"Approximately 8 of the 319 million people in the United States read the Wall Street Journal, about 2 percent of the population. If you look at the language — standardized English — being fed into many natural language processing units, it’s based on the language of that 2 percent. "
It's hard to take an article seriously when it opens with a logical fallacy. Yes, the WSJ uses standard English and only 8 million people read (subscribe to?) it. That does not mean that only 8 million people in the US use so-called standard English. There are also a lot of people who don't subscribe to that particular publication but still speak or write in "standard" English and that group is much larger than the group of WSJ readers.
Then why bring up WSJ at all? What about all the other papers? and what about all the other sources of training data for NLP research?
WSJ dataset is not the only such dataset. NLP is not very good with standard English yet and usually doesn't generalize from topic to topic. Dialects and other languages - especially those without formal rules - will come when we can deal with standard English.
I'm looking for formal rules. I'm not saying AAVE doesn't exist, I'm saying its rules (such as they are) are not formalized. Tons of rule books exist for Standard English and part of learning it is to memorize the correct rules. I don't believe any formal equivalent exists for AAVE.
What, then, are your parameters for formal? It has to be in a book, codified and signed off on? If so, aren't these kind of antithetical to AAVE on the face of it?
> It has to be in a book, codified and signed off on?
The point is for rules to be formal then the must be formalized somehow and codified. This could be online or in a book, but the point is that there must be some clear delineation between when the rules are followed and when they are broken. Otherwise what's the meaning of "formal" rules?
> If so, aren't these kind of antithetical to AAVE on the face of it?
Yes. My argument is that AAVE, almost by definition, is an informal dialect without formalized rules.
This is getting rather off topic, but I think it relates to the original point of why NLP might start with Standard English even if you are not biased. A large corpus of Standard English text (such as from the WSJ) will generally be very internally consistent precisely because it follows a set of formal rules codified into a style guide. As there is no such equivalent for AAVE, even gathering a large and internally consistent corpus of AAVE text seems prohibitively difficult. That being said, I do hope researchers are working on gathering text from Twitter to build up new training sets.
The point is for rules to be formal then the must be formalized somehow and codified
So every form of English is an "informal dialect" then? Because this ain't French with the Académie publishing strict rules for use of the language. Do you also say that the languages of remote Amazon tribes aren't "real languages" because they don't have a formal government body publishing written rules?
Or do you just want to bash on AAVE and are grasping at straws for reasons why?
> So every form of English is an "informal dialect" then?
Yes, the majority of spoken English does not follow the rules of Standard English. Pretending that such rules don't exist is willful ignorance though: the WSJ obviously write a more formalized version of English than teenagers do in text messages.
> Do you also say that the languages of remote Amazon tribes aren't "real languages" because they don't have a formal government body publishing written rules?
Nowhere did I say that AAVE is "not a real language" because it's less formalized than Standard English. Prior to spelling reforms, English itself was extremely inconsistent and informal, but I certainly don't pretend that it wasn't a language.
> Or do you just want to bash on AAVE and are grasping at straws for reasons why?
I'm not trying to bash AAVE. In fact, I'd even posit that the reason AAVE isn't more codified is perhaps because of racial bias which treated it simply as "incorrect" English instead of a separate dialect worthy of formalization. Pretending that all languages are equally formalized is simply willful ignorance though.
"Formal rules", in the context you've chosen to speak in, are defined by this upstream comment:
> NLP is not very good with standard English yet and usually doesn't generalize from topic to topic. Dialects and other languages - especially those without formal rules - will come when we can deal with standard English.
The rules you're talking about, that get printed in books and studied, are not linguistic rules. Crucially, this means they are not widely observed in printed standard English, which in turn means they can't be relevant to training a language model to understand printed standard English.
The "formality" you seem to want to talk about has no place in this discussion. It is not relevant to any language. gordonguthrie is correct to point out that the assumption lqdc13 is trying to make is false. You are wrong to contradict him using a meaning of "formal rules" that you brought to the conversation yourself. It had a meaning -- a completely unrelated meaning -- before you showed up.
