Comments for Multidisciplinary perspectives http://donosborn.org/blog ... on agriculture, NRM, education, ICT, language & development Sun, 06 Jul 2008 19:32:02 +0000 http://wordpress.org/?v=2.3.3 Comment on Economics of language and the “long tail” effect by Kamusi: The Semantic Search Engine of African Languages - Appfrica http://donosborn.org/blog/2008/04/10/economics-of-language-the-long-tail-effect/#comment-76 Kamusi: The Semantic Search Engine of African Languages - Appfrica Wed, 02 Jul 2008 20:53:31 +0000 http://donosborn.org/blog/2008/04/10/economics-of-language-the-long-tail-effect/#comment-76 [...] has some big plans in mind. In his piece called “The End of English” he references the economics of language and loosely describes in perfect detail what makers of semantic web applications all over the world [...] […] has some big plans in mind. In his piece called “The End of English” he references the economics of language and loosely describes in perfect detail what makers of semantic web applications all over the world […]

]]> Comment on Paradigm shift on machine translation? by Don http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-75 Don Tue, 01 Jul 2008 21:59:17 +0000 http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-75 Thanks Job. Maybe I should use the term "watershed" (or "tipping point"?) in trying to describe what has been changing. From the view of a lot of users, apparently, there is a difference. Van de Meer's use of the term "revolutionary" in writing about changes in attitudes about the utility of MT is what had me thinking along the lines that at least some people's mindsets had shifted. Thanks Job. Maybe I should use the term “watershed” (or “tipping point”?) in trying to describe what has been changing. From the view of a lot of users, apparently, there is a difference. Van de Meer’s use of the term “revolutionary” in writing about changes in attitudes about the utility of MT is what had me thinking along the lines that at least some people’s mindsets had shifted.

]]> Comment on Paradigm shift on machine translation? by Don http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-74 Don Tue, 01 Jul 2008 21:54:30 +0000 http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-74 Thanks Andrew, I appreciate the feedback and info, I was not aware of Language Weaver's work on Hausa (among African languages they've also done something on Somali as well as Arabic- the latter to & from several languages). In general, though my impression is that outside of South Africa and in particular the Meraka Institute, and apart from a very few individual experts, there has been little attention to MT for African languages. The hurdles in terms of lack of corpora and in some cases lack of stable orthographies are part of the reason, but a commitment to finding a way to leverage HLT for African languages is the key need. I do think that a program focusing on supporting a new generation of African experts in NLP/HLT could yield benefits for the experts, the field, and most of all the diverse language communities of Africa. Thanks Andrew, I appreciate the feedback and info, I was not aware of Language Weaver’s work on Hausa (among African languages they’ve also done something on Somali as well as Arabic- the latter to & from several languages). In general, though my impression is that outside of South Africa and in particular the Meraka Institute, and apart from a very few individual experts, there has been little attention to MT for African languages. The hurdles in terms of lack of corpora and in some cases lack of stable orthographies are part of the reason, but a commitment to finding a way to leverage HLT for African languages is the key need. I do think that a program focusing on supporting a new generation of African experts in NLP/HLT could yield benefits for the experts, the field, and most of all the diverse language communities of Africa.

]]> Comment on Paradigm shift on machine translation? by Job van Zuijlen http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-73 Job van Zuijlen Tue, 01 Jul 2008 14:03:18 +0000 http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-73 The pursuit of useful rather than the ultimate high-quality machine translation is not new and a not a paradigm shift of the last few years. I worked as an MT researcher in the eighties and already at that time a more practical approach to MT development was prevalent. The pursuit of useful rather than the ultimate high-quality machine translation is not new and a not a paradigm shift of the last few years. I worked as an MT researcher in the eighties and already at that time a more practical approach to MT development was prevalent.

