Our app is based around the user putting in less, but still getting the same, if not better quizzes and flashcards as our competitors. For a user to accomplish this, they first have to know what they want to do. For Flashcards, all they have to do is take a picture of a study guide with the word you want, and then the definition. There will also be an option where your word can be plugged into the google search engine, where the user can then accept or reject the definition we give to them. The process of the text on the picture being transferred to editable text is called OCR. It is defined below. If a user wishes to get a read aloud of a textbook, then they just snap a picture of the desired pages, or paragraphs to be read aloud, and they will immediately have a read aloud at their fingertips. OCR will be the program we use to do this. As for the quizzes, they will be short answer, and the way the questions will generates varies, based upon what subject the user is studying. There are two ways to form questions: 1) If your study guides or worksheets already have questions ( In subjects like Math or Language Arts ), the computer will simply copy them onto your phone, and display them ( still using OCR ). Then, you will be able to answer them. 2) The second way is mainly geared toward Social Studies, and will be generated from a textbook. Since most of social studies is based around people, and dates, we will use an AI to locate these, and simply form a question such as "What happens in 1967?", or "Who is Mark Twain?", etc. A definition of AI, and how we will train it can be found below the OCR section.
OCR is Optical Character Recognition. This means that it will take a picture of handwritten or printed characters, and it will output an editable document. So, how exactly does this work? Here are the basics. OCR is a software. It first scans each letter from an image one by one. Next, it searches its library of fonts and font sizes to find what best matches the text. Once it has found a credible font, it matches characters, including the alphabet punctuation, and numbers. Think of it this way. Before OCR, a computer will see a PDF as just black symbols on a white background. However, after OCR, the computer will actually know what text been displayed. At this stage, you will have an editable document displaying your text. But OCR can still do more . This includes the fact that it is capable of turning its generated text into speech. You can see an example of Google Keep's OCR program below.
AI, or artificial intelligence, is the basis of how are quizzes will be made. To understand how an AI could make a block of text, into questions, you must first understand how the human brain does. When we ask a question, it is almost always relevant to the topic at hand, but we usually do not know the answer. So, for a second imagine that you are an AI. What would you be looking for to form these questions? For social studies, maybe any dates, like 1967, or any indicators like AD or BC. With these dates, the AI simply asks, what happened in 1967? As for the training of the AI to find dates, we would just run all number combinations through it, which should be easy with basic coding knowledge Some other good basis' for questions would be names. We would run names through the AI, so they would get familiar with finding them. Questions could also be created if the user has a study guide with questions already on it. In that case, the AI wouldn't have to do anything, the computer would only have to display the questions.