Text on video sites

Recently, when I was watching videos on video sites, I noticed that text is displayed at the bottom of the screen. While I was watching one movie, I could understand the content, but I couldn't really explain the grammar so I asked one of the staff.

According to the staff, it was indeed grammatically incorrect. It was incorrect, but the text was "correctly" displayed.

Computer doesn't make any mistake. I have a mixed feeling about this...

Above is translation of an article "動画サイトのテキスト" dated October 16, 2017

Translation by Hiroko Matsuda 

Books I read: The 4-Hour Work Week

Title: Title:
The 4-Hour Workweek, Expanded and Updated: Expanded and Updated, With Over 100 New Pages of Cutting-Edge Content

Author: Tim Ferriss
Published by: Harmony; Expanded (2009/11/18)

A well known book. I read it a long time ago, but I hadn't written about it in this blog, so here it is.

The book introduces various ways to produce the same result but only work 4 hours a week. Countless methods of how to reduce time are being introduced. Just by reading the title, I thought 4 hours is just plainly impossible but as I read ahead, I started to think "I see I see!" on a lot of the points. 

By the way, there was news the other day on the professional baseball team that didn't make it to the A class this year going in to the fall camp. Apparently, it's an excruciating training and practice. By reading this book, I feel that the idea of cutting time is well received within the companies, but there is an impression that extensive practice is required to get the results in the sports industry. Rather than swinging the bat 1000 times a day, they should think about getting the same result by swinging 200 times per day or think of a practice regimen where practicing 3 hours a day would produce the same result as practicing for 8 hours a day. Even amateur players, for example, in high school baseball, they are always having morning practice, afternoon practice, evening practice, and after meal self-practice. Basically, the instruction method is to build the body for better performance, but I can't help but to think whether there is a better way.

I'm writing this is because I feel like the talk on cutting the time in terms of business is being done too much. I think cutting the time in sports should be the next topic and if someone writes about this, it would be an instant best-seller!

Above is translation of an article "読んだ本：The 4-Hour Work Week" dated November 22, 2017
Translation by Hiroko Matsuda

Translated> How many machine translation engines should we use? 20170924

I'm back from the Machine Translation Summit. Looking back at my notes, I wrote down things like these on post editing (well, typed on PC actually).

1. Domain specific machine translation engine is better.
2. It's possible to construct our own engine as long as corpus can be accumulated.
3. Specific operation details are trade secret.

In regards to 1, it's no use having machine translation aimed for general purpose. Accuracy will be higher with machine translation engine of specific field type. 

Regarding 2, corpus does not exist for language pairs other than En/Ja and Ja/En, so we need to construct our own. For latter pair, the server fee is becoming lower, and it's possible for smaller companies like ourselves to possess in-house server. In addition, with Japanese clients, there are many cases that clients cannot request projects if the server is not within the country.

For 3, much consideration is needed. All we need to do is to combine multiple machine translation engines and use the best one. Some CAT tools don't allow multiple APIs to be constructed, so that is one problem to be considered. Whether the server is located domestically or internationally should be publicized, but the combination of engines and how the translation is operated may be considered trade secret.

To be continued...

Translated> Time schedule for PSLT 2017 is up! 20170904

(This article has been originally written on September 4th, 2017)

For those attending the event, please feel free to come visit until the end.

The schedule is on here.

September 22, 2017

9:50-10:00 
Opening 

Session 1 (Chair: Katsuhito Sudoh) 

10:00-10:45
[Invited Talk] Bruno Pouliquen (World Intellectual Property Organization)
WIPO Translate: Patent Neural machine translation in 10 languages publicly available 

10:45-11:30
[Invited Talk] Masa Kajiki (MK Translation Firm)
Neural Machine Translation and Patent Translation from LSP Perspective 

[Lunch 11:30-13:10] 

Session 2 (Chair: Takehito Utsuro) 

13:10-13:55
[Invited Talk] Andrew Finch (Apple Inc.)
Statistical and Neural Models for Transliteration

Translated> Arrived at the Machine Translation Summit 20170921

(This article has been originally written on September 21, 2017)

I've finally arrived at the Machine Translation Summit being held at Nagoya University. The summit has been going on since the 19th. I'm finally participating on the 21st. Majority of seminars have already finished. I was a bit sad.

Today (21st), I left Osaka early in the morning and arrived Nagoya University around 9'oclock. I signed up for wifi when I signed in at the reception desk.

Received these items at the reception desk! Nagoya University bag!

I had lunch on campus. It was a lunch meal.

