Pages in topic:   < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16] >
Complimentary TM-Town membership for ProZ.com members
Thread poster: Jared Tabor
Jennifer Forbes
Jennifer Forbes  Identity Verified
Local time: 10:20
French to English
+ ...
In memoriam
Inevitably penalised ... May 8, 2016

Tom in London wrote:

Kevin Dias wrote:

Mirko, this past week you received a job message through TM-Town to localize a mobile game from English to Italian. You were the ONLY translator to be contacted by this client. This client found you through Nakōdo. If quantity was the only factor s/he would not have contacted you. You are not 1st in the traditional directory for English to Italian, nor even English to Italian video games.

This client reached out to you because you have translated similar material in the past to what they need translated...and it is clear from your TM-Town profile that you are professional who can do the job right. You have 9 positive reviews, you are a certified PRO, you are a member of 2 professional associations, you have 5 years of experience in the industry.

Seems like a clear example to me of the system working how it is intended.


And what about other equally good translators who have not subscribed to TM Town? Please confirm that they are not penalised.


It looks to me as if Prozians who have not subscribed to TM Town will inevitably be "penalised", in that those who are TM Town members have uploaded masses of evidence of their prowess, giving them an obvious advantage over standard (bog standard?) Prozians who have not done so, when outsourcers are seeking translators.
For various reasons (paranoid or otherwise), I don't want to join TM Town. I don't work with TMs - except the one in my brain (known as "wet ware", I believe). My membership of Proz is up for renewal soon and I admit I'm beginning to wonder whether I should renew, much as I enjoy so many of Proz's other features. The value of Proz membership seems to have been somewhat downgraded by the arrival of TM Town.


 
Tom in London
Tom in London
United Kingdom
Local time: 10:20
Member (2008)
Italian to English
Not May 8, 2016

Jenny Forbes wrote:


It looks to me as if Prozians who have not subscribed to TM Town will inevitably be "penalised"......I admit I'm beginning to wonder whether I should renew, much as I enjoy so many of Proz's other features. The value of Proz membership seems to have been somewhat downgraded by the arrival of TM Town.


Jared says not.

jared wrote on 28 Apr

Be careful with speculation


Tom in London wrote:

It does seem that unless you sign up for this TMTown thing, and pay extra, you will actually end up being disadvantaged in your Proz ranking.

jared said in the same post
This is incorrect.


So I'm happy that as a non-TM Town user I am not suffering in any way. However doesn't this mean that there is no advantage in using TM Town ?

[Edited at 2016-05-08 19:55 GMT]


 
Mirko Mainardi
Mirko Mainardi  Identity Verified
Italy
Local time: 11:20
Member
English to Italian
Not a matter of time... May 8, 2016

Kevin Dias wrote:

The fact is, we are just getting started. The first translator signed up for TM-Town in November 2014. The first year I was the only developer. Nate joined in October of last year. Ask anyone that has been around TM-Town since the beginning and I think they will tell you we have come a long way. We will still have a long way to go and a lot to do though, no doubt - that is the exciting and fun part.


The problem in using a metric such as TUs for ranking (with the underlying weaknesses to manipulation discussed at length before) is structural. Time won't change that...

Also, could you comment on why you chose TUs instead of words for ranking purposes, as previously asked?


Mirko Mainardi wrote:
quantity is basically the only deciding factor for visibility


Again, I disagree. Please re-read my prior post on the matter. Our goal is to move away from the traditional directory and get more end clients and agencies to use Nakōdo to find translators. Nakōdo recommends translators based on the type of work they have translated in the past. While quantity does play a role, it is definitely not the only deciding factor for visibility.


Although you just cherry picked a half sentence from my post, ignoring the rest, that doesn't change the fact that, right now, if you do a directory search on TM-Town, translators are ordered by type of membership ("professional" vs. "non professional"(!?)...) AND then by the number of uploaded TUs. That's a fact, not an opinion. Good thing you plan to "phase out" that concept in the future, but right now, that's how things work. Besides, you yourself wrote that, with the present system, you are actively "encouraging translators to load more work"... so, one could think that right now you're just trying to hoard as many TUs as possible, offering "visibility incentives" for this, and, based on what you wrote, you plan to change that in the future... Very transparent.

