Content translation/Deletion statistics comparison

Data on the deletion ratio of Wikipedia articles created with and without Content Translation.

Wikis with higher deletion ratios for CX created articles

We reviewed wikis where the deletion rate of articles created with content translation is higher than the deletion rate for articles created with other tools as part of T286636 during a specified timeframe. This data is updated quarterly (every three months) to assess the evolution of deletion rates as improvements are made. This timeframe was selected to capture a sufficient time for editors to review content and limit seasonality effects.

Data comes from mediawiki_history and reflects the deletion ratios of main namespace articles that were created using Content Translation compared to the deletion ratios of main namespace articles created without the tool. Bots were excluded. We also removed wikis where 15 or fewer articles were created with content translation during the reviewed timeframe to reduce noise in the data and focus on wikis with more representative data.

Summary

Time period	Overall CX articles deletion ratio 1	Overall non-CX articles deletion ratio 2	Number of wikis with higher cx article deletion ratios3	Wiki with the highest deletion ratio difference	Per Wiki deletion ratio results
January 2023- March 2023	4.0%	5.8%	26 wikis	Tibetan Wikipedia (88.2% cx; 6.12% non-cx)	Q3 2023 wiki results
October 2022 - December 2022	3.8%	5.4%	32 wikis	Cebuano Wikipedia (61.1% cx; 18.5% non-cx)	Q2 2022 wiki results
July 2022 - September 2022	3.2%	4.3%	25 wikis	Western Frisian Wikipedia (100% cx, 3.24% non-cx)	Q1 2022 wiki results
April 2022 - June 2022	3.3%	4.7%	17 wikis	Javanese Wikipedia (69.6% cx; 0.35% non-cx)	Q4 2022 wiki results
January 2022 - March 2022	3.1%	5.0%	13 wikis	Estonian Wikipedia (29.4% cx; 11.7% non-cx)	Q3 2022 wiki results
October 2021 - December 2021	2.9%	4.3%	16 wikis	Lithuanian Wikipedia (47.4% cx; 15.3% non-cx)	Q2 2021 wiki results
July 2021 - September 2021	2.4%	3.8%	13 wikis	Javanese Wikipedia (96.2% cx; 12% non-cx)	Q1 2021 wiki results
April 2021 - June 2021	3.9%	5.4%	15 wikis	Moroccan Arabic Wikipedia (36% cx; 1.64% non-cx)	Q4 2021 wiki results
January 2021 - March 2021	3.4%	8.9%	13 wikis	Hawaiian Wikipedia (39% cx; 1.2% non-cx)	Q3 2021 wiki results
1 Defined as the ratio of all deleted articles created with CX to all articles created with CX across all wikis. 2 Defined as the ratio of all deleted articles created without CX to all articles created without CX across all wikis. 3 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period.

Quarterly Results

January 2023 through March 2023 (Q3 2023)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: January 2023 through March 2023 (Q3)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
bowiki	17	98	15	6	88.24%	6.12%	−82.11%
iuwiki	202	44	202	20	100.00%	45.45%	−54.55%
htwiki2	84	91	35	3	41.67%	3.30%	−38.37%
gdwiki	18	24	9	3	50.00%	12.50%	−37.50%
jvwiki2	83	295	32	15	38.55%	5.08%	−33.47%
lawiki2	17	642	5	40	29.41%	6.23%	−23.18%
crwiki	36	24	36	19	100.00%	79.17%	−20.83%
yiwiki	31	68	16	21	51.61%	30.88%	−20.73%
iswiki2	25	1451	5	50	20.00%	3.45%	−16.55%
kuwiki2	59	1049	11	39	18.64%	3.72%	−14.93%
ltwiki2	49	5816	9	567	18.37%	9.75%	−8.62%
tumwiki	22	2103	2	34	9.09%	1.62%	−7.47%
bswiki2	55	1163	9	111	16.36%	9.54%	−6.82%
arywiki2	41	408	8	52	19.51%	12.75%	−6.77%
zh_yuewiki2	28	2471	3	104	10.71%	4.21%	−6.51%
aswiki	45	816	5	41	11.11%	5.02%	−6.09%
mznwiki	47	594	4	21	8.51%	3.54%	−4.98%
lnwiki	16	77	4	16	25.00%	20.78%	−4.22%
afwiki2	65	1236	6	63	9.23%	5.10%	−4.13%
hywiki2	286	3437	31	243	10.84%	7.07%	−3.77%
etwiki2	105	4759	12	367	11.43%	7.71%	−3.72%
ttwiki2	32	574	2	22	6.25%	3.83%	−2.42%
mywiki	86	2620	4	99	4.65%	3.78%	−0.87%
enwiki	545	161216	35	9568	6.42%	5.93%	−0.49%
kswiki	46	249	1	5	2.17%	2.01%	−0.17%
swwiki2	70	598	9	76	12.86%	12.71%	−0.15%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

