Internet Search Engines
- Some Comments for Fast, Efficient Full-name Searching for Family History Purposes

This page was last updated on 24 April 2001   Back to Hugh Reekie's Home Page

The purpose of this document is to recommend a structure for name entry in various genealogical documents on the 'net, in order for family historians to find matches as conveniently and fast as possible,using appropriate and available resources. As the web is changing at an incredible pace, this document reviews status, new features and makes suggestions as seem to be sensible, taking present trends into account. Its primarly relevance is for the United Kingdom. This is a draft with rather scrappy formatting; it is incomplete.

Note - some of the links are obsolete

This document is made up of the following Sections;

1. Introduction and basic assumption

One important assumption or premice is made: in the near future MANY MANY out-of-copyright books will be "on the web" using scanners; they will be indexed by some of the search engines; this process has already started with maps (e.g. Ancestry Home Town Daily).

This rapid major increase of information is both an opportunity and a curse: an opportunity because persons living in the 1700 -1900 time period will be mentioned in some of these books; but also a curse because both users and the search engines themselves will have to me much more sophisticated for effective, succesful, searching. The writer believes that Meta-tag, keyword and partial-document searchingwill have minimal usefulness for this application; it is already reported that the num,ber of indexed web pages is already 50,000,000 (March 98) or some seach engines - ref 111.

For a few months now I have been experimenting with the possibility of using the web to find a person, or ancestor, using a direct search. I have found it not onlypossible, but quick, useful and informative; but I have only found names that have been placed in web pages in specific text formats. I can find Hugh REEKIE and Hugh Reekie quite easily, but to find Reekie, Hugh and REEKIE Hugh requires one or two separate searches. Multiple searches of the two names may still not readily find text such as "REEKIE - looking for ancestors of Hugh, Andrew and Ian Reekie" - here you would get a search result using REEKIE or Hugh or Ian Reekie. I am aware that specific word searches using "within adjacent text" constraints are sometimes possible, which would alleviate this problem somewhat.

I have done some experiments with my own family history web page design (Fife Surname Pages) and checked out a few search engines,with their various specialist search choices, and I have come up with some conclusions. You may jump to the last section to read the them if you wish - Section 7.

I have not been able to locate a simple list of the key parameters of the various major Internet Search Engines; some engines are more crorrectly termed searchable catalogues, for good reason. But a good reference is

http://www.SearchEngineWatch.com/resources/index.html - ref 444 - other general references are in Section 8.

I have had difficulty answering such basic questions as:

Some of this information is not easy to obtain, and my search has been limited; it is not in any way complete.

8. References

The information above has been garnered from various places, including those listed below. The - 111, - 222 convention has been used as a convenient reference method rather than the usual "Ref 1" or "superscript 1" indications.

Comments, submissions and suggestions welcome - Hugh Reekie h.reekie@ieee.org