This is a lovely little Minna no Nihongo based Japanese vocabulary learning tool and a game.
This game is similar in some ways to anki and other apps. I've made this game with the use of AI. The input dataset was compiled using externally sourced dataset for the popular book Minna no nihongo's digital version of chapter wise words, verbs, adverbs, kanji, etc. The example usage of the words, in Hiragana and Kanji and the meaning of the sentence was made using AI to lookup for it across the web.
Here the content is mainly based on Minna no Nihongo. The user is free to use custom data input as long as it meets the input excel sheet format and does not break the game logic.
Input excel sheet format is also added here in the repo. The columns are:
- Serial Number
- Kanji
- Hiragana
- Romaji
- English Meaning
- Tags
- Chapter
- JLPT Level
- Example Sentence in Kanji
- Example Sentence in Hiragana/Katakana
- Example Sentence in Romaji
- Meaning of Example Sentence
- Remarks
- A learning mode
- A quiz mode
- Search for a word and learn more about it if its found in the dataset. This first shows a list of matches, then leads to learning mode for a selected word out of the matches.
Open the game html in a web browser, it will ask for an excel file to be uploaded to initialize the game. and set up the JLPT levels, Chapters, tags selection view and show the main game home page view. From this page user is supposed to configure the game. Choose which or how many questions to pick for the current round.
User can filter the dataset to work with for the quiz or the learning mode based on 3 filters with work to ADD questions to the current round using OR logic amongst them.
- JLPT Levels
- Chapters of the book (Minna no Nihongo)
- Tags (comma separated properties of a word. For example, verb, adverb, polite form, numbers, counting, days of the week. months of the year, etc)
Each of these show the current records belonging to that level or chapter or tag for user's convenience.
Chapters selection view can be collapsed for convenience. Chapter name can be alphanumeric. Tags selection view can be collapsed if needed. There are too many tags, so sometimes it helps to collapse them if the current focus is to play the game based on JLPT levels or a Chapter. Number of questions to play with right now can be selected using a scrubber style progress bar component or tap on the all button to select all possible questions from the dataset based on the current filters based. Default is to pick 10 questions if that meets the filter criteria.
Records are selected from the dataset based on the configuration. They become the input source for the Learning mode and Quiz mode. Search works on the entire dataset.
User can choose primary column to focus on: either Hiragana or Kanji or Meaning (in english). This is a radio button component in the configuration section. If user chooses Meaning as the primary focus, the secondary column by default is Hiragana, it can be toggled to Kanji using an on-off toggle button component.
Learning mode has a sub configuration. To work with the dataset in an increasing order, i.e. sequentially OR in a random order. User can decide this based on a toggle button selection.
Start learning and Start Quiz are two nice big call to action buttons to begin the respective mode of the game.
On starting the learning mode using a call to action button, the dataset is iterated in a wizard style view with previous and next button below. The focused word is shown highlighted and in larger font in the middle. The rest of the record vaules for this word as displayed neatly below. In the learning mode, 'Example sentence' is redundant text for the Kanji, Hiragana, romaji and meaning of the example sentence, so give it a decent heading and show the rest of the 4 related fields under it neatly. Show the JLPT level, chapter, tags and remarks also below if they exist.
The primary word is shown highlighted and in larger font in the middle. Secondary column word for the record is the correct answer. Out of the current question set valid as per the configuration, 3 more random options are generated and shown to the user. Immideate feedback is shown to the user based on selection, that the choice is right or wrong by highlighting the option. if correct is chosen, it can be a shade of green, and wrong option can be a shade of red. at the same time then highlighting the correct option to teach what was supposed to be picked.
Learning mode and quiz mode to show the progress and indicate how many questions left out of the current round's total.
Allow user to press back button to go to main configuration page from search view or the learning mode view. Allow user to quit the quiz. Allow user to restart the learning mode or the quiz mode. Allow exiting to main page or the configuration page also from learning mode or the quiz mode.
Search can be done with even a single character. User needs to tap the search button to initiate it (no auto search or debounce feature). Try to maintain focus in the search bar or screen update when searching. Search looks within the dataset's below columns:
- Hiragana
- Kanji
- Romaji
- Meaning
Highlight the match of the search keyword within the scrollable search results list then allow the user to tap on a result and switch the UI to learning mode for that selected word.
Use pastel colors for the buttons and navigation. Use an elegant font face like Noto Sans JP for the Japanese language characters. Use UI sans serif elsewhere.
The game has an html file you can download and open in your phone or pc browser. it needs an excel file which has all the words with hiragana, kanji, translations and other information.
Lastly, you can add your own content if you want...just open the excel file and add data to the end and give it a unique tag or chapter name to easily look for it.
Sorry if there are any mistakes in the content or game behaviour, please do let me know and I will try to quickly make the corrections. Feature requests are currently not being taken.