October 2022 through December 2022 (Q2 2022)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: October 2022 through December 2022 (Q2)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
cebwiki	18	281	11	52	61.11%	18.51%	−42.61%
lawiki2	17	764	6	43	35.29%	5.63%	−29.67%
arzwiki2	54	13791	16	238	29.63%	1.73%	−27.90%
hywiki2	38	1491	14	179	36.84%	12.01%	−24.84%
swwiki	45	1180	13	53	28.89%	4.49%	−24.40%
ltwiki2	36	2801	14	492	38.89%	17.57%	−21.32%
ttwiki2	32	68978	9	5436	28.12%	7.88%	−20.24%
iswiki2	20	1128	4	76	20.00%	6.74%	−13.26%
astwiki2	102	1420	15	56	14.71%	3.94%	−10.76%
idwiki	296	14138	80	2348	27.03%	16.61%	−10.42%
arywiki2	39	610	9	84	23.08%	13.77%	−9.31%
fiwiki2	42	8888	7	659	16.67%	7.41%	−9.25%
etwiki	54	4603	8	342	14.81%	7.43%	−7.38%
twwiki	47	125	4	2	8.51%	1.60%	−6.91%
orwiki	91	336	8	7	8.79%	2.08%	−6.71%
skwiki2	88	1817	18	258	20.45%	14.20%	−6.26%
tgwiki2	232	1743	21	63	9.05%	3.61%	−5.44%
afwiki	118	2121	9	77	7.63%	3.63%	−4.00%
htwiki2	26	192	2	8	7.69%	4.17%	−3.53%
zh_yuewiki2	63	2678	4	87	6.35%	3.25%	−3.10%
dewiki	302	52333	48	6711	15.89%	12.82%	−3.07%
eowiki	248	6424	10	70	4.03%	1.09%	−2.94%
brwiki2	90	2464	3	17	3.33%	0.69%	−2.64%
kuwiki2	312	1728	19	63	6.09%	3.65%	−2.44%
jvwiki2	30	930	1	17	3.33%	1.83%	−1.51%
bswiki	90	680	11	74	12.22%	10.88%	−1.34%
glwiki	602	3222	29	114	4.82%	3.54%	−1.28%
simplewiki	103	12489	28	3272	27.18%	26.20%	−0.99%
ckbwiki2	343	1145	17	49	4.96%	4.28%	−0.68%
lvwiki2	45	3406	4	285	8.89%	8.37%	−0.52%
zhwiki	1162	37882	79	2488	6.80%	6.57%	−0.23%
ocwiki	49	984	1	20	2.04%	2.03%	−0.01%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