> Crucially, this means they are not widely observed in printed standard English, which in turn means they can't be relevant to training a language model to understand printed standard English.
I agree that they're not widely observed in written English, but they are consistently observed in the WSJ, which was the origin of this entire debate.
As lqdc13 pointed out, NLP still isn't even consistently good at understanding standard English. One could reasonably posit that that's due to the inherent ambiguity and inconsistency of most writing and that focusing on a narrower, standardized document corpus (the WSJ) you could get better initial results. What, exactly, is controversial about that? Do you really think that the language of the WSJ is no more consistent and formalized than the language of Twitter users?
Thanks for the link. I found the regional breakdowns interesting.
That being said, I think this research emphasizes that AAVE is anything but standardized. That's not meant as a pejorative statement: it's just acknowledging that, like most languages in history, AAVE has not gone through a process of codification and standardization to formalize it.
Sure, but formal rules are not equivalent to a prescriptive grammar. AAVE has formal, consistent rules, described by linguists. You can make mistakes in AAVE just as in standard English (see: "African American Vernacular English Is Not Standard English with Mistakes" https://web.stanford.edu/~zwicky/aave-is-not-se-with-mistake... )
like most languages in history, AAVE has not gone through a process of codification and standardization to formalize it
I'm not sure what you're getting at - prescriptive grammars of the codified form you're describing are the products of their political and economic circumstances. There isn't some Hegelian trajectory of linguistic validity, where all variants aspire towards legalism.
Formality, if present, can be equally derived from observation and description of usage, using the linguistic analysis of the kind found in the articles I linked above. Language, when viewed in total, has shared elements and patterns that supersede the narrow focus of prescriptive impositions. It's why linguists can observe things such as that the dropped copula exists in both AAVE and Russian.
"I read the Wall Street Journal, about 0.00000031 percent of the population. If you look at the language — standardized English — being fed into many natural language processing units, it’s based on the language of that 0.00000031 percent."
Equally true, right? But somehow it seems just a little misleading.
It's hard to take a poster seriously when they show willful ignorance of the way language serves as a marker of class, ethnicity and race, and instead have to manufacture quibbles to try to discredit an article that might cause slightly less comfort with the notion of the Glorious AI Future™ -- Coming Soon™!
> When researchers evaluated their model against natural language processing tools...researchers found that the software flagged African-American English as “not English” at a much higher frequency than standard English.
Likewise, notice the deceptive switch between "English" and "standard English." So basically language bots trained to identify standard English are correctly identifying non-standard English dialects as not standard English.
If the training corpus for a machine learning model contains stereotypes and biases, then the output of the model will reflect those prejudices.
When that model is used in large-scale applications, it will not just repeat those biases, it will amplify them. It propagates biased language, which feeds back into the model in a self-reinforcing vicious cycle of increasing bias.
Should we attempt to detect prejudices in machine learning models and actively counter them, creating a self-reinforcing virtuous cycle of balance and tolerance?
----
As an aside, for an article about language, this one has a lot of typos. Doesn't anyone proofread before posting?
Especially as it relates to this: "This means that blogs or websites that employ African-American language could actually be pushed down in search results because of Google’s language processing."
Isn't the ideal system to not assume anything about the content it's analyzing, but to base the model on behavioral patterns instead? In the case of Google ranking, either a site has traffic/low bounce/backlinks/social cred or it doesn't--why is Google attempting to 'read' and analyze the content itself? (Other than to serve ads of course)
> Isn't the ideal system to not assume anything about the content it's analyzing ...
Because Google no longer just searches for literal strings of text, ranked by links.
They now infer your meaning and find results that match your intent -- even if the exact search string doesn't exist in the search results.
For more on this, read about Hinton's work on "thought vectors" [1].
So if Google can't determine the meaning of African American language as well as it can standard English, then that content is likely to be less visible in search results.
I can't say I would think that Google does a particularly good job at interpreting the intent of my search queries. I often need to reformulate them to get what I'm after. But I doubt that my searches are similar to those of the general public so that is understandable.