]]> Comment on Paradigm shift on machine translation? by Andrew Joscelyne http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-72 Andrew Joscelyne Tue, 01 Jul 2008 05:59:52 +0000 http://donosborn.org/blog/2008/06/30/paradigm-shift-on-machine-translation/#comment-72 Thanks for this useful summary. There are a number of interesting developments that are pushing MT into practice, rather than being the endless “holy grail” tech project with a 5-year window. • Statistical methods are speeding up engine production (no rules to craft manually), • the blogosphere and social networking generally are thrusting people’s noses against the language barrier in their everyday computing, • the open source community is getting to grips with simple and powerful tools to handle translation, • and large companies with global communication needs cannot get their stuff out quick enough with the laborious workflows found in commercial translation contexts, so they are looking closely at MT. On the language front, Asia Online is working closely on a number of Asian languages, India has a strong R&D base and seems to be developing solutions (not just PhD theses) for their language spread. Language Weaver (MT company) has a Hausa-English engine, and South Africa must surely be looking at EU extensive experience in using tools and tech to run a multilingual country. The key hope though is that communities will hijack the technology and push MT into new uses for new languages, so that we have a broad spectrum of experiments from which to select and build what Jaap van der Meer likes to call “translation out of the wall” – MT as a utility! Thanks for this useful summary.
There are a number of interesting developments that are pushing MT into practice, rather than being the endless “holy grail” tech project with a 5-year window.
• Statistical methods are speeding up engine production (no rules to craft manually),
• the blogosphere and social networking generally are thrusting people’s noses against the language barrier in their everyday computing,
• the open source community is getting to grips with simple and powerful tools to handle translation,
• and large companies with global communication needs cannot get their stuff out quick enough with the laborious workflows found in commercial translation contexts, so they are looking closely at MT.

On the language front, Asia Online is working closely on a number of Asian languages, India has a strong R&D base and seems to be developing solutions (not just PhD theses) for their language spread.
Language Weaver (MT company) has a Hausa-English engine, and South Africa must surely be looking at EU extensive experience in using tools and tech to run a multilingual country.

The key hope though is that communities will hijack the technology and push MT into new uses for new languages, so that we have a broad spectrum of experiments from which to select and build what Jaap van der Meer likes to call “translation out of the wall” – MT as a utility!

]]> Comment on Reflecting on “Computing’s Final Frontier” by Paradigm shift on machine translation? | Multidisciplinary perspectives http://donosborn.org/blog/2008/02/26/reflecting-on-computings-final-frontier/#comment-71 Paradigm shift on machine translation? | Multidisciplinary perspectives Tue, 01 Jul 2008 03:37:33 +0000 http://donosborn.org/blog/2008/02/26/reflecting-on-computings-final-frontier/#comment-71 [...] and in its potential for less-widely spoken languages, and having broached the topic on this blog once previously, I thought I’d take a moment to briefly review these articles. They are (links lead to [...] […] and in its potential for less-widely spoken languages, and having broached the topic on this blog once previously, I thought I’d take a moment to briefly review these articles. They are (links lead to […]

]]> Comment on Distant rumblings by Don http://donosborn.org/blog/2008/05/19/distant-rumblings/#comment-70 Don Mon, 23 Jun 2008 11:11:42 +0000 http://donosborn.org/blog/2008/05/19/distant-rumblings/#comment-70 Hi Mackenzie, The aftershocks continue there, though you don't hear about them in the news. A couple of times I've been on the line recently to Chengdu and they say that there was just another tremor. Re the Bisharat site, I'm actually revising the list of links as part of a major revision of the site - just haven't had the time to finish it yet. If there are questions re specific items, let me know. The links at the <a href="http://www.panafril10n.net/" title="PanAfriL10n.org" target="_blank" rel="nofollow">PanAfrican Localisation project site</a> are much more up to date. Hi Mackenzie, The aftershocks continue there, though you don’t hear about them in the news. A couple of times I’ve been on the line recently to Chengdu and they say that there was just another tremor.

Re the Bisharat site, I’m actually revising the list of links as part of a major revision of the site - just haven’t had the time to finish it yet. If there are questions re specific items, let me know. The links at the PanAfrican Localisation project site are much more up to date.