July 2022 through September 2022 (Q1 2022)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: July 2022 through September 2022 (Q1)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
fywiki	23	1110	23	36	100.00%	3.24%	−96.76%
arzwiki2	19	16331	9	198	47.37%	1.21%	−46.16%
lawiki2	32	611	17	45	53.12%	7.36%	−45.76%
iswiki	27	1113	8	41	29.63%	3.68%	−25.95%
arywiki	41	621	12	25	29.27%	4.03%	−25.24%
jvwiki2	18	2656	3	8	16.67%	0.30%	−16.37%
astwiki2	18	1030	4	64	22.22%	6.21%	−16.01%
ltwiki	40	3546	8	290	20.00%	8.18%	−11.82%
lvwiki	31	3743	4	156	12.90%	4.17%	−8.74%
htwiki2	21	421	3	27	14.29%	6.41%	−7.87%
aswiki	48	526	8	47	16.67%	8.94%	−7.73%
fiwiki	90	8700	11	685	12.22%	7.87%	−4.35%
tgwiki2	203	836	17	37	8.37%	4.43%	−3.95%
thwiki	20	3994	2	250	10.00%	6.26%	−3.74%
dawiki	96	5689	15	719	15.62%	12.64%	−2.99%
ckbwiki2	253	1845	19	88	7.51%	4.77%	−2.74%
brwiki	30	2562	1	19	3.33%	0.74%	−2.59%
kuwiki2	105	3469	4	67	3.81%	1.93%	−1.88%
ttwiki	46	1785	1	12	2.17%	0.67%	−1.50%
cywiki	69	143028	1	158	1.45%	0.11%	−1.34%
hywiki	137	1530	19	194	13.87%	12.68%	−1.19%
cswiki	400	8837	33	672	8.25%	7.60%	−0.65%
skwiki	59	2003	7	231	11.86%	11.53%	−0.33%
zh_yuewiki	39	5413	1	122	2.56%	2.25%	−0.31%
nlwiki	291	15380	31	1623	10.65%	10.55%	−0.10%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

April 2022 through June 2022 (Q4 2022)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: April 2022 through June 2022 (Q4)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
jvwiki	23	2283	16	8	69.57%	0.35%	−69.21%
mnwiki2	21	553	17	283	80.95%	51.18%	−29.78%
astwiki2	16	880	5	70	31.25%	7.95%	−23.30%
etwiki2	23	3689	6	400	26.09%	10.84%	−15.24%
kuwiki	75	3276	11	49	14.67%	1.50%	−13.17%
htwiki	35	3632	3	7	8.57%	0.19%	−8.38%
cywiki	57	459	6	21	10.53%	4.58%	−5.95%
lawiki	30	564	3	26	10.00%	4.61%	−5.39%
ruwiki	1224	440658	86	7452	7.03%	1.69%	−5.34%
bewiki2	217	4546	9	80	4.15%	1.76%	−2.39%
ckbwiki2	46	991	3	42	6.52%	4.24%	−2.28%
pswiki	36	1032	1	10	2.78%	0.97%	−1.81%
azwiki	214	4583	53	1059	24.77%	23.11%	−1.66%
arzwiki2	324	19663	10	321	3.09%	1.63%	−1.45%
eowiki2	274	6956	7	120	2.55%	1.73%	−0.83%
nqowiki	21	45	1	2	4.76%	4.44%	−0.32%
tgwiki	196	1106	8	42	4.08%	3.80%	−0.28%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

January 2022 through March 2022 (Q3 2022)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: January 2022 through March 2022 (Q3)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
etwiki	17	4503	5	528	29.41%	11.73%	−17.69%
skwiki	29	2587	8	424	27.59%	16.39%	−11.20%
mnwiki	17	378	11	214	64.71%	56.61%	−8.09%
bewiki2	165	5291	15	99	9.09%	1.87%	−7.22%
ltwiki2	31	2996	10	754	32.26%	25.17%	−7.09%
astwiki	27	1833	3	105	11.11%	5.73%	−5.38%
ckbwiki	69	7850	4	41	5.80%	0.52%	−5.27%
eowiki	202	10469	4	67	1.98%	0.64%	−1.34%
bclwiki	176	508	2	3	1.14%	0.59%	−0.55%
bswiki	145	2143	7	92	4.83%	4.29%	−0.53%
ttwiki	497	95099	2	33	0.40%	0.03%	−0.37%
arzwiki2	659	55894	4	272	0.61%	0.49%	−0.12%
skrwiki	295	526	2	3	0.68%	0.57%	−0.11%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

