About glossjar

glossjar is a word trainer for the most common Romanian words. You learn a few new ones each day, rate how well you knew each one, and it comes back around the time you'd start to forget it.

How it works

Each day brings 10 new words from the 5000+ most frequent ones (change the number in Settings), plus the older words that are due. After each card you press one of four buttons, and your choice sets the next review.

1Again

You didn't know it. It counts as a lapse.in ~1 minute

2Hard

You got there, but slowly. The next gap is shorter.a bit sooner

3Good

You knew it. The gap grows each time you get it right.in a few days

4Easy

You knew it at once. The gap jumps ahead.in weeks

Why it works

Without review, most of a new word is gone within a few days. Forgetting is fast at first and slows down later.

A review just before that point makes the memory stronger and the next drop slower, so the gaps keep growing: a minute, a day, a few days, weeks.

That's why ten minutes every day works better than an hour once a week.

Where the words come from

The word list is the most frequent Romanian words, counted and lemmatised by us from the Romanian news corpora of the Leipzig Corpora Collection (ron_news_2022_1M, ron_news_2024_1M) and the OpenSubtitles 2018 word list. Genders, plurals, forms and stress come from dexonline, the meanings were checked against English Wiktionary. Names, numbers and abbreviations are left out.

Leipzig Corpora Collection, Wortschatz Leipzig, University of Leipzig: CC BY 4.0. FrequencyWords (OpenSubtitles 2018): Hermit Dave, CC BY-SA 4.0. dexonline: GNU GPL. English Wiktionary via kaikki.org: CC BY-SA 4.0.

wortschatz.uni-leipzig.de · FrequencyWords · dexonline.ro · kaikki.org · CC BY 4.0 · CC BY-SA 4.0

D. Goldhahn, T. Eckart and U. Quasthoff: Building Large Monolingual Dictionaries at the Leipzig Corpora Collection: From 100 to 200 Languages. In: Proceedings of the 8th International Language Resources and Evaluation Conference (LREC 2012).

Recordings come from Lingua Libre speakers on Wikimedia Commons (CC BY-SA 4.0, CC BY 4.0 or CC0), credited on each card. Words without one are read by a synthetic, machine-generated voice, not a human recording: Google Cloud Text-to-Speech (Chirp 3 HD, Despina).

Lingua Libre · Wikimedia Commons · CC0 · Google Cloud Text-to-Speech