3 ms·
Compare this to one of Google's blog post promoting their MT research: https://research.googleblog.com/2016/09/a-neural-network-for-machine.html https://researc
by anewhnaccount2 9y ago
Compare this to one of Google's blog post promoting their MT research: https://research.googleblog.com/2016/09/a-neural-network-for-machine.html https://research.googleblog.com/2016/09/a-neural-network-for...
It is:
1) More accurate, compared to hyperbole like e.g. "Bridging the Gap between Human and Machine Translation" we have right there in the title the domain: news.
2) A more impressive result. This result is on an independently set up evaluation framework, compared to Google's which used their own framework.
Compare further the papers:
https://arxiv.org/pdf/1609.08144.pdf https://arxiv.org/pdf/1609.08144.pdf
https://www.microsoft.com/en-us/research/uploads/prod/2018/03/final-achieving-human.pdf https://www.microsoft.com/en-us/research/uploads/prod/2018/0...
These researcher appear to have been much clearer about what they're actually claiming, and also used more standard evaluation tools (Appraise) and methodology rather than something haphazardly hacked together.
- snovv_crash 9y agoThe outputs of Microsoft Research are really good. At least in my field, it is one of the few places where if they published something you can be sure of being able to reproduce the results using only what is described in the paper, no secret sauce required.
- londons_explore 9y agoGoogles was nearly 2 years ago though... Thats a long time in this field
- golfer 9y agoIf you are claiming that Microsoft is pure as the driven snow with regard to making exaggerated and hyperbolic claims, then you clearly know little about Microsoft.
- gumby 9y agoI don't read that claim being made. Microsoft Research has been generally left alone to do good computer science, in which "left alone" has included "by the marketing department"
- Hydraulix989 9y agoThat conclusion doesn't logically follow.
- romaniv 9y agoMicrosoft Research also seems to be much more willing to create usable demos of their work.
- Eridrus 9y agoI haven't looked at this work, but I found the write up of this work pretty misleading: https://www.microsoft.com/en-us/research/blog/microsoft-researchers-achieve-new-conversational-speech-recognition-milestone/ https://www.microsoft.com/en-us/research/blog/microsoft-rese... The issue with that evaluation was that machines were much better at not making some trivial mistakes humans don't care about (eg transcribing umm, err, etc), but were more likely to get the meaning wrong. Kudos to MS for doing the error analysis and publishing that info, but I found the reporting of it misleading.