It's even worse than that - it explicitly ignores quoted keywords in many cases. The only intent they care about is what guides you towards their interests.
Thanks, I did not know that was an option. It should make a few different queries quite a lot nicer. I was starting to get scared that I was wishing for the old Alta Vista.
But if I searched in the dialect of the content I would get the content. This is simply segregating the internet. Not "pushing down" rankings.
As an aside - this is a slippery slope conversation. Some people believe language requires rules. Other, more special people, think language is fluid and anarchist. The latter of those two do not make search engines.
For your aside, I don't think it's true. Search engine makers are well aware that language usage is what it is, and not what the thesaurus or grammar book says it is. Unfortunately, using machine learning to work out synonyms means you're going to miss out on minority usages, such as specialized academic jargon, or communities with millions of speakers which have slang and conversational grammar unlike "standard" English.
> Some aspects of communication are likely to prove more challenging, Hinton predicted. “Irony is going to be hard to get,” he said. “You have to be master of the literal first. But then, Americans don’t get irony either. Computers are going to reach the level of Americans before Brits.”
Showing only (or prioritising) same-language results helps with relevancy. A German speaker who searches for “gift” is not looking for ideas for presents, and an Italian speaker who searches for “peperoni” is not looking for the American variant of salami.
A further problem, though, is that search engines try to understand the query and the content searched at a deeper level than mere keyword matches. For example, they may try to pick up synonyms (if I search for “movies in <local area>”, perhaps results for “films” might be helpful too) and associations.
> why is Google attempting to 'read' and analyze the content itself?
So that when you search for "cheesecake recipes", you don't get a well-sourced, high-traffic article from a respectable website describing in detail how the new business initiative by Cheesecake Factory is a recipe for disaster.
* Disclaimer: I pulled that example out of thin air. There's actually no such article.
> When that model is used in large-scale applications, it will not just repeat those biases, it will amplify them. It propagates biased language, which feeds back into the model in a self-reinforcing vicious cycle of increasing bias.
Which is precisely why social media (e.g. your newsfeed) is such an awful source for news, especially when you consider how low the standards are on digital publishing these days...
So humans do this too naturally - this is how new languages develop over hundreds of years and groups evolve and split. Aside from that, I'm a little nervous about seeing machines accelerate this process.
> If the training corpus for a machine learning model contains stereotypes and biases, then the output of the model will reflect those prejudices.
Depends what you mean by "the corpus". Let's take the example of loan applications. Maybe the training data contains credit scores, biographical information, and personal essays about the applicants from loan officers.
Let's say loan officers tend to use more negative language when talking about black people.
Any decent machine learning system given racial data will not adopt those biases. In fact, it will counteract them. It will determine that e.g. the semantic content of the essays is an indicator of loan suitability, and it will also notice if it has to e.g. add something to the essay score for black people to make the best decisions.
As long as the output data is objective, competent ML systems will account for any bias in the input data.
> Should we attempt to detect prejudices in machine learning models
We don't have to. If the output examples we are training on are true, the ML algorithm won't adopt any incorrect biases.
This is literally the point of machine learning. You want to make the best possible decision. If your algorithm has incorrect human biases, you're losing money. ML naturally accounts for this stuff.
I don't understand what you mean by "the algorithm" or "true relationships".
Do you know what I mean? There's truth as in what the data says, and there's truth as in what's underlying before social processes " corrupt " the truth. This paper examines ways to get at the latent truth.
Do you think that's a useful endeavor? It's dismissive to say it has anything to do with what the "author likes".
> and there's truth as in what's underlying before social processes " corrupt " the truth.
Ah, so you're defining "truth" as "whatever I imagine the world would be like if society worked the way I wished it did". There's another word for that; "fantasy".
> If the output examples we are training on are true, the ML algorithm won't adopt any incorrect biases.
This is rarely the case when working wtih real data, and thus inspecting whether our models are biased against protected classes is probably one of the most important things an ML practitioner should do.