]]> Comment on Distant rumblings by Mackenzie http://donosborn.org/blog/2008/05/19/distant-rumblings/#comment-69 Mackenzie Thu, 19 Jun 2008 19:40:23 +0000 http://donosborn.org/blog/2008/05/19/distant-rumblings/#comment-69 O, Don, wish I had connected before our trip to China. We went through Chengdu like five days before the earthquake. Also want to thank you for keeping the pages on Bisharat from web sites that no longer exist. :) Really helping with my doc work! O, Don, wish I had connected before our trip to China. We went through Chengdu like five days before the earthquake.
Also want to thank you for keeping the pages on Bisharat from web sites that no longer exist. :) Really helping with my doc work!

]]> Comment on Economics of language and the “long tail” effect - part 2 by Don http://donosborn.org/blog/2008/04/17/economics-of-language-and-the-%e2%80%9clong-tail%e2%80%9d-effect-part-2/#comment-50 Don Thu, 01 May 2008 19:32:54 +0000 http://donosborn.org/blog/2008/04/17/economics-of-language-and-the-%e2%80%9clong-tail%e2%80%9d-effect-part-2/#comment-50 Belated thanks on this Martin. I'll reply there later (need to recover a forgotten password). I think a common issue from the perspectives of both articles is whether information technology reduces the pressure to abandon the languages out on the "long tail" by reducing costs and facilitating use of diverse languages. The impact of this on a language like English at the "head" of the distribution may not be that significant. The role of the head of the distribution with regard to the tail is an interesting one to explore - and indeed, in looking at the long tail in commerce, Chris Anderson considers the head to be essential to successful sales down the tail. There is not necessarily a competition, and in fact the opposite given that people buy more than one of things like books and recorded music. People attracted by a bestseller may pick other less-known products that strike their interest, especially when the transaction costs (to seller and buyer) are reduced enough to make them available. To a certain significant degree, languages of wider communication (LWCs) with lots of speakers (those at the head of the long tail distribution) have an advantage for learners. But learning, say English, does not mean needing to forget a less widely spoken language. On the other hand, speaking or learning a "bestseller" language like English doesn't appear to lead to study of less widely spoken languages. Whether there is any interrelationship between learning at one part or another of the distribution in the way there is apparently in the case of marketing best-sellers and low-volume specialty titles is a question that I'll try to come to later with reference to Louis-Jean Calvet's discussion of language "constellations." Belated thanks on this Martin. I’ll reply there later (need to recover a forgotten password).

I think a common issue from the perspectives of both articles is whether information technology reduces the pressure to abandon the languages out on the “long tail” by reducing costs and facilitating use of diverse languages. The impact of this on a language like English at the “head” of the distribution may not be that significant.

The role of the head of the distribution with regard to the tail is an interesting one to explore - and indeed, in looking at the long tail in commerce, Chris Anderson considers the head to be essential to successful sales down the tail. There is not necessarily a competition, and in fact the opposite given that people buy more than one of things like books and recorded music. People attracted by a bestseller may pick other less-known products that strike their interest, especially when the transaction costs (to seller and buyer) are reduced enough to make them available.

To a certain significant degree, languages of wider communication (LWCs) with lots of speakers (those at the head of the long tail distribution) have an advantage for learners. But learning, say English, does not mean needing to forget a less widely spoken language. On the other hand, speaking or learning a “bestseller” language like English doesn’t appear to lead to study of less widely spoken languages.

Whether there is any interrelationship between learning at one part or another of the distribution in the way there is apparently in the case of marketing best-sellers and low-volume specialty titles is a question that I’ll try to come to later with reference to Louis-Jean Calvet’s discussion of language “constellations.”

]]> Comment on Economics of language and the “long tail” effect - part 2 by Martin Benjamin http://donosborn.org/blog/2008/04/17/economics-of-language-and-the-%e2%80%9clong-tail%e2%80%9d-effect-part-2/#comment-49 Martin Benjamin Mon, 21 Apr 2008 22:05:15 +0000 http://donosborn.org/blog/2008/04/17/economics-of-language-and-the-%e2%80%9clong-tail%e2%80%9d-effect-part-2/#comment-49 Thanks, Don, for starting an intriguing discussion. I have posted in response (more a tangent than a reply) here: http://www.kamusiproject.org/en/node/151 Cheers, Martin Thanks, Don, for starting an intriguing discussion. I have posted in response (more a tangent than a reply) here: http://www.kamusiproject.org/en/node/151

Cheers,
Martin

]]>