As far as I could see, the Nakōdo search engine is based, once again, on quantity, and "relevance" (compared to a sample text), and, again, not on quality, so I really don't see much to agree or disagree about, especially considering that this too is completely open to the same type of manipulation and abuse I mentioned in my previous post. E.g. If I want to be found by customers in the legal field, all I have to do is find a huge amount of legal documents, machine (or otherwise) "translate" them and upload them or analyze them with your offline tool. Et voilà, I'll rank #1 in both the directory AND the Nakōdo searches in that pair and for that speciality... Easy as pie.
So, in light of that, how is a SEO-like approach that disregards quality and is open to abuse any better that the simple quantitative search used by the directory search?

I do agree that such a tool could prove useful, but only if combined with quality and "reliability" indicators (and even then, with caution), or as an extension to ProZ features. This, however, goes beyond visibility. Being "visibile" and being "good" are (unfortunately) two entirely different things, although the first one still plays an equally (if not even more) decisive role in choosing a product/service, as you surely know better than me...

Mirko, this past week you received a job message through TM-Town to localize a mobile game from English to Italian. You were the ONLY translator to be contacted by this client.


That's interesting, because, after telling them my rate, they replied saying they had to check other quotes before getting back to me. Truly shocking...

This client reached out to you because you have translated similar material in the past to what they need translated...and it is clear from your TM-Town profile that you are professional who can do the job right.


Thanks for the vote of confidence, appreciated. However, I believe that could also be easily explained by the fact that the first translator appearing on the VG search link you quoted does state her rates, which are incidentally higher than the budget specified by that client (and the rate I proposed is higher than that, too)... In addition to that, I DID upload "Deshi analyses" of several of my gaming-related TMs, so I expect there to be relevant terminology for matching in them (I can't speak for other translators, of course).

BTW, one thing I did not understand when setting up my profile, is why you have categories such as "games", "gaming" and "video games", which sound kind of confusing (and redundant), both for translators and clients...

A classification system such as that used on ProZ would probably be more intuitive and comprehensible.

In closing, I'd like to restate what I wrote on page 3 of this thread: "the main point is advertising your services and be found, so I guess the same could be said for TM-Town". I welcome increased exposure and "opportunities to be found", and that's the main reason why I decided to take Proz up on their "gracious" offer, although, as I repeatedly wrote, I am convinced the "ranking system" you have devised is deeply flawed, mainly because it disregards quality while blindly awarding quantity and because it easily allows "cheating".

Edited for "exposition"

[Edited at 2016-05-09 08:44 GMT]


 
Mirko Mainardi
Mirko Mainardi  Identity Verified
Italy
Local time: 11:20
Member
English to Italian
Don't really care... May 8, 2016

Tomas Mosler, DipTrans IoLET MCIL MITI wrote:

Kevin Dias wrote:
...this past week you received a job message through TM-Town to localize a mobile game from English to Italian.

Is TM Town itself publicly naming a translator who was approached within an allegedly confidential and secure environment?

Or did the translator in question approve this announcement in advance?


The translator in question didn't, but he doesn't really care either...

He just hopes he won't unknowingly become a(n even more) published "case study"


 
..... (X)
..... (X)
Local time: 19:20
It is combined with quality and reliability indicators May 8, 2016

Mirko Mainardi wrote:
I do agree that such a tool could prove useful, but only if combined with quality and "reliability" indicators (and even then, with caution), or as an extension to ProZ features.


It is currently combined with quality and "reliability" indicators and also extensions of ProZ features.

Take a look at your TM-Town profile:
• Certified PRO Badge (indicator of quality & reliability + extension of ProZ feature)
• 9 WWA positive reviews (indicator of quality & reliability + extension of ProZ feature)
• KudoZ points summary (extension of ProZ feature)
• Member of 2 professional associations (indicator of quality & reliability)

Mirko Mainardi wrote:
Also, could you comment on why you chose TUs instead of words for ranking purposes, as previously asked?


A translation unit is well defined and easy to calculate measure of quantity. Word count on the other hand is a much grayer area, see my blog post on that issue. Not only that, but then you run into issues across different languages. Languages with character counts (Japanese, Chinese, etc.) would not be easily compared against Roman languages.


 
Tomas Mosler, DipTrans IoLET MCIL MITI
Tomas Mosler, DipTrans IoLET MCIL MITI  Identity Verified
Czech Republic
Local time: 11:20
Member (2008)
English to Czech
SITE LOCALIZER
Nakodo May 9, 2016

"...Nakōdo will only search translation memory (TM) files or aligned source and target documents. Therefore, if you only have unaligned source and target documents in your TM-Town account you will not show up in the results."
https://www.tm-town.com/faq

Two questions:
1) Does Nakodo accept only true bilingual data (source plus target document or TM), or can be Deshi fil
... See more
"...Nakōdo will only search translation memory (TM) files or aligned source and target documents. Therefore, if you only have unaligned source and target documents in your TM-Town account you will not show up in the results."
https://www.tm-town.com/faq

Two questions:
1) Does Nakodo accept only true bilingual data (source plus target document or TM), or can be Deshi files used for that? (As discussed earlier, the understanding of "aligned" seems to differ.)

