Test of Text-to-Speech (TTS) Abbreviations in Waze Clients

[ img ] This is the place to discuss issues that are relevant for locations in the US. For any other discussions, please use the main forums.

Moderator: MapSir

Re: Roman numerals

Postby dbraughlr » Thu Jun 05, 2014 5:16 pm

I favor eliminating all Roman numerals as abbreviations. The disadvantages far outweigh any advantages. Pronunciation of "XIV" can be addressed when the pronunciation of "Louis" is addressed.
dbraughlr
 
Posts: 569
Joined: Tue Aug 13, 2013 2:24 am
Has thanked: 164 times
Been thanked: 98 times

Re: Roman numerals

Postby dbraughlr » Wed Jun 04, 2014 7:32 pm

GizmoGuy411 wrote:I agree with sketch in that I am really not very concerned about Roman Numerals and that the letters themselves are more important at least for North America using the English voices.

We all agree.

GizmoGuy411 wrote:What I am MORE concerned with is why these variations exist in the first place and if they are indicators of some bad RegEx that needs to be addressed at the Waze end, as other parts of the world may need Roman Numerals.


Scrapping Roman numerals is a good way to avoid the bugs and problems they cause.

GizmoGuy411 wrote:Actually I don't think there are many more instances than those I listed as potential conflicts with Roman Numerals. This was limited to valid Roman Numerals, not simply every possible combination of letters used in Roman Numerals. I'll edit that in the prior post.


The highest number you gave in the prior post was MM. In order for that range of Roman numerals to work, all 2000 combinations from "I.", "II.", to "MCMXCIX.", "MM." would have to be defined as abbreviations.
That seems unrealistic. That's why I suggested an upper limit like 30.
dbraughlr
 
Posts: 569
Joined: Tue Aug 13, 2013 2:24 am
Has thanked: 164 times
Been thanked: 98 times

Re: Roman numerals

Postby GizmoGuy411 » Wed Jun 04, 2014 3:10 pm

dbraughlr wrote:
GizmoGuy411 wrote:Therefore we may need to follow suite for the letters used for Roman Numerals too.

From what you listed, there are thousands of them. That's clutter.
What range do you really intend to cover?

GizmoGuy411 wrote:So maybe Roman Numerals could use a period after them instead of quotes to differentiate them.

That doesn't seem likely, at least not if you intend to cover a range beyond 1 to 30.

What's wrong with using Louis the 14th instead of Louis XIV.? If anything, our experience with N/E/S/W should tell us that it was a bad precedent that should not be expanded.


I agree with sketch in that I am really not very concerned about Roman Numerals and that the letters themselves are more important at least for North America using the English voices.

What I am MORE concerned with is why these variations exist in the first place and if they are indicators of some bad RegEx that needs to be addressed at the Waze end, as other parts of the world may need Roman Numerals.

So any submission I may to Waze at this point will be to just point out the inconsistencies. We have plenty of one off issues with TTS pronunciation that is something that may be able to be adjusted in the future via the editor on an individual basis.

Issues like "Sulgrave Dr" being pronounced in the past as "Sulgrave Doctor" however where of concern as in this case it may have been indicative of a wider problem that could effect other names as Doctor vs Drive was more of a global concern.

Actually I don't think there are many more instances than those I listed as potential conflicts with Roman Numerals. This was limited to valid Roman Numerals, not simply every possible combination of letters used in Roman Numerals. I'll edit that in the prior post.
Not sure I agree with you that there are thousands of RN combinations, as many combinations of the letters in RNs are not actually valid RNs.
GizmoGuy411
Global Champ Mentor
Global Champ Mentor
 
Posts: 1406
Joined: Wed Oct 13, 2010 3:14 am
Location: NW Ohio, SE Michigan, NW Indiana tri-state area
Has thanked: 654 times
Been thanked: 430 times

Re: Roman numerals

Postby dbraughlr » Wed Jun 04, 2014 7:37 am

GizmoGuy411 wrote:Therefore we may need to follow suite for the letters used for Roman Numerals too.

From what you listed, there are thousands of them. That's clutter.
What range do you really intend to cover?

GizmoGuy411 wrote:So maybe Roman Numerals could use a period after them instead of quotes to differentiate them.

That doesn't seem likely, at least not if you intend to cover a range beyond 1 to 30.

What's wrong with using Louis the 14th instead of Louis XIV.? If anything, our experience with N/E/S/W should tell us that it was a bad precedent that should not be expanded.
dbraughlr
 
Posts: 569
Joined: Tue Aug 13, 2013 2:24 am
Has thanked: 164 times
Been thanked: 98 times

Re: Roman numerals

Postby GizmoGuy411 » Wed Jun 04, 2014 6:55 am

dbraughlr wrote:Whether Louis XIV is read as lewis ex eye vee or lewis fourteenth, it won't be perfect.

I think we want C, D, I, L, M, V, and X pronounced as letters. It seems that Roman numerals are the exceptions and should require special notation if used at all.


While the logic that the Roman Numerals are the exceptions, and therefore should require special notations seems correct, we have to remember that we have already have set a precedence to use quotes around the letters N, E, S, and W, to get the them pronounced as letters instead of the cardinal directions.

