Last Chance to Read
Your Account
Sign In  or  Sign Up
Basket
Your Basket
Your basket is empty
Payment methods accepted on LCTR website
 
britishspy
 
You are here:   
 

Advanced Document Search

 
You can modify the options below the search box to get different search results. For more information and help about searching, please see the 'help' section below.
Search Options
Search criteria:     Fuzzy:    Fuzzy Tolerance:     Phonic:    Stemming:
Search from:        to end of:   

Searching the Database

 
You can search by one or more words across all content for all dates, either by "exact phrase", "all of these words", or "any of these words." If these give too many results you can restrict each search by including a filter on a date range.

The way in which the site is indexed means that you can also perform wildcard searching using the characters *, ?, and =.



  • * matches any number of characters. So for example searching for Fre*k, would find Fredk and Frederick.
  • ? replaces a single letter. So Sm?th would find Smith and Smyth
  • = matches any single digit for example 1=== would find 1234 or 1840 but not 2345 or A123
The wildcards can also be used in combination for example *annin* would find Channing, Canning, etc

However please note to speed searching, common words such as: “the”, “and”, “if”, and “my” have been excluded from index

Back to top

Some Strategies for Searching

 
You have searched the database but unfortunately you did not find anything or did not find as much as expected.

So recommended search strategies include checking short as well as long forms of the forename so Frederick Smith might appear as :

  • Fred Smith,
  • Fredk. Smith
  • Mr. F. Smith
  • Smith, Frederick
  • Smith, Fredk.
  • Smith, F.
A fundamental difficulty will always be that the original newspaper is unlikely to be in pristine condition and even if it were, that the optical character recognition (OCR) result may not be perfect.

Occasionally the first letter may be missed altogether (especially if printed in a different font from the rest of the text) so that "stranger" can become the word and rare surname "ranger."

Another challenge for newspapers in particular is that words can be split over more than one line to fit the fixed column width. So the word "requested" should it be printed at the end of the column can appear as "re-" at the end of one line and "quested" at the start of the line beneath it. So that a search for the rare surname "Quested" may result in a number of erroneous "hits."

So until the newspapers have been manually checked, if your surname begins with H or W consider searching using II or VV as the first characters (If the print is faint the horizontal bar in the H can be lost and the W becomes two V’s).

Other common misinterpretations* by OCR software include c for e, i for l, and o for p and q (if the tail of the p or q is lightly printed).

It is also worth mentioning that before he early years of the 19th Century (pre 1804 ?) the print font used by newspapers included a long serpentine “s” (looks like an f without the bar across). This was not used for capital letters or for letters at the end of a word. Therefore it should not impact searching for the surname "Smith" for example. However if an "s" appears within a word for example Clarkson, it would be worth searching for Clarkfon and Clarklon as well as Clarkson.

* For a more comprehensive review of potential misinterpretations which might occur with OCR software see separate FAQ
Back to top

Some Common Misinterpretations of Letters by OCR Software

 

Upper Case Characters

  • H = II or B or Il (I and lowercase L)
  • W = VV and more rarely AV
  • J = I
  • T = ‘I’ or ‘1’ and rarely X
  • R = J
  • V = Y



Lower Case Characters



  • d = cl or il (interchangeable)
  • e = c (interchangeable) and e = o and occasionally e = s
  • h can appear as b or li
  • m = in
  • n can frequently be interchanged with u
  • q = o
  • t = f
Back to top

Finally the J O S E P H   P R A T T effect

 
Occasionally, were a name appears in a heading and is capitalised, the OCR will interpret the name as a string of seperate Letters so that JOSEPH PRATT (the name were I first found this issue) has appeared as
  • J O S E P H  P R A T T
So as a final search strategy you might want to consider searching for a surname with spaces between each letter,


Back to top
Ask a Question
Name:
Email:
Tel:
Query: