Forum Improvement Proposal: Add Automated Translation

[allancto]

Proposal and Benefit to RChain

The RChain Worldwide Forum https://rchaincommunity.xyz/ (#588 by @ICA3DaR5) supports discussion groups in many languages, and we wish to improve communication between our various language groups. Manual translation documents has in the past been a primary way of reaching out to speakers of other languages. Manual translation more suitable for many purposes than automatic translation with existing tools, but in our context automatic translation may be more desirable:

• Automatic translation goes in both directions. People from different language groups can share ideas, comments, and views of the same document.
• Automatic translation is fast, translation is available almost immediately.
• Automatic translation is cost effective: we can expect to translate every document and every response almost instantly and at very low cost.
• Automated translation can improve our mutual understanding. We will all be able to learn a little of each other’s words and typography, views and values, perspectives and wisdom.

From a software design perspective there will be several issues to consider.

1. UX and layout. How can we minimize the number of clicks, can readers set default preferences yet easily override the defaults when they want to?
2. Performance. Should documents be translated automatically and stored for rapid viewing?
3. The goal is to utilize existing services such as google translate and google captioning to do the actual translation. There should be no need at this time us to develop translation software, only to apply the existing services. Perfect translation is not the goal, ability to communicate with each other by incorporating existing tools is the goal.

Potential added features:

1. Indexing. How can we make it easy to point “into” a translated document? (by paragraph, special word, manually generated pointer?)
2. Video. Can automated video transcription be handled within this project? A big benefit would be to use words to “index” into RChain videos, which would enable or at least facilitate queries like “show me the video clip starting at the time where Mike asks about validators”. A related possibility would be to automatically or semi-automatically include speaker identification within videos. (@coininterview)
3. Search. Related to indexing, it would be awesome to have google-class search tools (indeed potentially based on google api’s) to look for relevant RChain documents within all of our repositories. (@eknir)

I’m attaching the Help Wanted label here, we do not have a team yet. We are seeking contributors in full stack web development, backend development and UX to accomplish these goals.

Objectives, Timeline and Budget

How will we measure completion? [coded, merged and tested in the Forum] Estimated Timeline Required to Complete the Task: [target 4-8 weeks] Estimated Budget of Task: [$8000-$13000 depending on features] Suggested Budget Valuation Model: [TBD] Retrospective remarks: [fill in AFTER completion of task and BEFORE final voting]

Legal

Task Submitter shall not submit Tasks that will involve RHOC being transacted in any manner that (i) jeopardizes RHOC’s status as a software access token or other relevant and applicable description of the RHOC as an “asset”—not a security— or (2) violates, in any manner, applicable U.S. Securities laws.

[zsluedem]

@allancto Automated translation?Do you mean something like google translations?

[allancto]

@zsluedem yes exactly, i explicitly mean google translate or equivalent. The translation and transcription "engines" already exist, not perfect but quite usable, we just need to apply them to make cross language communication easier.

[Viraculous]

@allancto said

Perfect translation is not the goal, ability to communicate with each other by incorporating existing tools is the goal.

It thinks considering this with respect to scale, automated translation is quite important. Communication is not all about getting the accurate word but transmission of usable information and for a community to cooperate at scale there is a need for communication aid like the automated translation which is cost efficient as well. @allancto checkup the link it's not going through.

[allancto]

@Viraculous that link was broken and now fixed, thank you! -@allancto https://rchaincommunity.xyz/

[allancto]

@all here is a link provided by @ICA3DaR5 which gives some samples of working within the Discourse framework, or similarly this.

Also a specific project around translation, "Translation Services Supported Microsoft Translator, Google Translate - Coming Soon" https://github.com/discourse/discourse-translator

@ICA3DaR5 thanks for these links. -@allancto

[David405]

@allancto I am concerned about the quality of automated translations......I think one of the major calls for concern of translated works over time has been the use of services such as google translate.

[deannald]

Perhaps I'm missing the point of this issue, but doesn't your browser already take care of this automated translation? I can reference any one of the pages produced and Google automatically translates the page (based on my preferences). Why are we reinventing the wheel here? What additional functionality are you proposing? If I wander into https://rchaincommunity.xyz/c/Italian-Italiano and click on any one of the articles posted there, I get automatic translation to English (or I right click "Translate to English"). Scaling this out to have storage (in how many languages?) and indexing/search just doesn't appear worth the benefit - better to add tags to articles.

This issue seems superfluous - I'd rather see development work happening on things where we get a lot more use of out it.

[deannald]

@David405 - I agree there are concerns with automated translations, but since this isn't providing advice of a critical nature (medical, emergency services, etc.) I think the use of automated translations works in our circumstances. But I know what you are saying - I was working on a chatbot for government services that translated from 70 different languages back and forth from English (using Azure Cognitive Services and AI) and the concern was that someone would ask for help and it would be a medical/police/public health emergency and it wouldn't be handle properly or the person would be given the wrong information or phone numbers, etc.

In RChain's case, I think it's safe to say that automatic translation is going to be fairly sufficient.

[dckc]

I see this has a Web Development label and no other issue does. Labels are for grouping issues. Until there's a pattern, a new label isn't worth the cost.

[ghost]

@deannald

Perhaps I'm missing the point of this issue, but doesn't your browser already take care of this automated translation? I can reference any one of the pages produced and Google automatically translates the page (based on my preferences). Why are we reinventing the wheel here? What additional functionality are you proposing?

For issue #588 I am going to use https://github.com/discourse/discourse-translator for convenience and user-friendliness, the budget for this won't be $8000-$13000, that would be crazy. @allancto probably has other things in mind on top of this which we need to discuss.

At the moment the Discourse Translator plugin uses Microsoft Translator with plans to include Google Translate soon.

[AyAyRon-P]

Are there any updates to this automated translation @allancto @ICA3DaR5 ?

[ghost]

@AyAyRon-P As of now I am trying to resolve an issue with the Microsoft Translator Text API, once resolved I'll post about it on issue #588

[allancto]

added WontFix because no devs stepped up and also wait to see outcome of @ICA3DaR5 investigation into existing plugin.

I will still support this issue if and when a great web developer steps in wanting to work on it. -@allancto

[AyAyRon-P]

Hi @allancto @ICA3DaR5

Are there any updates to this automated translation or can we close this?