October 2021 through December 2021 (Q2 2021)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: October 2021 through December 2021 (Q2)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
ltwiki	38	4431	18	676	47.37%	15.26%	−32.11%
lawiki2	21	1084	4	45	19.05%	4.15%	−14.90%
ocwiki	20	405	2	12	10.00%	2.96%	−7.04%
arzwiki	62	170066	4	219	6.45%	0.13%	−6.32%
hywiki	43	2328	9	344	20.93%	14.78%	−6.15%
swwiki	71	1673	4	19	5.63%	1.14%	−4.50%
fiwiki2	38	8220	5	715	13.16%	8.70%	−4.46%
thwiki2	69	7307	4	157	5.80%	2.15%	−3.65%
mrwiki2	98	3421	5	52	5.10%	1.52%	−3.58%
iswiki	20	706	2	54	10.00%	7.65%	−2.35%
afwiki	82	1309	5	63	6.10%	4.81%	−1.28%
nlwiki	236	14644	36	2175	15.25%	14.85%	−0.40%
orwiki	150	383	5	12	3.33%	3.13%	−0.20%
kmwiki	43	557	7	90	16.28%	16.16%	−0.12%
bewiki	154	5133	3	95	1.95%	1.85%	−0.10%
twwiki	82	274	3	10	3.66%	3.65%	−0.01%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

July 2021 through September 2021 (Q1 2021)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: July 2021 through September 2021 (Q1)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
jvwiki2	105	434	101	52	96.19%	11.98%	−84.21%
kuwiki	21	12573	3	69	14.29%	0.55%	−13.74%
skwiki	34	1341	9	266	26.47%	19.84%	−6.63%
fawiki2	1340	57805	169	3789	12.61%	6.55%	−6.06%
lvwiki	70	2224	10	232	14.29%	10.43%	−3.85%
lawiki	17	958	2	77	11.76%	8.04%	−3.73%
mrwiki2	66	5294	2	58	3.03%	1.10%	−1.93%
thwiki	34	7722	2	316	5.88%	4.09%	−1.79%
fiwiki	33	8375	3	662	9.09%	7.90%	−1.19%
aswiki	25	1666	1	57	4.00%	3.42%	−0.58%
zh_yuewiki	35	9546	1	218	2.86%	2.28%	−0.57%
eowiki	261	7012	3	59	1.15%	0.84%	−0.31%
ttwiki2	579	20023	3	43	0.52%	0.21%	−0.30%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

April 2021 through June 2021 (Q4 2021)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: April 2021 through June 2021 (Q4)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
arywiki	25	1461	9	24	36.00%	1.64%	−34.36%
jvwiki	890	721	162	22	18.20%	3.05%	−15.15%
kawiki2	48	4558	10	519	20.83%	11.39%	−9.45%
afwiki	100	1735	10	93	10.00%	5.36%	−4.64%
astwiki	37	14525	1	150	2.70%	1.03%	−1.67%
mrwiki2	99	5864	2	60	2.02%	1.02%	−1.00%
ttwiki	952	32396	10	22	1.05%	0.07%	−0.98%
crhwiki	75	1883	1	8	1.33%	0.42%	−0.91%
bewiki2	174	4215	5	87	2.87%	2.06%	−0.81%
hiwiki	385	2735	93	639	24.16%	23.36%	−0.79%
ckbwiki	94	3892	3	97	3.19%	2.49%	−0.70%
pswiki	47	335	2	13	4.26%	3.88%	−0.37%
swwiki	95	3305	2	59	2.11%	1.79%	−0.32%
urwiki	132	3043	7	158	5.30%	5.19%	−0.11%
fawiki	1897	81694	110	4699	5.80%	5.75%	−0.05%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period 2 Also identified in the prior quarter as a wiki with a higher deletion ratio for articles created with Content Translation

January 2021 through March 2021 (Q3 2021)