We have a good understanding of the error profiles of most ML algorithms. We can put tight bounds on the difference between predictions and validation data. If ML algorithms make mistakes, it's usually due to noise or low-quality data, not the algorithm itself.
There is no reason an ML algorithm would be biased against a "protected class". It doesn't know what those are. It's possible that the algorithm will uncover a truth that you don't like, like that there as differences in risk profiles across race or gender, but that doesn't mean the algorithm is incorrectly biased. It just means that reality is at odds with how you might want it to be.
Which data are you referring to? In most cases, the training data isn't human-generated, and if it is, we usually want to match human behavior as close as possible.
Virtually all data used to predict crimes or recidivism is fraught with human bias, for example. Not sure that we want to reproduce the bias of criminal justice system in any prediction problem involving this type of data.
How is recidivism data biased? I'm sure that the information gleaned from parole officers and cops might be biased, but as long as the ML system is trained on whether or not someone actually reverted to committing crimes, it should be able to detect bias on the part of P.O.s and other functionaries and give a more accurate determination as to someone's chances of recidivism.
Since most languages have dialects of similar "standard" form divergence, it's an NLP problem that would have to be addressed anyway. While slang might be hard to keep up with, the grammatical patterns of AAVE are well-documented and understood by linguists, and there's plenty of data to model.
The issue of prejudice stems from refusing to treat AAVE as a valid input to an AI, mirroring the racism in American culture that similarly devalues and excludes it. Output can conform to some approximation of "standard" form without being problematic, following the pattern of interaction between standard and dialect speakers. As a speaker of the standard, it's gauche and derisive to imitate or refuse or comprehend a dialect speaker, not to respond in your own form.
Was reading the article and, in true wiki fashion, click the links that seem relevant to get more context. As is often the case I ended up in some other corner of the internet watching interviews with Stephen Pinker. Something in all that input triggered a thought when I read your comment:
What is really the difference between spelling a word wrongly and using a non-dictionary word to express something. We except one as "slang" but will criticize the other because they don't comply with established rules. Isn't that what slang is? A break from established rules.
Anyways, just a thought, because I agree with you; if you publish - proofread. I just find it interesting to throw a Socrates argument at myself from time to time.
Btw and somewhat of topic.. this was the Pinker interview that was most interesting:
This is the same question as 'What values should we impose on society?' And that is an irreducible conflict. It's politics.
Programming a computer to perform a task is a powerful way to reveal hidden assumptions. What the problem of machine learning on human data reveals, is there is no objectivity. Our most 'objective' models will simply learn and then reinforce biases and prejudices and cause harm to people. We can't build objective, neutral, value-free models when it comes to human behavior, because humans change their behavior in response to the models. When we reject stereotypes we are imposing a set of values, just as much as when we embrace them. Machine learning forces us to confront the fact that this is exactly what we are doing.
There's a kind of philosophical crisis going on here. We need an entirely new way to think about the limits of objectivity in human sciences, and how to create ethical models in the presence of feedback between science and society. The language of objective physical science doesn't work.
It's interesting that they frame the problem as "woe be to the black users, the AI might discriminate against them". First, approximately all of Google's most valuable users (or any other ad-based platform) are non-black. Secondly, the problem is empirically the opposite.
Er, how are those couple doodles examples of the empirically opposite problem? They have nothing to do with AI. (Also... is any representation of people of color an instance of anti-white bias? Because otherwise, I don't see what the problem is with those doodles.)
Also: "approximately all of Google's most valuable users (or any other ad-based platform) are non-black"... "most valuable users"? Really? Putting aside whether that's even an accurate assessment in the financial terms in which it was presumably intended: Do ethics stop at treating well only the people with most profit potential? Is impact on business our only arbiter?
The problem opposite of artificial intelligence unintentionally erasing blacks is a company's built-in intelligences intentionally erasing whites, minimizing their achievements, and deconstructing their identities. Whites are not "allowed" to have achievements; they must be couched in terms of general social achievements which of course are (according to the doodles) aspirationally majority non-white, and ideally non-white and non-male.
