Knowledge Bomb 💣

A quiz game, a brain imagery generator, an ice breaker to galvanize undercurrent ideas, an alternative to Wikipedia "random article", available in 300+ languages 🌍🌏🌎 https://www.knb.wiki

---

The game content is data mined and analysed from Wikipedia articles and Wikidata items by multiple dimensions: languages clustering, hypernyms categorization, complexity scoring, etc.

Hypernym Hierarchy

All qualified Wikidata items / Wikipedia articles are categorized in 64 buckets.

The 64 top-level hypernyms of controlled vocabulary are presented for players to switch on/off to control the scopes of the next randomized quiz during game session by detonating 💣

Different views of the hypernym hierarchy:

• Tree path
• Tree path (monospaced node)
• Hilbert curve
• Hilbert curve (monospaced node)
  • Compact tiles

The design principle is: gameplay first, performance second, academic last; coarsely classifying as a decision tree, instead of networking as an ontologically precise semantic graph. The tree branches are grown and pruned to achieve balanced number of items/articles per node.

Knowledge Profile

For each game, a knowledge profile, or a brain imagery, will be generated to represent the player's mind for that quiz session. The more correct answer of the hypernym category, the more saturation of the cell color; the more wrong answer of the hypernym caterogy, the less saturation of the cell color. Example of knowledge profile :

2D language picker

The 🗺 2D language picker keeps the constant position of each language no matter choosing which language to display. It solves the problem of one dimentional language pickers (re)ordering certain languages alphabetically at various positions that hard to find, e.g. Deutsch - Aleman, Inaleman, Lialémani, Béésh bichʼahii, Ceruman, Duits, Eleman, Frangikos, German, Jerman, Krzyżacki, Mjymjecko, Немецкий, Német, Olmoni, Preisen, Rajchski, Saksa, 𐌸𐌹𐌿𐌳𐌹𐍃𐌺𐌰𐍂𐌰𐌶𐌳𐌰, Þýska, Tysk, Ubudage, Vācu, Weimarische, Yoeraman, Zėm, etc.

🗺 is also availble in game as a cheatsheet (hit top menubar ⌨ for hotkeys).

Design considerations

A lite version of the lexical distance map can be used as an convenient language picker for Google Translate:

List of Wikipedias - Languages ranking table

English is the first language in terms of information coverage and completeness. But which language is the second, and so on? Can we order languages in a quantitative manner?

The ranking is based on numbers of qualified Wikidata items per language. (wonder where's the swedish-cebuano?)

Ranking methodology

Rank	Language(Local)	Language	Wiki	WSM	Coverage%	Q60 🥇	Q50 🥈	Q40 🥉	Q30 🏅	Q20 🎖️	Articles	Solidness%	Speakers
1	English	English	en	9661.31	99.16%	5878	8817	13943	23112	41766	6896745	4.83%	1348000
2	Español	Spanish	es	9580.10	97.50%	5868	8770	13786	22630	39736	1984288	16.50%	586000
3	Français	French	fr	9566.99	97.23%	5864	8752	13730	22510	39587	2641672	12.36%	274000
4	Русский	Russian	ru	9539.45	96.67%	5851	8746	13749	22408	38749	2004567	16.19%	258000
5	Deutsch	German	de	9466.85	95.20%	5763	8577	13439	21913	38074	2951069	10.83%	135000
6	Українська	Ukrainian	uk	9464.87	95.16%	5871	8787	13770	22116	36206	1349145	23.68%	40000
7	中文	Chinese	zh	9441.68	94.70%	5869	8788	13722	21922	35746	1447000	21.97%	1120000
8	Português	Portuguese	pt	9424.59	94.36%	5789	8634	13492	21788	36365	1135120	27.91%	274000
9	Italiano	Italian	it	9402.41	93.91%	5744	8562	13334	21536	36703	1886839	16.71%	85000
10	日本語	Japanese	ja	9391.92	93.70%	5826	8688	13476	21451	35328	1433599	21.94%	128000
11	العربية	Arabic	ar	9373.57	93.34%	5877	8793	13596	21427	34137	1243026	25.21%	274000
12	Nederlands	Dutch	nl	9333.97	92.55%	5841	8664	13339	21137	33833	2169581	14.32%	30000
13	فارسی	Persian	fa	9326.38	92.40%	5872	8796	13564	21111	32892	1016108	30.53%	74000
14	Català	Catalan	ca	9317.07	92.22%	5838	8663	13372	20991	33361	761290	40.67%	10000
15	Polski	Polish	pl	9314.81	92.17%	5740	8517	13199	21034	34214	1631087	18.97%	45000
16	한국어	Korean	ko	9280.71	91.50%	5856	8701	13397	20813	32203	687727	44.67%	82000
17	Svenska	Swedish	sv	9196.05	89.84%	5729	8484	13022	20362	31313	2596249	11.62%	13000
18	Bahasa Indonesia	Indonesian	id	9070.28	87.40%	5758	8510	12826	19240	28464	707545	41.47%	199000
19	Čeština	Czech	cs	9068.74	87.37%	5672	8375	12777	19493	29087	554980	52.85%	14000
20	Türkçe	Turkish	tr	9041.95	86.85%	5729	8438	12759	19161	27986	622900	46.81%	88000
21	Suomi	Finnish	fi	9039.49	86.80%	5657	8349	12696	19221	28473	581857	50.09%	6000
22	עברית	Hebrew	he	8881.88	83.80%	5568	8131	12131	18061	26160	363836	77.33%	9000
23	Norsk (Bokmål)	Norwegian (Bokmål)	no	8847.43	83.15%	5564	8129	12161	17975	25479	636711	43.85%	5000
24	Српски / Srpski	Serbian	sr	8729.97	80.96%	5743	8294	12007	16587	21995	695799	39.07%	12000
25	Esperanto	Esperanto	eo	8685.88	80.14%	5655	8178	11764	16337	21828	359848	74.77%	180
322	see 297 more...

Qualified Wikidata items

A qualified item (examples) should be a general, universal, common concept or knowledge, which satisfies these criteria:

• With more than 20 languages versions of Wikipedias.
• Not an onomatology (anthroponymy, toponymy, hydronym, taxon etc.), which is not a particular person, place, biota, chemical, astronomical object, company, product, publication, song, team, sport match, event, year, decade, day, unicode, etc.
• Not a disambiguation page item.
• Can be hypernyms of items above (any level of parents of instance_of?/subclass_of*). Hypernyms are legitimate to have less than 20 languages versions, being small quantity of abstract philosophical terms backboning the semantic network.
• Qualified item/article should be verified through with cross check and proofreading for quality assurance.