Google and Its Smart Chinese Translation Neural Machine Thing

October 5, 2016

Google has a new neural translation system for Chinese. Read more here. It sort of works, but poetry is not its strong suit. Many Chinese student memorize Shi Jing’s “Cry of the Ospreys.” In Chinese, the first line of the poem is:


Google produces this translation of the line:

“Guan guanju dove, in the river of the continent.”


A standard English translation is:

Guan, guan, trill the ospreys, upon the island in the creek.

The standard English translation makes evident the sound of the ospreys from the island in the creek. Google sticks in a “dove” and dumps the island. Close enough for ospreys if not making the meaning clear to a non Chinese reader. Shi Jing is not around to offer an opinion which is probably a good thing.

Stephen E Arnold, October 5, 2016

Microsoft Changing Everything: At Least What Daesh Means in Redmond

August 30, 2016

I reported that Microsoft’s chief envisioning officer (I love that title) asserted that artificial intelligence will change everything. I pointed out that Microsoft has not been able to “change” China. Now Microsoft has learned that it cannot change the meaning of the word “Daesh,” which is one of the names of the Islamic State. I read “Bing Translates “Daesh” As “Saudi Arabia”, Angers Entire Kingdom.” The write up points out:

Bing Translation of “Daesh” the Arabic acronym for a global terrorist group backed by Kingdom of Saudi Arabia to “Saudi Arabia” has put the Microsoft Corporation in hot water with the Kingdom. Apparently, when the Arabic word


was typed into Bing Translate, the words “Saudi Arabia” would appear as the English translation, according to Khaberni. The so-called technical error caused an uproar in Saudi Arabia, where many Saudi social media users called for a boycott of Bing and Microsoft. The Microsoft Corporation has formally issued an apology to the Kingdom of Saudi Arabia, calling the error “unintentional”.

In what seems like the blink of an eye, Microsoft rolled out the bot which quickly learned to be somewhat interesting. The bot rolled away. Then Microsoft made Windows 10 Web cam hostile. Now Microsoft’s smart translation system has managed to anger the nation state Saudi Arabia. I assume Microsoft’s professionals anticipate smooth, seamless processing when entering the Kingdom from the USA. Now let’s think about the “change everything” statement. Doesn’t seem exactly correct, does it? How about some snap inspections of luggage to brighten one’s day? What’s the word for that? Sheesh? Oh, tay?

Stephen E Arnold, August 30, 2016

Statistical Translation: Dead Like Marley

June 16, 2016

I read “Facebook Says Statistical Machine Translation Has Reached End of Life.” Hey, it is Facebook. Truth for sure. I learned:

Scale is actually one reason Facebook has invested in its own MT technology. According to Packer [Facebook wizard’’], there are more than two trillion posts and comments, which grows by over a billion each day. “Pretty clearly, we’re not going to solve this problem with a roomful or even a building-full of human translators,” he quipped, adding that to have even “a hope of solving this problem, we need AI; we need automation.” The other reason is adaptability. “We tried that,” said Packer about using third-party MT, but it “did not work well enough for our needs.” The reason? The language of Facebook is different from what is on the rest of the Web. Packer described Facebook language as “extremely informal. It’s full of slang, it’s very regional.” He said it is also laden with metaphors, idiomatic expressions, and is riddled with misspellings (most of them intentional). Additionally, as in the rest of the world, there is a marked difference in the way different age groups communicate on Facebook.

I wonder if it is time to send death notices to the vendors who use statistical methods? Perhaps I should wait a bit. Predictions are often different from reality.

Stephen E Arnold, June 16, 2016

Be the CIA Librarian

May 3, 2016

Research is a vital tool for the US government, especially the Central Intelligence Agency which is why they employee librarians.  The Central Intelligence Agency is one of the main forces of the US Intelligence Community, focused on gathering information for the President and the Cabinet.  The CIA is also the topic of much fictionalized speculation in stories, mostly spy and law enforcement dramas.  Having played an important part in the United States history, could you imagine the files in its archives?