Wikis with higher deletion ratios for articles created with Content Translation
Reviewed Time Period: January 2021 through March 2021 (Q3)
Wiki project1	Created Articles		Deleted Articles		Deletion Ratios
	Created CX Articles	Created non-CX Articles	Deleted CX Articles	Deleted non-CX Articles	CX Articles Deletion Ratio	Non-CX Articles Deletion Ratio	Deletion Ratio Difference
hawwiki	64	85	25	1	39.06%	1.18%	−37.89%
kuwiki	204	4011	34	69	16.67%	1.72%	−14.95%
lawiki	25	923	3	48	12.00%	5.20%	−6.80%
ltwiki	23	2272	13	1135	56.52%	49.96%	−6.57%
fiwiki	83	9448	12	787	14.46%	8.33%	−6.13%
fiu_vrowiki	16	131	1	4	6.25%	3.05%	−3.20%
eowiki	123	5221	5	62	4.07%	1.19%	−2.88%
kawiki	122	5434	23	889	18.85%	16.36%	−2.49%
arzwiki	110	37033	3	335	2.73%	0.90%	−1.82%
thwiki	17	4635	1	208	5.88%	4.49%	−1.39%
bewiki	256	5225	9	115	3.52%	2.20%	−1.31%
mrwiki	164	4771	3	60	1.83%	1.26%	−0.57%
bswiki	52	1953	5	187	9.62%	9.58%	−0.04%
1 Excludes wikis with 15 or fewer articles created with Content Translation during the reviewed time period

Monthly deletion ratios for representative wikis (Jan 2016 - Jan 2019)

Monthly data about the deletion of articles created with and without Content Translation on several Wikipedias was prepared as part of T215397. This data is for whole months, not quarters, from January 2016 until January 2019.

The results can be found in a Google spreadsheet.

This spreadsheet is publicly shared and can be filtered, copied, etc. Note that only pages in the main namespace are counted. This may lead to discrepancies between this data and the data at Special:CXStats, which includes all namespaces.

The query

To examine the queries used to created and to run this yourself, see query 53775 in Quarry. To look at different languages and dates, replace the database name and the timestamp value.

The use of the CX deletion statistics comparison data

One of the ways the WMF Language team uses this data is to determine when to adjust the machine translation limit in the Wikis to enforce the review and modification of initial machine translation before articles are published to encourage quality translations. Below are some criteria for changing Machine Translation limits in the tool when there is a high deletion of CX articles in any Wikipedia.

SN	Instance/case	Problem (check Content translation/Deletion statistics comparison to determine this)	Frequency	Possible causes (To be populated based on findings, experiences and feedback)	Action (This column can change to iterate learnings)
	Translations deleted	More than 50% of articles are deleted. (Please note that the 50% applies when there are more than ten translated articles in a quarter or more; it will not apply if the articles created are less than 10)	Occurs in one quarter		Investigate
			Occurs consecutively in two quarters (5 to 6months)		Take action based on findings from earlier investigation. If no tangible findings, make Machine Translation (MT) limit more strict by 5% and monitor for changes.
		More than 75% of articles are deleted. (Please note that the 50% applies when there are more than ten translated articles in a quarter or more; it will not apply if the articles created are less than 10)	Occurs in one quarter.		Make MT limit more strict by 10% and monitor changes
			Extends to another quarter.		Community Consultation required to: Gather samples of problematic translations to increase Machine translation based on logic. Understand if there are underlying issues with the tool.
	Difference in deletions: translations (CX) vs. new articles (non-CX)	The difference between the two deletion ratio (non-CX and CX)  is above -50%	Occurs in a quarter		Make MT limit more strict by 5% and monitor changes.
			Extends to another quarter		Make MT limit more strict by another 5% and monitor changes.
		The difference between the two deletion ratio (non-CX and CX)  is above -80%	Occurs in a quarter.		Make MT more strict by 10% and monitor changes.
			Extends to another quarter		Community Consultation required to: Gather samples of problematic translations to increase Machine translation based on logic. Understand if there are underlying issues with the tool.
	Translations created	If articles created with CX decrease by 90% from previous month	First quarter		Investigate, check for technical issues.
			Three months consecutively		Community consultation, from the outcome, determine the next step. The next step can be making the MT limit less strict by 10% and monitor changes.