I went sightseeing, or rather walked around the campus. Truly, there was a memorial building since there were Nobel prize winner from the university. The entrance was free, so I checked it out.

At the event site, I saw one person I knew, so I chatted with him. In addition to translation company staff, there were many researchers from overseas. That's kind of obvious, though. People like me is probably a minority.

Unfortunately, a showcase called "Technology Showcase" has already finished the day before. I couldn't participate in the exhibition of each manufacturers.

I did not bring any paper notebooks. I input everything on google document via iphone. I brought a wifi-operating keyboard.

All I need is this and iphone.

I went back to the hotel as soon as today's session was over. This is because I need to practice for tomorrow's presentation. I brought some work with me too, so I did some work.

I can't explain well, but it's stimulating to be outside than being in the office.

I'll do my best tomorrow!

Translated> Finished presenting at The 7th Workshop on Patent and Scientific Literature Translation (PSLT 2017) 20170923

Thank you for those who participated!

This is where I made my talk.

Come to think of it, it was probably my first time doing a presentation in a university lecture hall.

It's pretty big!

Contents aside, I was happy that I was able to make some people laugh with the phrase "I will finish my presentation well before our lunch time" lol

About my presentation, afterwards, I was told "Bold move!" by one of the participants. I'll take it as a good thing...

There was catering.

I could have had my breakfast here!

Myself not knowing that you can have breakfast at the event site had my breakfast at the hotel.

I learned many things at this summit, but below are the main things I learned.

In neural machine translation of De/Ja and Ja/De, the translation might be done in the order of: De>En>Ja or Ja>En>De. Therefore, for Ja/De translation, some English is mixed in the translated German. This is something I heard from one person, there is no verification. He mentioned that it is possible that the language is translated to English first.

I also learned that domain-specific machine translation (domain-specific MT) works well. For example, in the MT developed by WIPO, sentences like "I am a boy" is not included in the corpus of the machine translation. This is because no such sentence will appear in patents. Therefore, when doing patent translation, MT designated for patent might be better. It seems that there will be a notification from WIPO around next week (I didn't hear this from any representative, so I will not publicize any contents)

Translated> Machine Translation Summit will be held from September 22nd at Nagoya University

From September 18th (Mon), 16th Machine Translation Summit will be held at Nagoya University. I'm extremely looking forward to it, but I'm participating from the 21st. My talk is on the 22nd.

I actually wanted to participate from the 18th, but it seemed that they wanted me to participate from the 21st...

I wanted to participate from the 18th because there are some interesting lectures and seminars. 22nd seems to be a workshop day, but I'm afraid that many people will go home since it's the last day of the event. One of my acquaintance who is participating from the beginning said he is leaving after the 21st.

I've never been to Nagoya University, so I'm excited to go there. Each university has its own atmosphere and I enjoy experiencing it. I'm thinking of going sightseeing in Nagoya on Saturday the 23rd.

By the way, the event I'm talking at is not the machine translation summit. It's The 7th Workshop on Patent and Scientific Literature Translation (PSLT 2017).

Finally submitted the abstract for The 7th Workshop on Patent and Scientific Literature Translation (PLST2017)!

Today, I finally submitted the abstract (created by staff and submitted by me).

The 7th Workshop on Patent and Scientific Literature Translation (PSLT2017) will be held on September 22 at Nagoya University (This article was originally written on August 10th). If you are around, please come join us!

Invited talks will be done by well known speakers from Apple and WIPO...

I'll most likely feel out of place, but I'll do my best nonetheless.

Here is what I'll be talking about:

Neural Machine Translation and Patent Translation from LSP Perspective

Masa Kajiki kajiki@mktranslaktionfirm.com

1. Abstract

Five descriptions of Neural Machine Translation from a translation company (LSP) perspective in the following order:

1) Description of the service (Hybrid Translation) based on neural machine translation.

2) How much faster the translation has become and the amount of editing needed for translations.

3) Problems regarding implementation of neural machine translation.

4) Possibility of Neural Machine Translation dramatically changing the business of translation companies.

5) Description of the company’s direction to end human translation by 2018 and utilize machine translation for all projects starting from 2019.

2. Introduction

According to some Internet news article, translation is an industry likely to disappear in the future. We will truly disappear if we continue translating in a conventional way. We cannot just stand and watch this happen. Just then, usage of Neural Machine Translation was permitted for commercial use for private translation companies. At this time, we decided to start a service utilizing the Neural Machine Translation. We have come up with a Hybrid Translation, which is a new service combining Neural Machine Translation, a cloud translation system and a post editor....