database
The subject database offers a wide range of subjects that users can choose, from math classes to even coding classes. Users are allowed to enroll in as many subjects as they please. Each subject will have a different time span (some may last weeks, while others last months), depending on the difficulty of the class. One of our missions is to help students really master their material. Thus, we may bring back topics from pre-requisite classes if a user does not seem comfortable with a prerequisite topic. For example, if a student is working on an algebra course, but is struggling to add fractions, our app will require the user to complete a few exercises on adding fractions before continuing his algebra course.
- Test
In order to understand a student’s specific learning style, we need them to take a quick test before they begin their course. The first 10 questions will be related to the subject they are pursuing. These questions will help us understand what level they are at currently so that they can avoid re-learning something they already know. The next 10 questions will be more general and would try to get to know the user better. This batch of questions should help us figure out their learning style, whether it be visual, example-based, etc. Our machine learning algorithm would utilize the responses from these 20 questions to generate the perfect curriculum for the student.
- Personalized Course
Once the user completes his test, he can now begin the curriculum which is specifically tailored to his learning style. Each week, students will learn throughout the week, and will have to complete a practice test on Sundays in order to move on to the next chapter. We are characterizing our users into 4 learning styles: visual, example-based, text-based, and auditory learning.
Logistics
Process
In order to develop this app, our team will first brainstorm and decide the basic features and functionality we want our app to offer. Once we have finalized our plans, we will begin designing our app. Our primary design tool will be Sketch, on which we can thoroughly design the layout of each screen in the app. Following this, we still start coding our app in XCode (for iOS devices) and Android Studio Code (for Android devices). In order to host all of our courses and user data, we will connect our app to Firebase, a cloud-based database made by Google. Once our app has been built, we will start beta testing it with users in order to get feedback quickly. We will then incorporate any suggestions into our app. Once we are confident that our app is complete, we will then launch it to the App Store and Google Play Store.
Materials
We will need Sketch, XCode, Android Studio Code, and Firebase for this application.
Labor
To build this app we will hire a group of software developers to help the process of programing this app smoother and reliable. In addition, we will hire a group of test subjects to test out our app before it reaches the market to see if students feedback on it. The test subjects will range from high school students to college students. To prevent potential bias, our test subjects will range from honors students to students with a 504 plan. These students’ feedback will help us refine our app so that we can release the best version possible to the market.
Cost to Build and Revenue Projections
We estimate that SmartClass will cost roughly $100000 – $110000 to create. For the product development, we will be hiring a team of 3 software engineers. We will be paying them around $30000 a year. In addition to that, it will cost us around $500 a year to keep the app on the App Store and Google Play Store and another $500 to host our data on Firebase. We also want to invest $10000 into our marketing efforts in order to help us build our user base. Lastly, we expect another $10000 to go into hiring beta users to test our app and give us feedback. Even though the cost of launching this app is quite expensive, we predict that we will be able to recover this money within a year. Like a lot other education apps, we will give user a one-week free trial with limited features to give them a test run of the app. After that one-week period, we will be charging $30 for 2 courses during one semester, $60 for 4 courses during one semester and $100 for 5+ courses during one semester. Additionally, we will also be selling our software as a bulk package to schools and universities as another source of extra income.
Other Innovations
Thirdspace Learning
Our Artificial Intelligence online learning application, SmartClass would implement machine learning and create a personalized way of learning for each individual student. It would track the strengths and weaknesses of the student and notify the teacher, in a similar manner to the online math learning platform called Thirdspace Learning. According to Eddington in Thirdspace Learning, “If the student misunderstands the task or if the teacher misses something important, Thirdspace wants the artificial intelligence to identify the problems and warn the teacher before the problem begins to aggravate” (Eddington, 2018). Like Thirdspace, SmartClass would also show the progress of each individual student and the task’s he or she is struggling with. Although our app would identify the strengths and weaknesses from the beginning and continue to monitor those specific weaknesses. For example, SmartClass would first assign a “Knowledge Check” to see, which areas a student would need the most assistance. It would assign that specific topic to the student after the teacher has taught it in the classroom. If the student continues to have problems in that certain task, then it would notify the teacher to assist the student individually. Additionally, even after learning the topic, a spaced repetition algorithm in our app would assign two questions from that topic approximately a month later as a review in order to optimize retention. Therefore, this would not only give the student confidence, but the teacher as well.
Thinkster Math
The training platform called Thinkster Math is another mathematical online learning system, which combines AI and machine learning to help teachers monitor the performance of a student. According to Sennaar, “The tutoring platform records, student work and tracks the steps a student chose to solve a math problem. As a result, math coaches can identify problem areas and inform strategies for improving student performance” (Sennaar, 2019). Through this method, Thinkster Math is able to efficiently gather each student’s performance data and create a learning experience based on it. SmartClass would also be able to identify the thought process and learning style of each individual student. Unlike Thinkster Math, SmartClass would assist students with more than one subject, which will also take into consideration how each student learns differently depending on the subject. For example, if a student began doing tasks related to learning a new language, our app would assign three different types of questions. The first would be a visual question with a picture. The second would be a voice recording of a short conversation in that specific language. The third would require the student to make some form of movement to solve the question, such as taking notes or drawing a picture. After assessing whether the student has a visual, auditory, or a kinesthetic learning style, SmartClass would then assign questions based on the data gathered to not only help the student learn quicker but also demonstrate to the teacher how the majority of the students prefer learning.