It's simply a well written and concise opinion (that's the opposite of a rant), completely relevant to the article in question, the contents of which, moderator 'dang' disagrees with. This is easily seen by his history of comments which exclusively censor and censure only one viewpoint.
And this is fine to hold an opinion as the expense of keeping an open mind. It is fine to use your power as a mod to advance your opinion. What is not fine is to hide your agenda underneath feigned impartiality. Your comment should read "White people can not complain about being snubbed by tech companies on this forum". It would be a far more direct and honest approach.
I'm unclear on what a "race rant" is or how this could qualify. The behavior I have described is the academic consensus on the appropriate treatment of race, it's not totally surprising that Google or anyone else would participate in that consensus. The footnotes of this article are a good starting point.
What I'm calling a rant is making grand statements on inflammatory topics that come from a place of resentment.
On HN, we're hoping for thoughtful statements that come from a place of curiosity. Those aren't just two different kinds of discourse, they're incompatible, like forest fires are incompatible with forests.
Edit: one privilege of HN moderation is getting devastating critiques of one's analogies from emailers. Ok, forest fires don't really make my point here. But there is an article called "When Analogies Fail" on the front page right now, so I'll leave it.
Google, probably more than any other company in the world, gets that their "most valuable users" aren't large coherent groups of any sort, Google wins by catering to everybody at once (aka "the long tail", to use a 2000s-ism). This is what drives their relentless focus on machine learning everything, because machine learning allows you to serve microscopic constituencies. You don't need machine learning to sell beer to white, suburban males.
"You don't need machine learning to sell beer to white, suburban males."
This is both bigoted and empirically false. Do you imagine suburban whites to be a particularly unsophisticated demographic? Imagine saying "you don't need machine learning to sell fried chicken to urban blacks".
When I first moved to London, one of my co-worker was a very exuberant young Scottish woman. Her accent was near impossible for me, and she fully realised this, but at one point we came to the mutual understanding that I did in fact pick up most of what she was saying, even though it was only because she talked so much, that she didn't really need to intentionally repeat what she was saying.
I thought most people would be able to get written Scots dialect, but I did struggle with Walter Scott's Rob Roy where one of the characters speaks it phonetically.
There's also a guy doing a newspaper column in Scots in The National, causing quite a bit of controversy.
Spoken language is another matter. I've had to interpret between an Ulsterman and an Afrikaaner both of whom were nominally speaking English as a first language.
Is this a result of the AI being written by a bunch of non-black people? It seems to me that this problem is more of a reflection of the people working on the code.
There may be some of that, but I think it has more to do with AI still being young, and it's easier to work with less ambiguous text for learning. That and everyday speech move much faster than formal writing. Now that my son has gone off to college and is less accessible as a resource, I expect to almost totally lose touch with the speech of "these kids today."
> I expect to almost totally lose touch with the speech of "these kids today."
Is this long term really a problem though? Regression to the mean: No one cares about the specific slang of the 60s anymore. And the hip slang of today will be out of fashion in 15 years.
In a similar vain the extreme T9 keyboard text message abbreviations of the youth in the 2000s did go out of fashion because of automatic spell checking and speech recognition of modern smartphones.
How long term? What's "a problem"? We know with certainty that given the passage of enough time, the speech of "these kids today" will be completely unintelligible to you.
>Do you think written Standard English is less ambiguous than any other English
Definitely. Things are clearer when you write formally and strictly adhere to grammatical rules. When I'm speaking casually I omit words and (ab)use punctuation much differently, and while that is perfectly acceptable to do, it is more complex and ambiguous to parse.
As far as I'm aware the same pattern exists in a lot of languages and dialects, and it's possible to define formal AAVE too.
I don't think there is anything "formal" about Standard English. Can you give an example of what you're talking about? Which grammatical rules do you have in mind?
I don't know what you mean by formal black English (AAVE) vs. ... casual black English?
>I don't think there is anything "formal" about Standard English.