2) What options do the translators, who want to protect confidentiality of their work, have to remain competitive in terms of TU/quantity ranking? (I mean, "Deshi-ing" or redacting 100k words is not very feasible.)

It is currently combined with quality and "reliability" indicators and also extensions of ProZ features.

Some of the indicators listed as indicators of quality actually say little (objectively, not case to case) about quality.

But what I'm curious about - are these indicators used to improve the (TU-count derived) TM Town ranking? E.g. if one translator had 1,000 TUs + one association, and another 900 TUs + two associations + 10 WWA ratings, could the latter one be first? Or are these indicators just attached to profiles to provide additional information once the profile is visited?
Collapse


 
Mirko Mainardi
Mirko Mainardi  Identity Verified
Italy
Local time: 11:20
Member
English to Italian
IF you're a ProZ member... May 9, 2016

Kevin Dias wrote:

Mirko Mainardi wrote:
I do agree that such a tool could prove useful, but only if combined with quality and "reliability" indicators (and even then, with caution), or as an extension to ProZ features.


It is currently combined with quality and "reliability" indicators and also extensions of ProZ features.

Take a look at your TM-Town profile:
• Certified PRO Badge (indicator of quality & reliability + extension of ProZ feature)
• 9 WWA positive reviews (indicator of quality & reliability + extension of ProZ feature)
• KudoZ points summary (extension of ProZ feature)
• Member of 2 professional associations (indicator of quality & reliability)


Because I was a (paying) ProZ member to begin with, and because I had entered that info, and I believe that neither is mandatory, is it? If I create a profile on TM-Town, pay for "professional" (again, a big "?" here) membership, write a lot of stuff (30 years of experience, member of this and that, completed 6 post-grad degrees and 3 PhDs, translated 40 millions words, climbed the Everest barefoot, etc.), upload a heap of TUs, I'll just rank first and look even more experienced, reliable, qualified, etc. than anyone else... Or are you actively vetting and checking such claims? And as written before, such check will be basically impossible for TMs (in most cases), and TMs are the fulcrum around which your entire platform revolves (as its name kind of subtly hints)...

Right now it seems to me you're heavily relying on ProZ for that aspect (and ProZ does carry out some sort of checks...), but since ProZ membership isn't mandatory, that doesn't really sound like a universal solution to the quality/reliability issue... Besides, as Tomas wrote: "Some of the indicators listed as indicators of quality actually say little (objectively, not case to case) about quality".

Also, I think you didn't get what I was referring to when I wrote "extension of ProZ features". What I actually meant is literally that, that is to say, the "Nakōdo search" as an additional tool to look for translators on ProZ, while now it is a separate tool, on a separate platform, with separate (and differently tiered) memberships, which may or may not overlap.


Mirko Mainardi wrote:
Also, could you comment on why you chose TUs instead of words for ranking purposes, as previously asked?


A translation unit is well defined and easy to calculate measure of quantity. Word count on the other hand is a much grayer area, see my blog post on that issue.


Sorry, but I don't see how that is relevant... Based on what I read, you have devised your own "word count analyzer", which should already be more than enough to solve word-count discrepancies for files uploaded to your platform, but, in any case, I would hardly consider a discrepancy of 36 words on 2.7k (as in your example) counted with different tools as an issue...

Besides, that actually doesn't explain at all why TUs are a "better" metric than words... "Translation units" are definitely quite the opposite of a "well defined measure"... While a TU can basically contain any number of words, from 1 to ∞ (depending on punctuation and segmentation rules), a word is a word (or "a character is a character"), which is a much more objective and "well defined measure".

In the example I made in a previous post, "if I translate a 1,000 words glossary, I could end up with a 1,000 TUs TM, while if I translate a cohesive source text of 1,000 words, I could end up with a 40/50 TUs TM". Do you really believe that, having translated that glossary, I would gain 20/25 times more "experience" than I would by translating the 1,000 cohesive text, and that I should be given "priority" for that? O_o

That is obviously an "extreme" example, but very real. And, by the way, it would be much easier to create "fake TMs" using single word (or "keyword") lists, which in turn would also maximize the number of resulting TUs...