Differ (EdTech Foundry)
SmartClass would also include a chatbot, which would interact with each student and assist them when answering questions. For example, Sennaar stated that the company EdTech Foundry, “released Differ, a chatbot designed to assist students in higher education. The system is designed to provide quick answers to student questions that tend to recur each semester…the chatbot occasionally suggests academic articles and further reading to students that pertain to their coursework. The chatbot also suggests direct ways that a student can contribute to their class, such as posting in forums” (Sennaar, 2019). The chatbot is able to act as a teaching assistant and increase student engagement by making useful recommendations. Furthermore, the app Duolingo also uses a chatbot for learning to speak fluently in a certain language. For instance, Eddington stated that the app is “capable of “receiving and responding to a few thousands of unique answers” instead of a more restricted amount of hints. Advantages include the “help me reply” function, which offers users possible answers if the person finds it difficult to answer. Unfamiliar and new words can also be translated or checked its pronunciation in real time during chats” (Eddington, 2018) . SmartClass would also include a chatbot that interacts with the student, regardless of any subject. For example, if a student were given a math question, but was unsure how to solve it, a hint would be given and the student can also see an example of a question similar to what they learned in class to help solve the current problem. A video would also be included at the end of each question. Additionally, if a student were to practice English and type in a sentence, then our chatbot would immediately recognize the various ways the sentence can be written and also suggest an ideal version of the sentence regardless of whether the student gets the question right or wrong.
Conclusion
There are many different learning styles, mainly VARK (visual, aural, read/write, kinesthetic) that each student has. It is difficult for teachers to accommodate to all students preference of learning however. We are proposing an app, infused with artificial intelligence, that will learn the user’s prefered method of learning and develop a lesson plan based off of that style. Though there are other innovations that are similar to our idea, our app will offer additional functionality that other apps don’t offer. For one, our app will be highly adaptable. In other words, our A.I. algorithm will monitor students’ learning plans and will make changes to their plans if needed. Additionally, we will allow students to enroll in multiple subjects on our platform instead of just one, and will also incorporate a chatbot for those who need help solving a problem or has questions. The cost to build this app ranges from $100000 – $110000. Even though it may be expensive to build and use, in the long-term, the effects it will have on the students will be greatly beneficial. Incorporating artificial intelligence into education will hopefully revolutionize the education system.
References
Dali Luo1, daliluodlledu89@21cn. co. (2018). Guide Teaching System Based on Artificial Intelligence. International Journal of Emerging Technologies in Learning, 13(8), 90–102. https://doi-org.ccny-proxy1.libr.ccny.cuny.edu/10.3991/ijet.v13i08.9058
Eddington, R. D., Siva, Tatt, S., & Goodman, S. M. (2018, November 10). Top 5 Artificial
Intelligence tutors in education. Retrieved April 24, 2019, from
Gray, A. (2008, August 20). A Brief History of Education. Retrieved from https://www.psychologytoday.com/us/blog/freedom-learn/200808/brief-history-education
Miller, M. (2011, September 08). Project Gutenberg founder and inventor of the ebook, Michael Hart, dies. Retrieved from https://www.zdnet.com/article/project-gutenberg-founder-and-inventor-of-the-ebook-michael-hart-dies/
Nadel, B. (2018). AI ACCELERATES: Machine learning will offer teachers deeper insight, but not replace them. District Administration, 54(1), 55–58. Retrieved from https://search-ebscohost-com.ccny-proxy1.libr.ccny.cuny.edu/login.aspx?direct=true&db=eue&AN=127021812&site=ehost-live
Popenici, S. stefan. popenici@cdu. edu. a., & Kerr, S. sker2923@uni. sydney. edu. a. (2017). Exploring the impact of artificial intelligence on teaching and learning in higher education. Research & Practice in Technology Enhanced Learning, 12(1), 1–13. https://doi-org.ccny-proxy1.libr.ccny.cuny.edu/10.1186/s41039-017-0062-8
Rouse, M. (2010). What is AI (artificial intelligence)? – Definition from WhatIs.com. Retrieved from https://searchenterpriseai.techtarget.com/definition/AI-Artificial-Intelligence
Sennaar, K. (2019, February 03). The Artificial Intelligence Tutor – The Current Possibilities of
Smart Virtual Learning. Retrieved April 24, 2019, from
https://emerj.com/ai-sector-overviews/artificial-intelligence-tutor-current-possibilities-sm
Sloan, W. M. (2012, July). What Is the Purpose of Education? Retrieved from http://www.ascd.org/publications/newsletters/education-update/jul12/vol54/num07/What-Is-the-Purpose-of-Education¢.aspx
Thompson, C. (2016, March 01). Steve Wozniak’s Apple I Booted Up a Tech Revolution. Retrieved from https://www.smithsonianmag.com/smithsonian-institution/steve-wozniaks-apple-i-booted-up-tech-revolution-180958112/
Worldeducationcom. (2016, May 12). Education is the process of facilitating learning, or the acquisition of knowledge, skills, values, beliefs, and habits… Retrieved from https://worldeducationcom.wordpress.com/2016/05/12/education-is-the-process-of-facilitating-learning-or-the-acquisition-of-knowledge-skills-values-beliefs-and-habits-educational-methods-include-storytelling-discussion-teaching-training-and-d-2/
Let’s Build Your App. (n.d.). Retrieved from https://www.clavax.com/howmuchtomakeanapp/lets-build-your-app
THE BLACKBOARD LEARNING SYSTEM. (2007). 35th ed. [ebook] Albany: The Journal of Educational Technology Systems, pp.301-314. Available at: http://uupinfo.org/research/working/bradford.pdf [Accessed 13 May 2019].