"Standard" English as she is spoke? No. But a3n was specifically talking about the formal version often called "Standard written English". It is very prescriptive and less flexible, which makes it easier to parse.
> Can you give an example of what you're talking about? Which grammatical rules do you have in mind?
Let's start with "every rule involving punctuation".
I'm not trying to be dense, but I still don't see it. Punctuation helps with identifying boundaries of clauses, whether they're dependent or not, whether a sentence is a question. Is this what NLP struggles with? I don't think so. One of the big unsolved problems is understanding the referents of pronouns (and other proforms, like "I will [do so] too"). Style guides and generally prescriptivism does not help that.
Another ambiguity is parse ambiguities like " I saw the girl with the telescope." What does "formal" English say about that?
>
Do you think written Standard English is less ambiguous than any other English, or written language?
Absolutely. It follows a strict set of formal rules and you can readily find sources to learn those rules. Heck, copyeditors regularly debate the specific nuances of how to consistently write Standard English.
Other dialects (and spoken language) are far less formalized. Where is the AP Stylebook or Strunk & White for AAVE?
Partly that, but my guess would be that it's because it's largely trained from written English (not the sort you see on Twitter and in SMS), and written English is usually one of the standard varieties, and the standard varieties bear little resemblance to AAVE.
I would guess the same problem occurs with other dialects which are sufficiently different from the standard.
I expect it's mostly a reflection on the learning data in language processing. Almost any corpus of text you can easily find for experimenting is close to standard English. Dialects very rarely appear in published texts unless it's just to point out that a specific person being quoted uses that dialect.
Choose a student of any colour. When they need a text corpus for research, do you think they'll reach for a different, known, large text collection that matches their background?
There is nothing about a computer algorithm that renders it immune to the power structures around it. The people who train these systems don't even realize they are creating a biased model. Borrowing from postcolonial theory, this is related to the concept of epistemic violence. The way we organize human information, the way we categorize and evaluate viewpoints and ways of understanding the world is a battleground for imperialism. Intent does not matter. The algorithm has no intent, the researchers don't intend to create a vehicle of oppression, and yet it happens anyway. They have created a difference in economic utility between standard, hegemonic English, and marginalized dialects of English which inevitably will have social ramifications. If this study weren't conducted, would we have ever found out?
The best part about black twitter is how dynamic and innovative it is. There is a certain playfulness with language which is frowned upon in many academic settings & especially by linguistic prescriptivists, but healthy languages change over time.
And "more diverse datasets" isn't enough. I get that you need to train an AI on a set of old data, but any truly good NLP would be able to adapt to new terms and phrases as they're invented. Not saying it's an easy problem to solve, but many people I've spoken to don't seem to even be aware that language recognition is necessarily a moving target.
The article refers to another article about "Black Twitter"[1], but the examples are all about hashtags. It doesn't seem to be a dialect; it's about subject matter of interest to the community.
For context, Olga Russakovsky is the lead author behind the ImageNet Large Scale Visual Recognition Challenge, This is one of the most popular computer vision benchmarks in the world. Olga has expertise dealing with issues of dataset bias.
A large number of ImageNet classes are dogs and birds, so most of the representational power in CNNs trained on this set are devoted to distinguishing between similar-looking dog breeds and bird species.
This, I suspect but do not know, may be a failing of the current direction of NL AI solutions. Selecting and using datasets to train NLP systems is a hard thing, especially if you truly treat it as a pattern to be learned rather than an enhancement of an underlying description of the particular language you're trying to learn.
Are we headed towards a time when deeply trained networks seem to work in an acceptably large number of cases, but how and why are unknown and susceptible to fatal flaws that rarely but catastrophically reveal themselves? I suspect that is the case with the current path of machine learning but hope I am wrong. It just seems to me that the inevitable results of certain efforts in machine learning are magic boxes that "work" but no one understands why or how, which smacks of the days of medicine prior to an understanding of the germ theory of disease where certain efforts at sanitation due to the belief in "ill humors" did improve health but were based on utterly false underlying theories, and those successes tended to reinforce other false solutions that did not improve health but in fact was harmful.
