Chi usa i Bot?

21件のコメント

• 2
  
  carla m. 2021年04月29日 08:36

  Ciao Elena,

  ho notato anche io almeno un traduttore che ieri ha preso uno dei lavori in tedesco per poi fare scadere il tempo.

  Se hai fatto una ricerca nel forum avrai visto che il problema/dubbio riguardo i bot è stato sollevato varie volte e in varie combinazioni linguistiche in questi anni. Il team ha affermato di poter individuare e prendere provvedimenti contro i traduttori che usano bot, ma il dubbio che varie persone continuino a usufruirne rimane... Forse dei traduttori sono stati eliminati ma a noi non è dato sapere.

  Hai scritto anche al support? Immagino di sì.

• 1
  
  marcodnd 2021年04月29日 08:49

  Lo avevo notato anche io, ma potrebbe anche trattarsi del bug segnalato il 16 aprile: "It's been brought to our attention that some translators are currently experiencing a bug that prevents them from submitting a collection ... ".

• 2
  
  carla m. 2021年04月29日 09:03

  Non so, ma ne dubito. Come dice Elena, un traduttore vede subito che il testo non è nella lingua della propria combinazione, quindi non dovrebbe neanche provare a tradurre. Il bug che tu citi rallenta il salvataggio e/o invio, qui è questione di rifiutare immediatamente. Spero che il team risolva entrambi i problemi al più presto ;-)

• 1
  
  marcodnd 2021年04月29日 09:26

  Sì, ma io avevo comunque notato una maggiore incidenza di questi "exceeded alloted time" proprio qualche giorno prima che fosse segnalato il bug, quindi sarebbe una strana coincidenza.

  Quei lavori in tedesco sono comunque stati presi da decine e decine di traduttori, presumo per la maggior parte 'umani', per poi essere abbandonati immediatamente. Immagino che molti facciano clic immediatamente su Start Translating senza neanche guardare il testo. Alcuni avranno riscontrato il bug di cui si parla e hanno fatto scadere il tempo.

  By the way, ma quanto ci vuole a togliere questi lavori in tedesco? Ancora li vedo, dopo quasi 24 ore.

• 2
  
  Elena F 2021年04月29日 09:27

  @Carla, sì, ho già segnalato. Spesso mi capita che la mattina "acchiappo" lavori piccoli che durante la notte sono stati ostaggio di due o tre persone! Sono sempre gli stessi 4 utenti, ormai ho imparato gli ID. Sarebbe facile tenere traccia di questi comportamenti e mandare avvertimenti automatici agli utenti per poi sanzionarli se continuano.

  @Marco non credo si tratti di quel bug, la funzione Decline funziona benissimo e non ha senso che una persona tenga un lavoro in una lingua sbagliata per 4 ore

   

• 3
  
  Katrina Paterson 2021年04月29日 09:27

  Hello everyone, I'm Katrina, your Community Experience Associate. I'm chipping into the discussion in English because I don't know Italian (though I can see from your words that it's a very beautiful language). 

  Elena, thank you for sharing your experiences and also for your thoughts about improvements that could be made to our system. We always welcome any and all feedback, and I'll be sure to pass your comments on to the relevant team. 

  As Carla suggested, if you ever notice what you believe is unusual activity, we would always strongly recommend that you contact the Support team about this (support@gengo.com), since they will be the best equipped to look into specific cases. In addition to this, any information that you or any other translator provides them with is valuable, since this helps us to monitor the system as a whole. Of course, all of you are welcome to share your thoughts here as well (provided that you don't reveal sensitive information about other translators, or customers).

  In terms of the bug issue that Marco mentioned - we're aware of this and currently looking into it. 

  I hope these comments have helped! Thanks again to all of you for your feedback :) 

• 2
  
  Elena F 2021年04月29日 09:30

  @marco abbastanza sospetto, il pulsante decline funziona, perché restare bloccati 4 ore con un lavoro sbagliato??

  Se non vuoi vederli più questi lavori in tedesco, accettali e poi fai decline, così spariranno dalla tua vita :)

• 2
  
  Elena F 2021年04月29日 09:33

  @Katrina thank you for your fast reply. I have contacted support about this matter a month ago. I wish you could implement an automated system that warns users that keep holding jobs without working on them.

• 1
  
  Katrina Paterson 2021年04月29日 11:38

  @Elena - no problem! We'd always say that it's worth contacting Support again anyway, providing them with as much specific information as you can about the unusual activity, including language pair and translator ID. The more information they have, the better they'll be able to look into the particular situation that you describe, and the better they'll be able to monitor the system as a whole, since reporting the situation will then create a record of it. We really do appreciate your suggestions for implementing an automated system, and I've passed that feedback further along, so when time and resources allow for it, this may be something that we can look into. 

  If you have any other concerns, please don't hesitate to reach out to me directly at katrina.paterson@lionbridge.com

  Thanks again for all of your ideas and for sharing your experience with the forum. 

• 2
  
  Elena F 2021年04月29日 11:50

  Hi Katrina, unfortunately the job is no longer available for me, so I can't see any details at the moment. If Gengo wants to investigate further it's pretty easy to find the jobs I'm talking about as they are EN-IT jobs flagged as wrong language by many translators in the past couple of days.

  Glad to know you will take my suggestions into consideration!

