On 07/28/2014 12:04 PM, Karen Coyle wrote:
> On 7/27/14, 4:03 PM, Stuart Yeates wrote:
>> On 07/28/2014 08:58 AM, Karen Coyle wrote:
>>> With big data, the alphabet is not the right tool.
>> I believe that this reflects a flawed assumption.
>> Not all BIBFRAME-using collections will be 'big.'
> That's absolutely right. So with a mix of big and little, can we use the
> same rules?
There are places where lexical collation is clearly wrong approach
("Your discovery-layer search for 'bear' returned 1.5 million hits, here
are the first 20: 'Aardvark bears of Africa', 'Aardvark bears of
Antarctica' ...") and some where are less clear ("Your search for 'Barak
Obama' did not match any name records, the closest matches were: 'Barack
Obama / Barack Hussein Obama II (1961-)', ...").