104 comments
[ 3.0 ms ] story [ 153 ms ] threadIt's hard to take an article seriously when it opens with a logical fallacy. Yes, the WSJ uses standard English and only 8 million people read (subscribe to?) it. That does not mean that only 8 million people in the US use so-called standard English. There are also a lot of people who don't subscribe to that particular publication but still speak or write in "standard" English and that group is much larger than the group of WSJ readers.
It IS "the" language of that two million. Just not ONLY those two million.
WSJ dataset is not the only such dataset. NLP is not very good with standard English yet and usually doesn't generalize from topic to topic. Dialects and other languages - especially those without formal rules - will come when we can deal with standard English.
The point is for rules to be formal then the must be formalized somehow and codified. This could be online or in a book, but the point is that there must be some clear delineation between when the rules are followed and when they are broken. Otherwise what's the meaning of "formal" rules?
> If so, aren't these kind of antithetical to AAVE on the face of it?
Yes. My argument is that AAVE, almost by definition, is an informal dialect without formalized rules.
This is getting rather off topic, but I think it relates to the original point of why NLP might start with Standard English even if you are not biased. A large corpus of Standard English text (such as from the WSJ) will generally be very internally consistent precisely because it follows a set of formal rules codified into a style guide. As there is no such equivalent for AAVE, even gathering a large and internally consistent corpus of AAVE text seems prohibitively difficult. That being said, I do hope researchers are working on gathering text from Twitter to build up new training sets.
So every form of English is an "informal dialect" then? Because this ain't French with the Académie publishing strict rules for use of the language. Do you also say that the languages of remote Amazon tribes aren't "real languages" because they don't have a formal government body publishing written rules?
Or do you just want to bash on AAVE and are grasping at straws for reasons why?
Yes, the majority of spoken English does not follow the rules of Standard English. Pretending that such rules don't exist is willful ignorance though: the WSJ obviously write a more formalized version of English than teenagers do in text messages.
> Do you also say that the languages of remote Amazon tribes aren't "real languages" because they don't have a formal government body publishing written rules?
Nowhere did I say that AAVE is "not a real language" because it's less formalized than Standard English. Prior to spelling reforms, English itself was extremely inconsistent and informal, but I certainly don't pretend that it wasn't a language.
> Or do you just want to bash on AAVE and are grasping at straws for reasons why?
I'm not trying to bash AAVE. In fact, I'd even posit that the reason AAVE isn't more codified is perhaps because of racial bias which treated it simply as "incorrect" English instead of a separate dialect worthy of formalization. Pretending that all languages are equally formalized is simply willful ignorance though.
> NLP is not very good with standard English yet and usually doesn't generalize from topic to topic. Dialects and other languages - especially those without formal rules - will come when we can deal with standard English.
The rules you're talking about, that get printed in books and studied, are not linguistic rules. Crucially, this means they are not widely observed in printed standard English, which in turn means they can't be relevant to training a language model to understand printed standard English.
The "formality" you seem to want to talk about has no place in this discussion. It is not relevant to any language. gordonguthrie is correct to point out that the assumption lqdc13 is trying to make is false. You are wrong to contradict him using a meaning of "formal rules" that you brought to the conversation yourself. It had a meaning -- a completely unrelated meaning -- before you showed up.
I agree that they're not widely observed in written English, but they are consistently observed in the WSJ, which was the origin of this entire debate.
As lqdc13 pointed out, NLP still isn't even consistently good at understanding standard English. One could reasonably posit that that's due to the inherent ambiguity and inconsistency of most writing and that focusing on a narrower, standardized document corpus (the WSJ) you could get better initial results. What, exactly, is controversial about that? Do you really think that the language of the WSJ is no more consistent and formalized than the language of Twitter users?
http://s3.amazonaws.com/academia.edu.documents/41737546/The_...
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.516....