If you have a penchant for information, the US government, and a library degree then maybe you should apply to the CIA’s current job opening: as a CIA librarian.  CNN Money explains one of the perks of the job is its salary: “The CIA Is Hiring…A $100,000 Librarian.”  Beyond the great salary, which CNN is quick to point out is more than the typical family income.  Librarians server as more than people who recommend decent books to read, they serve as an entry point for research and bridge the gap between understanding knowledge and applying it in the actual field.

“In addition to the cachet of working at the CIA, ‘librarians also have opportunities to serve as embedded, or forward deployed, information experts in CIA offices and select Intelligence Community agencies.’  Translation: There may be some James Bond-like opportunities if you want them.”

Most of this librarian’s job duties will probably be assisting agents with tracking down information related to intelligence missions and interpreting it.  It is just a guess, however.  Who knows, maybe the standard CIA agent touts a gun to the stacks?


Whitney Grace, May 3, 2016
Sponsored by, publisher of the CyberOSINT monograph

Online Translation: Google or Microsoft?

March 1, 2016

HI have solved the translation problem. I live in Harrod’s Creek, Kentucky. Folks here speak Kentucky. No other language needed. However, gentle reader, you may want to venture into lands where one’s native language is not spoken or written. You will need online translation.

Should I forget Systran and other industrial strength solutions of yesteryear. Today the choice is Google or Microsoft if I understand “2 Main Reasons Why Google Translate Is Ahead of Microsoft and Skype.” (The link worked on February 22, 2016. If it does not work when you read this blog post, you may have to root around. That’s life in the zip zip world today.)

Reason one is that Google supports more languages than Microsoft. The total is 100 plus. The write up is sufficiently amazed to describe the language support of the Alphabet Google thing as “mind blowing.” Okay.

Reason two is that Google’s translation function works on smartphone. The write up points out:

You can hand-write, speak, type, or even take a picture of a given language and Google Translate will translate it for you. Not only this but on Android, some of the translation features are available offline. So, some features are accessible even if you do not have access to the internet.

The write up does not dig too deeply into Microsoft’s translation capability. If you are interested in Microsoft’s quite capable and useful services, navigate to the Microsoft Language Portal. Google is okay, but one service may not do the job a person who does not speak Kentucky requires.

Stephen E Arnold, February 27, 2016

When Google Translate Is Not Enough

September 16, 2015

I read a delightful article called “The British Library Is Crowdsourcing the Translation of a Mysterious 13th-Century Sword Inscription.” I am not too keen on edged weapons. Nevertheless, I am interested in becoming sharper when it comes to translation methods.

The write up states:

+NDXOXCHWDRGHDXORVI+ This inscription, engraved on a 13th-century double-edged sword owned by the British Museum, is the medieval mystery of the moment. Stumped by its cryptic engraving, last week the British Library tapped the interwebs for its crowd wisdom, asking commenters to help decode the meaning.

What makes the article entertaining is the fact that the British Library, backed with the formidable talents of British universities where linguistics absolutely thrives is turning to the hoi polloi for assistance.

And assist did the rustics. Consult the original article for the full span of human ingenuity. Here’s the comment I enjoyed from a non rustic:

“Everything is explained in Winnie the Pooh.”

A Google search reveals more questions:



Stephen E Arnold, September 16, 2015

Captain Page Delivers the Google Translator

January 15, 2015

Well, one of the Star Trek depictions is closer to reality. Google announced a new and Microsoft maiming translate app. You can read about this Bing body blow in “Hallo, Hola, Ola to the New, More Powerful Google Translate App.” Google has more translation goodies in its bag of floppy discs. My hunch is that we will see them when Microsoft responds to this new Google service.

The app includes an image translation feature. From my point of view, this is helpful when visiting countries that do not make much effort to provide US English language signage. Imagine that! No English signs in Xi’an or Kemerovo Oblast.

The broader impact will be on the industrial strength, big buck translation systems available from the likes of BASIS Tech and SDL. These outfits will have to find a way to respond, not to the functions, but the Google price point. Ouch. Free.

Stephen E Arnold, January 15, 2015

Machine versus Human Translations

January 7, 2015

I am fascinated with the notion of real time translation. I recall with fondness lunches with my colleagues at Ziff in Foster City. Then we talked about the numerous opportunities to create killer software solutions. Translation would be “solved”. Now 27 years later, progress has been made, just slowly.

