Memora: An application for learning information by spaced repetition

Annotation:

The purpose of this work is to develop a mobile application for learning information by spaced repetition, which allows users to create decks of cards containing text and images, import and export them, set up repetition times, and receive push notifications to optimize the memorization process. The application implements an interval repetition algorithm to improve learning efficiency.

stack:

• Android Studio
• Kotlin
• Jetpack Compose

TO-DO:

• class hierarchy
• main menu screen
• deck creation button
• create deck
• deck display after clicking
• card creation: text, picture, difficulty
• display the card in the deck
• view card when clicked
• mark as viewed
• popup menu when you press three dots view edit info delete
• customize push notifications
• implement 2 interval counting algorithms for two difficulties: difficult: frequent and fast. easy: infrequent and slow.

dev log:

03.01.25

init the project, made git branches

04.01.25

1. created design prototype: https://www.figma.com/design/BUK3l9sfjKn4nrAt1FZQty/Memora?node-id=601-10&p=f
2. using Jetpack Compose created top bar and Add Button

02.03.2025

let's start from the beginning

described our models with data classes

19.04.25

1. refactored the whole project to match oop coursework requirements (all files are single classes)
2. moved some components to separate classes for being reusable (graphical, )

TO-DO 0. ci/cd to build artifacts (apk)

1. make deck preview better (3-dot button to rename and delete, preview picture) (DeckItem in DeckListScreen) DeckListViewModel????????
2. unify creation dialog and make it fancier class CreateDialog() : Component
3. learning screen with preview, it will be similar to editing but we cant edit and there would be only 2 buttons (done and exit), LearningScreen
4. make the unified editng and creation screen where we can add 4.5. content as Content, not string only ` plan on this
5. adaptive algorithm.. supermemo?
6. offline push-notifications
7. imports mesh (refactoring?)
8. about screen.
9. settings screen where we can choose our intervals for fixed algorithm
10. JsonProcessor for exporting one deck, all decks, importing deck from file (JSON and .memoradeck (Bytestream))
11. Foreground Service
12. documentation

25.05.25

1. now we can rename and delete deck, 3dots, deck name preveiww looks nicer

2. card item look way better, learning screen(s), adaptive algorithm (based on how hard was it)

Changed the review intervals according to difficulty: For the first review (reviewCount = 1): Difficult (1): 1 hour Medium (3): 6 hours Easy (5): 1 day For second repetition (reviewCount = 2): Difficult (1): 6 hours Medium (3): 1 day Easy (5): 3 days For subsequent repetitions: Difficult: interval * (ease - penalty) Medium: interval * ease Easy: interval * (ease + bonus) Improved handling of learning quality: For difficult learning (quality < 3): Significantly reduce ease (penalty * 2) Reset interval to 6 hours When medium (quality = 3): Slightly decrease ease (penalty * 0.5) At easy (quality > 3): Increase ease in proportion to quality Now cards that the user marks as difficult will appear much more frequently for repetition, and easy cards will appear less frequently. This should make the learning process more efficient and adapted to individual memorization.