Just finished updating the English lexicons for the Office 2007 speller. Here are some highlights:
1. "Teixeira" is recognized, in response to an ESPN Nate Silver column forwarded to us:
...T-E-I-X ... as a baseball writer, I have a vested interest in seeing that Big Tex becomes famous enough that his last name is added to the Microsoft Word spell-check dictionary, like Steven Spielberg ("i" before the "e") or Gerard Depardieu (French). If Teixeira can continue to combine an Eddie Murray-like bat with a Keith Hernandez-like glove, he should be well on his way.
2. More place names are recognized. We support 4 English dialects: US, UK, Canada, and Australia. All users will see an increased coverage in Australian and South African place names, because of the data we were able to purchase and review (thanks to my colleague Ann Brockenbrough). This is the beginning of more directed coverage of these market areas.
3. More user-requested words are recognized. The Australian sub had a contest, to elicit words not yet in the speller for Australian. I'd award the prize to the mate who suggested emu. Can't believe we didn't have that in any of English spellers (though we had emus).
We also processed a bunch of data from user dictionaries in Hotmail and the Office 2007 Beta. This requires permission from the users, of course, and will be an increasingly valuable resource for new items. Words added include names (Kyla, Kaitlin, Grinch, Jolie), companies (Mozilla, PeopleSoft), products (Celebrex, Percocet), and newly inflected words (homeschooling, scrapbooking, texted). Post your requests here!
4. More legitimate repeated words are recognized. Word usually flags repeated words, unless we identify them as exceptions, often place names like Bora Bora and Walla Walla. In Office 2007 we'll permit the following Australian place names (in all dialects) without flagging the repeated words: Baw Baw, Curl Curl, Gin Gin, Kin Kin, Lang Lang, Mia Mia, Mooney Mooney, Nar Nar, Obi Obi, Wangi Wangi, Wirra Wirra.
Of course, there may be very rare instances where an erroneous repetition of these words will go unidentified, e.g. I liked the part in Star Wars where Obi Obi Wan teaches Luke. But, for the most part, users should not notice this as a problem.
-- Mari Broman Olsen (Developer)