Extending this concept to "non fraudulent" scenarios, here's another example: localization strings are generally much shorter than those resulting from the segmentation of "normal" text, so, someone who specializes in, say, software localization, will end up having a much higher number of TUs than someone translating, for instance, marketing or legal copy, although the latter could very well have translated many more source words...

Not only that, but then you run into issues across different languages. Languages with character counts (Japanese, Chinese, etc.) would not be easily compared against Roman languages.


How is that even an issue? There's no reason why word/character counts in different languages should be compared to begin with, and I frankly don't see a problem with that, as nobody would be comparing the number of words translated by an EN>IT translator with the number of characters translated by a JP>EN translator... The comparison would only take place within each language pair, as it should. The same applies to directory and Nakōdo searches, which also are (obviously) pair specific.


Edited for quote tags.

[Edited at 2016-05-10 00:14 GMT]


 
Mirko Mainardi
Mirko Mainardi  Identity Verified
Italy
Local time: 11:20
Member
English to Italian
List of source words May 9, 2016

Tomas Mosler, DipTrans IoLET MCIL MITI wrote:

"...Nakōdo will only search translation memory (TM) files or aligned source and target documents. Therefore, if you only have unaligned source and target documents in your TM-Town account you will not show up in the results."
https://www.tm-town.com/faq

Two questions:
1) Does Nakodo accept only true bilingual data (source plus target document or TM), or can be Deshi files used for that? (As discussed earlier, the understanding of "aligned" seems to differ.)


While waiting for a reply from Kevin, and considering I have exclusively uploaded Deshi analyses, I believe those also work with the Nakōdo search, since someone using it will do so by inputting a sample of source text. Not sure how accurate that search is, though, as it seems it would be necessarily limited to single and isolated words rather than sentences, parts of sentences or even composite terms.


2) What options do the translators, who want to protect confidentiality of their work, have to remain competitive in terms of TU/quantity ranking? (I mean, "Deshi-ing" or redacting 100k words is not very feasible.)


Actually, I did that... Sure, it takes quite a lot of time for "big" TMs, but it does work, and the output basically is a list of "keywords"...

It is currently combined with quality and "reliability" indicators and also extensions of ProZ features.


Some of the indicators listed as indicators of quality actually say little (objectively, not case to case) about quality.

But what I'm curious about - are these indicators used to improve the (TU-count derived) TM Town ranking? E.g. if one translator had 1,000 TUs + one association, and another 900 TUs + two associations + 10 WWA ratings, could the latter one be first? Or are these indicators just attached to profiles to provide additional information once the profile is visited?


Indeed, that definitely is part of the problem of how ranking works (IMO).


 
..... (X)
..... (X)
Local time: 19:20
If you don't like it... May 9, 2016

I've explained about how TM-Town works and our rationale for various features and mechanisms in great detail in this thread. If you don't like how it works, you don't have to use it. Even better, you are free to create your own web application that works the way you want it to work (whether that is using source words or not having a concept of "ranking").

 
Maxi Schwarz
Maxi Schwarz  Identity Verified
Local time: 05:20
German to English
+ ...
On that issue May 10, 2016

Mirko Mainardi wrote:

Tomas Mosler, DipTrans IoLET MCIL MITI wrote:

Kevin Dias wrote:
...this past week you received a job message through TM-Town to localize a mobile game from English to Italian.

Is TM Town itself publicly naming a translator who was approached within an allegedly confidential and secure environment?

Or did the translator in question approve this announcement in advance?


The translator in question didn't, but he doesn't really care either...

He just hopes he won't unknowingly become a(n even more) published "case study"

I waited before posting until I knew whether you had authorized that announcement. Frankly, I was shocked. My own transactions with clients are private. I would never imagine that the management of any site where my name is listed, and who has access to my private information, would see fit to publish that information. And not just in the body of a post, even in the subject line. I don't know if this falls under confidentiality, but it does seem to fit into privacy, trust, and judgment.


 
Tomas Mosler, DipTrans IoLET MCIL MITI
Tomas Mosler, DipTrans IoLET MCIL MITI  Identity Verified
Czech Republic
Local time: 11:20
Member (2008)
English to Czech
SITE LOCALIZER
No matter if paying ProZ members like TM Town or not, they paid for it May 10, 2016

Kevin Dias wrote:
I've explained about how TM-Town works and our rationale for various features and mechanisms in great detail in this thread.