Every once in a while an old technical cardboard box gets hauled out from under the car port. There are old ideas that just don’t have an affordable, reliable, practical solution. After rummaging in the box, the enthusiasts put it back on the shelf and move on to the next YouTube video.

I read “The Battle of the Translators: Man vs Machine.” The write up tackles Skype’s real time translation feature. Then there is a quick excursion through Google Translate.

The passage I noted was:

So, while machine translations may be great for rudimentary translations or even video calls, professional human translators are expert craftsmen, linguists, wordsmiths and proofreaders all wrapped in one. In addition to possessing cultural insight, they also are better editors who shape and perfect a piece for better public consumption, guaranteeing a level of faithfulness to the original document — a skill that not even the most cutting-edge machine translation technology is capable of doing just yet. Machine translators are simply not yet at the level of their chess-playing counterparts, which can beat humans at their own game. As long as automatic translators lack the self-awareness, insight and fluency of a professional human translator, a combination of human translation assisted by machine translation may be the optimal solution.

I include a chapter about automated translation in CyberOSINT: Next Generation Information Access. You can express interest in ordering by writing benkent2020 at yahoo dot com. In the CyberOSINT universe, machine translation exists cheek-by-jowl with humans.

For large flows of information in many different languages, there are not enough human translators to handle the work load. Machine based translations , therefore, are an essential component of most cyber OSINT systems. For certain content, a human has to make sure that the flagged item is what the smart software thinks it is.

The problem becomes one of having enough capacity to handle first the machine translation load and then the human part of the process. For many language pairs, there are not enough humans. I don’t see a quick fix for this multi-lingual talent shortfall.

The problem is a difficult one. Toss in slang, aliases, code words and phrases, and neologisms. Stir in a bit of threat with or without salt. Do the best you can with what you have.

Translation is a thorny problem. The squabbles of the math oriented and the linguistic camps are of little interest to me. Good enough translation is what we have from both machines and humans.

I don’t see a fix that will allow me to toss out the cardboard box with its musings from 30 years ago.

Stephen E Arnold, January 7, 2015

Remember Bing Translator?

November 13, 2014

Short honk: Microsoft offers an online translation service. It was called Bing once. That name has gone the way of the dodo. Details are here: “Bing Translator Picks Up an Update, Drops Bing Name and Adds Offline Translation for Vietnamese.” Just Bing it, but make sure you know the current name. Is this what MBAs learn today?

Stephen E Arnold, November 13, 2014

LinkedIn Enterprise Search: Generalizations Abound

November 11, 2014

Three or four days ago I received a LinkedIn message that a new thread had been started on the Enterprise Search Engine Professionals group. You will need to be a member of LinkedIn and do some good old fashioned brute force search to locate the thread with this headline, “Enterprise Search with Chinese, Spanish, and English Content.”

The question concerned a LinkedIn user information vacuum job. A member of the search group wanted recommendations for a search system that would deliver “great results with content outside of English.” Most of the intelligence agencies have had this question in play for many years.

The job hunters, consultants, and search experts who populate the forum do not step forth with intelligence agency type responses. In a decision making environment when inputs in a range of language are the norm for risk averse, the suggestions offered to the LinkedIn member struck me as wide of the mark. I wouldn’t characterize the answers as incorrect. Uninformed or misinformed are candidate adjectives, however.

One suggestion offered to the questioner was a request to define “great.” Like love and trust, great is fuzzy and subjective. The definition of “great”, according the expert asking the question, boils down to “precision, mainly that the first few results strike the user as correct.” Okay, the user must perceive results as “correct.” But as ambiguous as this answer remains, the operative term is precision.

In search, precision is not fuzzy. Precision has a definition that many students of information retrieval commit to memory and then include in various tests, papers, and public presentations. For a workable definition, see Wikipedia’s take on the concept or L. Egghe’s “The Measures Precision, Recall, Fallout, and Miss As a function of the Number of Retrieved Documents and Their Mutual Interrelations, Universiiteit Antwerp, 2000.

In simple terms, the system matches the user’s query. The results are those that the system determines containing identical or statistically close results to the user’s query. Old school brute force engines relied on string matching. Think RECON. More modern search systems toss in term matching after truncation, nearness of the terms used in the user query to the occurrence of terms in the documents, and dozens of other methods to determine likely relevant matches between the user’s query and the document set’s index.

