Transcript Translation is very difficult

Translation in the 21st Century
Impacts of MT and social media on
language services
January 1954
Tower of Babel
Divide & Unite
Language as Instinct
Translation is very difficult …
October 2004
Reality Check
… TAUS founded.
Let a Thousand MT Systems Bloom!
Effectiveness of Data
Profit from sharing
(“Pirate’s Dilemma”)
MT is here to stay
Translation as a Utility
… as a Human Right
Five years ahead …
Technology in 5 Years
Hybrid Systems
Targeted Correction
(communities, games)
Real-time Training
TMs in 5 Years
ownership?
Cleaning
Semantic Clustering
Corpus Linguistics
Preserve Endangered Languages
Exciting new perspectives …
Profession in 5 Years
Choices
End to Repetitive Tasks
Productivity 5 to 10 Times Higher
Supplemented by Non-Professional Volunteers
Applications in 5 Years
Translation out of the Wall
Virtuous Circle
Spoken Translation
Industry in 5 Years
Thinking about drivers/trends
From TAUS Copenhagen Forum (May 2010)
Certain






Explosion in new content
Shift from text to text and
multi-media (word counts
go down)
Mobile user, hand held
devices
Real time/Just in time
demand
Cross-lingual translation
challenges
Balance of cost, timeliness
and quality
Uncertain



Open (collaborative) vs
Closed (competitive)?
Fee vs free?
Human vs Machine?
(incremental step or technology
breakthrough)
Industry in 5 Years
Machines
SWOT
Content disruption
Openassessment
Data
(Collaborative)
Innovation dilemma
Embedding technology
?
Human & Machine
Closed
(Competitive)
SWOT for Enterprise Language Service
S
W
• High leverage from TM
• Well established process and
management
O
• Opening new markets with MT
• Engaging with users & communities
• Convergence with video and speech
• Search engine optimization
• Translation of user generated content
• Quality inconsistent (local flavor missing)
• Lack of flexibility, reactive rather than
creative
T
• Rigid landscape (vendor lock-in)
• Not scalable to expand quickly
• Inability to ensure quality in new markets
• Lack of corporate awareness of new locales
Content Disruption
Sales
Localization industry
Web
“Battle for words”
UI
Manuals
Support
Social media
Knowledge Base
User generated content
New technologies
and solutions
Innovation Dilemma
S
• High leverage from TM
• Well established process and
management
O
•
•
•
•
•
Opening new markets with MT
Community/user feedback
Convergence with video and speech
Search engine optimization
Translation of user generated
content
• Quality inconsistent (local
flavor missing)
• Lack of flexibility (reactive, rather
than creative)
W
T
• Rigid landscape (vendor lock-in)
• Not scalable to quickly support new
markets
• Inability to ensure quality in new
markets
• Lack of corporate awareness of new
locales
Innovation Dilemma
S•
High leverage from TM
• Well established process and
management
O
•
•
•
•
•
Opening new markets with MT
Community/user feedback
Convergence with video and speech
Search engine optimization
Translation of user generated
content
• Quality inconsistent (local
W
flavor missing)
• Lack of flexibility (reactive, rather
than creative)
T
• Rigid landscape (vendor lock-in)
• Not scalable to quickly support new
markets
• Inability to ensure quality in new
markets
• Lack of corporate awareness of new
locales
Business Model Attributes
Old Model
1. One translation fits all
5. Word-based pricing
2. Project-based translation
6. GMS system
3.TM is core
7. Cascaded supply chain
4. One-directional
8.Translate-Edit-Proof
New Model
1. Quality differentiation
5. SaaS – Value-add
2. Continuous translation
6. MT embedded
3. Data is core
7. Community – user
4. Multi-directional
8. Post-edit – Real-time – Peer review
Enterprises in 5 Years
Need a Language Strategy
not just reducing word rates
“Dual Linguaspheres”
Enterprises in 5 Years
Machines
MT embedded.
Data is core.
Open Continuous translation.
(Collaborative)
Community/user.
SaaS + Value-add.
?
Project-based.
Cascaded supply chain.
Closed
(Competitive)
Word-based pricing (text).
GMS workflow systems.
TM is core.
Human & Machine
Strategic Roadmap
Machines
MT utility
SaaS + Value-add.
Closed
(Competitive)
Project-based.
Data is core.
Continuous translation.
Cascaded supply chain.
Word-based pricing (text).
Quality differentiation
One translation fits all
Community/user.
TM is core.
2011
2012
2013
Human & Machine
Open
(Collaborative)
Language Data Management
The Power of Data for Translation
1.
2.
3.
4.
Terminology mining and dictionary building
Customize automated translation
Global market and customer analysis
Quality management
in 5 Years
Industry Think Tank
& Innovation Partner
Research
Consulting
Events - workshops
www.translationautomation.com
in 5 Years
Industry Data Resources
Billions of Words
Translation Matching
TM Cleaning
Matching Scores
www.tausdata.org