Therefore we may need to follow suite for the letters used for Roman Numerals too.

Of course there are other conflicts to consider as well including: (examples using valid Roman Numerals only, and not every combination of the letters used in Roman Numerals)

CD = cd or 400 ?
CM = centimeter or 900 ?
DC = District of Columbia or 600 ?
MD = Maryland or 1500 ?
MI = Michigan or 1001 ?
MIX = Mix or 1009 ?
MM = millimeter or 2000 ?

So maybe Roman Numerals could use a period after them instead of quotes to differentiate them.
GizmoGuy411
Global Champ Mentor
Global Champ Mentor
 
Posts: 1406
Joined: Wed Oct 13, 2010 3:14 am
Location: NW Ohio, SE Michigan, NW Indiana tri-state area
Has thanked: 654 times
Been thanked: 430 times

Re: Roman numerals

Postby dbraughlr » Mon May 12, 2014 6:34 pm

Whether Louis XIV is read as lewis ex eye vee or lewis fourteenth, it won't be perfect.

I think we want C, D, I, L, M, V, and X pronounced as letters. It seems that Roman numerals are the exceptions and should require special notation if used at all.
dbraughlr
 
Posts: 569
Joined: Tue Aug 13, 2013 2:24 am
Has thanked: 164 times
Been thanked: 98 times

Re: Re: Test of Text-to-Speech (TTS) Abbreviations in Waze C

Postby GizmoGuy411 » Tue Nov 06, 2012 12:58 pm

xteejx wrote:Has anyone asked Waze to look at their TTS service and grab all this information for us? There's gotta be an easier way than testing manually.

Sent from my GT-I9100P using Tapatalk 2


Yes we asked for that at the Palo Alto meetup last January and they said it was not available.


Sent from my Galaxy Nexus using Tapatalk 2
GizmoGuy411
Global Champ Mentor
Global Champ Mentor
 
Posts: 1406
Joined: Wed Oct 13, 2010 3:14 am
Location: NW Ohio, SE Michigan, NW Indiana tri-state area
Has thanked: 654 times
Been thanked: 430 times

Re: Re: Test of Text-to-Speech (TTS) Abbreviations in Waze C

Postby CBenson » Sun Nov 04, 2012 11:30 am

Timbones wrote:
AlanOfTheBerg wrote:If we make it a standard that "St" is "street" and "St." is "saint" then we know exactly how it will be pronounced regardless of where it is in the street name.

Is this going happen? Did the request get put in to HQ?

I don't know. But if it is/has, "Dr" as "drive" and "Dr." and "doctor" should get the same request.
CBenson
Waze Global Champs
Waze Global Champs
 
Posts: 10330
Joined: Wed Nov 03, 2010 9:13 pm
Location: Crownsville, MD, US
Has thanked: 1069 times
Been thanked: 2355 times

Re: Re: Test of Text-to-Speech (TTS) Abbreviations in Waze C

Postby GizmoGuy411 » Fri Sep 21, 2012 10:51 pm

Kandralla wrote:
GizmoGuy411 wrote:Penn for Pennsylvania was moved from the main list to the State abbreviation list where it belongs, along with other alternate state abbreviations. Yes we definitely want it removed from the TTS!


Am I correct in my assumption that this means that William Penn Highway will now be called "William Penn Highway" as oppoesed to "William Pennsylvania Highway"?

If so thanks, that was annoying.


No. Nothing has changed except where we had "Penn" in the test list.

The list has no impact on how Waze handle TTS. It is merely to learn how TTS currently works as Waze has no documentation themselves. The TTS results are actually a product of a company called Nuance.

We still do not know if, how or when our list may be able to help Waze influence the TTS package.

Sent from my Galaxy Nexus using Tapatalk 2
GizmoGuy411
Global Champ Mentor
Global Champ Mentor
 
Posts: 1406
Joined: Wed Oct 13, 2010 3:14 am
Location: NW Ohio, SE Michigan, NW Indiana tri-state area
Has thanked: 654 times
Been thanked: 430 times

Re: Re: Test of Text-to-Speech (TTS) Abbreviations in Waze C

Postby rleejr79 » Fri Sep 21, 2012 10:24 pm

Kandralla wrote:
GizmoGuy411 wrote:Penn for Pennsylvania was moved from the main list to the State abbreviation list where it belongs, along with other alternate state abbreviations. Yes we definitely want it removed from the TTS!


Am I correct in my assumption that this means that William Penn Highway will now be called "William Penn Highway" as oppoesed to "William Pennsylvania Highway"?

If so thanks, that was annoying.


As a fellow Pennsylvanian with a 'Penn St.' , +1.
~ I don't always use my phone to respond in forums. When I do, I Swype on my GNex with Tapatalk 2
rleejr79
 
Posts: 188
Joined: Wed Sep 23, 2009 8:50 am
Location: Southwestern PA
Has thanked: 6 times
Been thanked: 6 times

PreviousNext

Return to United States

Who is online

Users browsing this forum: No registered users