With a known corpus like ABI/INFORM in the early 1980s, a trained searcher testing search systems can craft queries for that known result set. Then as the test queries are fed to the search system, the results can be inspected and analyzed. Running test queries was an important part of our analysis of a candidate search system; for example, the long-gone DIALCOM system or a new incarnation of the European Space Agency’s system. Rigorous testing and analysis makes it easy to spot dropped updates or screw ups that routinely find their way into bulk file loads.

Our rule of thumb was that if an ABI/INFORM index contained a term, a high precision result set on SDC ORBIT would include a hit with that term in the respective hit. If the result set did not contain a match, it was pretty easy to pinpoint where the indexing process started dropping files.

However, when one does not know what’s been indexed, precision drifts into murkier areas. After all, how can one know if a result is on point if one does not know what’s been indexed? One can assume that a result set is relevant via inspection and analysis, but who has time for that today. That’s the danger in the definition of precision in what the user perceives. The user may not know what he or she is looking for. The user may not know the subject area or the entities associated consistently with the subject area. Should anyone be surprised when the user of a system has no clue what a system output “means”, whether the results are accurate, or whether the content is germane to the user’s understanding of the information needed.

Against this somewhat drab backdrop, the suggestions offered to the LinkedIn person looking for a search engine that delivers precision over non-English content or more accurately content that is not the primary language of the person doing a search are revelatory.

Here are some responses I noted:

  • Hire an integrator (Artirix, in this case) and let that person use the open source Lucene based Elasticsearch system to deliver search and retrieval. Sounds simplistic. Yep, it is a simple answer that ignores source language translation, connectors, index updates, and methods for handling the pesky issues related to how language is used. Figuring out what a source document in an language with which the user is not fluent is fraught with challenges. Forget dictionaries. Think about the content processing pipeline. Search is almost the caboose at the end of a very long train.
  • Use technology from LinguaSys. This is a semantic system that is probably not well known outside of a narrow circle of customers. This is a system with some visibility within the defense sector. Keep in mind that it performs some of the content processing functions. The technology has to be integrated into a suitable information retrieval system. LinguaSys is the equivalent of adding a component to a more comprehensive system. Another person mentioned BASIS Technologies, another company providing multi language components.
  • Rely on LucidWorks. This is an open source search system based on SOLR. The company has spun the management revolving door a number of times.
  • License Dassault’s Exalead system. The idea is wroth considering, but how many organizations are familiar with Exalead or willing to embrace the cultural approach of France’s premier engineering firm. After years of effort, Exalead is not widely known in some pretty savvy markets. But the Exalead technology is not 100 percent Exalead. Third party software delivers the goods, so Exalead is an integrator in my view.
  • Embrace the Fast Search & Transfer technology, now incorporated into Microsoft SharePoint. Unmentioned is the fact that Fast Search relied on a herd of human linguists in Germany and elsewhere to keep its 1990s multi lingual system alive and well. Fast Search, like many other allegedly multi lingual systems, rely on rules and these have to be written, tweaked, and maintained.

So what did the LinkedIn member learn? The advice offers one popular approach: Hire an integrator and let that company deliver a “solution.” One can always fire an integrator, sue the integrator, or go to work for the integrator when the CFO tries to cap the cost of system that must please a user who may not know the meaning of nus in Japanese from a now almost forgotten unit of Halliburton.

The other approach is to go open source. Okay. Do it. But as my analysis of the Danish Library’s open source search initiative in Online suggested, the work is essentially never done. Only a tolerant government and lax budget oversight makes this avenue feasible for many organizations with a search “problem.”

The most startling recommendation was to use Fast Search technology. My goodness. Are there not other multi lingual capable search systems dating from the 1990s available? Autonomy, anyone?

Net net: The LinkedIn enterprise search threads often underscore one simple fact:

Enterprise search is assumed to be one system, an app if you will.

One reason for the frequent disappointment with enterprise search is this desire to buy an iPad app, not engineer a constellation of systems that solve quite specific problems.

Stephen E Arnold,November 11, 2014

Next Page »