Notes

Accuracy of OCR and overProof is measured in comparison with the human corrections. We know human corrections in this sample are incomplete, and themselves contain errors, but they are the best we could find automatically from the NLA newspapers corpus, tagged as completely corrected then further filtered to those with at least 3 corrections, at least 40% of lines corrected and lowest third percentage of non-dictionary words.
Accuracy is measured by a separate process from that used to colour words in this output: the colouring process is heuristic, and not completely accurate.
Colour legend:
Text - OCR text corrected by human and/or overProof
Text - human and/or overProof corrections
Text - discrepencies between human and/or overProof
Text - human corrections not applied by overProof
Identified overProof corrections are calculated by the statistical calculation process, and shows those words changed by overProof which ALSO match human corrections. As human corrections are often wrong and incomplete, so too is this list.
Identified overProof non-corrections are calculated by the statistical calculation process, and shows those words in the overProof output which DO NOT MATCH human corrections. As human corrections are often wrong and incomplete, so too is this list. Words marked as [**VANDALISED] are those which have been changed by overProof but not by the human correction; as before, a missed human correction will be (incorrectly) classified as vandalisation by overProof.
Searchability of unique words refers to the distinct words in an article, and how many are present before and after correction. It is measure of how many of the words within an article could be used to find the article using a search engine.
Weighted Words refers to a calculation in which common words count for little (a fraction of a word) and unusual words count for more, in proportion to the log of the inverse of their frequency in the corpus. It may be an indicator of how well distinctive words in an article can be searched before and after correction.

Article ID 79264355, Article, OLD BOTTLE MESSAGE, page 14 1956-05-31, The Central Queensland Herald (Rockhampton, Qld. : 1930 - 1956), 83 words, 4 corrections

Raw OCR	Human Corrected	overProof Corrected
OLD BOTTLE	OLD BOTTLE	OLD BOTTLE
MESSAGE	MESSAGE	MESSAGE
PERTH. May 27.-Probablv	PERTH. May 27.-- Probably	PERTH. May 27. Probably
throw overboard from tone of	thrown overboard from one of	throw overboard from tone of
Augtmllafe fim troopamps tn	Australia's first troopships in	Augtmllafe from troopships in
Wort<f Wir I, a bottle eog	World War I, a bottle con-	World War I, a bottle dog
tln!n<- three faded sheet of	tainer three faded sheets of	trials- three faded sheets of
notepi^per has been found at	notepaper has been found at	notepaper has been found at
Doubtful Island Bay. on the	Doubtful Island Bay, on the	Doubtful Island Bay on the
sooth coast of Western Aus	south coast of Western Aus-	south coast of Western Australia.
tralia.	tralia.
One note read: "Off to the	One note read: "Off to the	One note read: "Off to the
Dardanelles. 30/11/1*--half	Dardanelles. 30/11/15--half-	Dardanelles. 30/11/1*--half
way tfr Western Australia"	way to Western Australia"	way to Western Australia"
Then «ame the namea, H	Then came the names, H	Then came the names, H
Notices. Q. Holeenberger tods	Noakes, G. Holzenberger and	Notices. Q. Holeenberger tons
Captain Bromlden, with mn«	Captain Bromiden, with some	Captain Bromlden, with many
others Impossible to decipher	others impossible to decipher.	others impossible to decipher
The bottle was found un	The bottle was found un-	The bottle was found in
corked <on a lonely south	corked on a lonely south	worked on a lonely south
coaet besich.	coast beach.	coast beach.

Identified overProof corrections	BEACH NOTEPAPER WORLD IN NAMES WAR SHEETS TROOPSHIPS CAME PROBABLY
Identified overProof non-corrections	NOAKES AND THROWN BROMIDEN SOME AUSTRALIAS UNCORKED [**VANDALISED] FIRST HOLZENBERGER CONTAINER

	Word count	OCR accuracy %	overProof accuracy %	Errors corrected %
All Words	74	67.6	85.1	54.2
Searchability of unique words	58	67.2	82.8	47.4
Weighted Words		70.8	84.5	47.0

Article ID 84389063, Article, One snake got away, page 7 1956-09-29, The Argus (Melbourne, Vic. : 1848 - 1957), 61 words, 4 corrections

Raw OCR	Human Corrected	overProof Corrected
One snake i	One snake	One snake i
got away j	got away	got away j
; NATHALIA, Friday:	NATHALIA, Friday:	; NATHALIA, Friday:
I Jack and Walter Broom	Jack and Walter Broom	I Jack and Walter Broom
\ were mustering stock at	were mustering stock at	were mustering stock at
Skeleton Creek when	Skeleton Creek when	Skeleton Creek when
the.v saw two brown	they saw two brown	they saw two brown
snakes.	snakes.	snakes.
They killed one and	They killed one and	They killed one and
the other got away.	the other got away.	the other got away.
; The dead snake was	The dead snake was	The dead snake was
7ft. OJin. long, with 6Ain.	7ft. 0½in. long, with 6½in.	7ft. 6in. long, with 54in.
girth, with black spots	girth, with black spots	girth, with black spots
along its back.	along its back.	along its back.
It was the largest	It was the largest	It was the largest
snake seen in this area.	snake seen in this area.	snake seen in this area.

Identified overProof corrections
Identified overProof non-corrections

	Word count	OCR accuracy %	overProof accuracy %	Errors corrected %
All Words	52	98.1	100.0	100.0
Searchability of unique words	41	100.0	100.0	100.0
Weighted Words		100.0	100.0	0.0

	Word count	OCR accuracy %	overProof accuracy %	Errors corrected %
All Words	126	80.2	91.2	55.8
Searchability of unique words	99	80.8	89.9	47.6
Weighted Words		82.9	90.9	46.9

NLA Trial Articles from 1956

Article ID 79264355, Article, OLD BOTTLE MESSAGE, page 14 1956-05-31, The Central Queensland Herald (Rockhampton, Qld. : 1930 - 1956), 83 words, 4 corrections

Article ID 84389063, Article, One snake got away, page 7 1956-09-29, The Argus (Melbourne, Vic. : 1848 - 1957), 61 words, 4 corrections

Accumulated stats for 2 articles from year 1956