Can you link me to a post in which you explained
- if Nakodo, unlike the standard directory search, accepts only truly bilingual data, or if it accepts the same input as the translation data for the directory search,
- if the "quality indicators" are in any way reflected in the standard ranking, and
- how are the "reference TMs" (public domain, e.g. DGT), which a user uploaded into TM Town before the separated "reference TM" category was introduced, currently treated in terms of publicly displayed number of TUs? Thanks.

If you don't like how it works, you don't have to use it.

Is this some new mantra? As explained earlier, this is not a relevant argument. The moment ProZ spent money of paying ProZ members on a service, promotes it and enables it to display data from ProZ profiles, these members surely have a right to comment, make suggestions or raise questions, be it about one particular ProZ service or another (especially if such a service couldn't be used on equal conditions by all members) and no matter if they personally use it or not.


 
Elizabeth Tamblin
Elizabeth Tamblin  Identity Verified
United Kingdom
Local time: 10:20
French to English
I'm afraid I haven't read the whole thread May 10, 2016

As I cannot hope to match the large quantities of work many of my colleagues have produced, is there any point at all in my uploading any work to TM-Town? I was thinking of completing some of those sample translations, but is it worth the effort?

To be frank, it seems highly doubtful that any prospective client will ever contact me based on my TM-Town membership, if quantity of work is the main criterion.


 
Mirko Mainardi
Mirko Mainardi  Identity Verified
Italy
Local time: 11:20
Member
English to Italian
No need to take your ball and leave May 10, 2016

Kevin Dias wrote:

If you don't like it...

I've explained about how TM-Town works and our rationale for various features and mechanisms in great detail in this thread. If you don't like how it works, you don't have to use it. Even better, you are free to create your own web application that works the way you want it to work (whether that is using source words or not having a concept of "ranking").


Actually, you just unilaterally stated your point of view, over and over, but you are not willing to discuss and question it based on input from others, as your reply above clearly shows. But indeed, the ball is (or however was) yours. Thank you for your time.


At any rate, I still hope someone else will read and take into consideration the doubts and concerns expressed here by me and many other ProZ members when deciding how to move forward with this new ProZ+TM-Town combine...


 
Tom in London
Tom in London
United Kingdom
Local time: 10:20
Member (2008)
Italian to English
And yours May 10, 2016

Mirko Mainardi wrote:

Thank you for your time.

.....deciding how to move forward with this new ProZ+TM-Town combine...


And thank you for yours.

Jared has already made it clear in a previous post that those who pay the extra and get the TmTown thing will have no advantage over those who don't.

So I can't see what the point of it is.


 
Mirko Mainardi
Mirko Mainardi  Identity Verified
Italy
Local time: 11:20
Member
English to Italian
You're welcome ;-) May 10, 2016

Tom in London wrote:

Mirko Mainardi wrote:

Thank you for your time.

.....deciding how to move forward with this new ProZ+TM-Town combine...


And thank you for yours.

Jared has already made it clear in a previous post that those who pay the extra and get the TmTown thing will have no advantage over those who don't.

So I can't see what the point of it is.


I see what you did there but, anyway, if you're referring to this: http://www.proz.com/forum/prozcom_translator_coop/300979-complimentary_tm_town_membership_for_prozcom_members-page9.html#2547308 then it seems to me Jared only said that your "It does seem that unless you sign up for this TMTown thing, and pay extra, you will actually end up being disadvantaged in your Proz ranking" was incorrect... And indeed, as things stand, those are two separate rankings, and I don't think TM-Town membership is (currently) affecting directory rankings on ProZ...

However, since this also seems to be a situation which will inevitably evolve (somehow) in the near future, due to the fact that ProZ did acquire TM-Town, "the point is" exactly what I previously wrote, that is to say, the hope that "someone else will read and take into consideration the doubts and concerns expressed here by me and many other ProZ members when deciding how to move forward with this new ProZ+TM-Town combine", hopefully in order to avoid scenarios similar to the one you were referring to with your words I quoted above, and, possibly, to make TM-Town a useful, reliable, equitable and reputable tool for ProZ members (as opposed to a "competing" platform that could actually damage some of them...).

So... what was your point again?


 
Pages in topic:   < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16] >


To report site rules violations or get help, contact a site moderator:


You can also contact site staff by submitting a support request »

Complimentary TM-Town membership for ProZ.com members






Protemos translation business management system
Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!

The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.

More info »
Trados Business Manager Lite
Create customer quotes and invoices from within Trados Studio

Trados Business Manager Lite helps to simplify and speed up some of the daily tasks, such as invoicing and reporting, associated with running your freelance translation business.

More info »