   

• 3
  
  dunkel 2021年04月29日 14:27

  Hi Katrina,

  I already reported something like this in the past (a "translator" exceeding allotted time - most probably a bot, as suggested by Elena). Contacting the support to let them know about those issues is not a big deal, so I will do it again if needed. But this makes me wonder: Gengo does not have a system to automatically detect them?

  Or, speaking about automatic control systems, a way to give a prompt heads-up to the team when a project is flagged over and over again? Those collections in the wrong language pairs mentioned before kept on bouncing back and forth for hours. I had to accept and refuse them to stop the continuous chiming of the feeder.

• 1
  
  Katrina Paterson 2021年04月29日 15:01

  Hi Dunkel, thanks for your comments. At the moment, we look at unusual translator activity on a case-by-case basis, which means that if any of our translators contact us to report seeing something strange, we then look into the issue and act accordingly. As you and Elena have mentioned, the idea of creating a system to automatically detect unusual activity is a good one, and I have passed this feedback along. For now, the best thing to do if you have doubts about another translator's activity is to make a note of that person's translator ID and contact Support with that particular piece of information, and with any other details that may be helpful. The more specific information we have, the better we are able to investigate situations like these. (This goes for all language pairs, not just English to Italian.)

  Dunkel, your second point about creating an automatic system to alert us when a project has been repeatedly flagged is also very interesting and I'll be sure to forward that suggestion on, too. Could you tell us a little more about how such a system would work?

• 3
  
  dunkel 2021年04月29日 15:19

  Dear Katrina,

  Thank you for your reply and for your openness. However, I must admit that the question Could you tell us a little more about how such a system would work? is leaving me a bit puzzled - then what is the purpose of flagging a collection in the first place?

• 1
  
  Katrina Paterson 2021年04月29日 15:29

  Hi Dunkel, sorry if I wasn't very clear there. So basically, do you mean that the automated system that you are proposing would alert us automatically if a translator flagged a job for any reason (wrong source text, not enough context, unsuitable content, etc.) and then once this job had been flagged, it would not appear on the dashboard again?

  Katrina Patersonにより編集されました 2021年04月29日 15:30
• 1
  
  dunkel 2021年04月29日 15:50

  Well, I supposed that flagging a job would already send alerts to relevant Gengo team members, in order to let them know there is something wrong (shouldn't things work like this?).

  However, it seems like the staff is unaware of collections flagged, otherwise they would not stick around for hours. And please allow me to add that, sooner or later, somebody will probably take a chance and try to translate them even if they have issues. So, it is in Gengo's interest to remove those jobs promptly, also not to keep the customer waiting.

  Of course, it would also be a big plus if flagging a job means not to see it anymore in our dashboard.

• 1
  
  carla m. 2021年04月29日 15:52

  Hi Katrina, I was under the impression that flagging a job would automatically signal to the team that something is wrong with a collection, so that they can fix it/alert the customer/take it off the dashboard if necessary. If this is not the case, I can only repeat what Dunkel already said: "what is the purpose of flagging a collection in the first place?".

• 1
  
  Katrina Paterson 2021年04月30日 06:58

  Hi Dunkel and Carla, thanks for your comments. So, the way it works is that once a translator flags a job (for whatever reason), it will not be automatically cancelled BUT the Support team will be alerted and they will then look into the the job and decide which action to take. While this process is taking place, the job will continue to appear on the system and be available to translators. I hope this clarifies things. 

• 2
  
  dunkel 2021年04月30日 07:53

  Hello Katrina, thank you for your clarification. Then I can only suggest a simple but much needed improvement - the Support team should intervene more quickly. Unfortunately, flagged jobs tend to stick around for hours, which could be detrimental for both customers (long delivery times) and possibly quality (as somebody could decide to take those projects nonetheless).

• 2
  
  carla m. 2021年04月30日 08:21

  Thanks Katrina. Sure, I don't expect the immediate cancellation of a job at the first flag, but as dunkel says we're talking about jobs lingering for hours/half a day on the dashboard, and being there even after two or three flags and up to 30 translators accepting and rejecting them (yep, I had time on my hands and checked until I finally decided to reject the jobs myself in order to make them disappear...). We know that's not the only problem the team has to deal with: the bug that slows down collections is even more annoying, and the eternal doubt that some people are using bots - which is what this thread was really about - is extremely annoying :-). However, fixing it would be a big plus for all the people involved (team/clients/translators). Have a good day everyone!

• 2
  
  Katrina Paterson 2021年04月30日 10:24

  Hi Dunkel and Carla, thanks also to both of you for your thoughts. 

  So just to sum up everything that we've been saying - now that we've established how the job flagging process happens, and clarified that each job that is flagged is investigated, it seems that both of your concerns are that these jobs are staying on the system for a while despite being flagged multiple times. I think that this ties back to Dunkel's previous comment from earlier in the conversation, in reference to the suggestion about creating 'a way to give a prompt heads-up to the team when a project is flagged over and over again', and I do apologise if I confused things by querying this comment. 

  We really do appreciate both of you sharing these thoughts, and we will certainly make sure to pass them on. Carla, you're right that the conversation has shifted a little from the original topic of the bots, but in any I hope that it's been useful for everyone, and I'd always emphasise that we're here to listen to what you have to say about any topic. 

  Thanks again to all of you that have shared their thoughts on this thread, and I wish everyone a good day too.

• 0
  
  captnemo 2022年03月13日 17:17

  Tedesco >> Italiano ??