EDIT:
If you want a really deep dive, you can also check out "African American English: A Linguistic Introduction" from Cambridge University press:
www.cambridge.org/us/academic/subjects/languages-linguistics/sociolinguistics/african-american-english-linguistic-introduction?format=PB&isbn=9780521891387
That being said, I think this research emphasizes that AAVE is anything but standardized. That's not meant as a pejorative statement: it's just acknowledging that, like most languages in history, AAVE has not gone through a process of codification and standardization to formalize it.
like most languages in history, AAVE has not gone through a process of codification and standardization to formalize it
I'm not sure what you're getting at - prescriptive grammars of the codified form you're describing are the products of their political and economic circumstances. There isn't some Hegelian trajectory of linguistic validity, where all variants aspire towards legalism.
Equally true, right? But somehow it seems just a little misleading.
Likewise, notice the deceptive switch between "English" and "standard English." So basically language bots trained to identify standard English are correctly identifying non-standard English dialects as not standard English.
If the training corpus for a machine learning model contains stereotypes and biases, then the output of the model will reflect those prejudices.
When that model is used in large-scale applications, it will not just repeat those biases, it will amplify them. It propagates biased language, which feeds back into the model in a self-reinforcing vicious cycle of increasing bias.
Should we attempt to detect prejudices in machine learning models and actively counter them, creating a self-reinforcing virtuous cycle of balance and tolerance?
----
As an aside, for an article about language, this one has a lot of typos. Doesn't anyone proofread before posting?
Isn't the ideal system to not assume anything about the content it's analyzing, but to base the model on behavioral patterns instead? In the case of Google ranking, either a site has traffic/low bounce/backlinks/social cred or it doesn't--why is Google attempting to 'read' and analyze the content itself? (Other than to serve ads of course)
Because Google no longer just searches for literal strings of text, ranked by links.
They now infer your meaning and find results that match your intent -- even if the exact search string doesn't exist in the search results.
For more on this, read about Hinton's work on "thought vectors" [1].
So if Google can't determine the meaning of African American language as well as it can standard English, then that content is likely to be less visible in search results.
[1] https://www.theguardian.com/science/2015/may/21/google-a-ste...
Bing seems to do the same thing. I really wish there were a grep for the web because both engines get it wrong even when I try to exclude results.
It's basically grep for the web, since it uses the actual query instead of the meaning of it.
As an aside - this is a slippery slope conversation. Some people believe language requires rules. Other, more special people, think language is fluid and anarchist. The latter of those two do not make search engines.
My guess is that it's unlikely to provide sources that are in one of these dialects, as the link density will be low.
> Some aspects of communication are likely to prove more challenging, Hinton predicted. “Irony is going to be hard to get,” he said. “You have to be master of the literal first. But then, Americans don’t get irony either. Computers are going to reach the level of Americans before Brits.”
A further problem, though, is that search engines try to understand the query and the content searched at a deeper level than mere keyword matches. For example, they may try to pick up synonyms (if I search for “movies in <local area>”, perhaps results for “films” might be helpful too) and associations.
So that when you search for "cheesecake recipes", you don't get a well-sourced, high-traffic article from a respectable website describing in detail how the new business initiative by Cheesecake Factory is a recipe for disaster.
* Disclaimer: I pulled that example out of thin air. There's actually no such article.
Which is precisely why social media (e.g. your newsfeed) is such an awful source for news, especially when you consider how low the standards are on digital publishing these days...
No.
Depends what you mean by "the corpus". Let's take the example of loan applications. Maybe the training data contains credit scores, biographical information, and personal essays about the applicants from loan officers.
Let's say loan officers tend to use more negative language when talking about black people.
Any decent machine learning system given racial data will not adopt those biases. In fact, it will counteract them. It will determine that e.g. the semantic content of the essays is an indicator of loan suitability, and it will also notice if it has to e.g. add something to the essay score for black people to make the best decisions.
As long as the output data is objective, competent ML systems will account for any bias in the input data.
> Should we attempt to detect prejudices in machine learning models
We don't have to. If the output examples we are training on are true, the ML algorithm won't adopt any incorrect biases.
This is literally the point of machine learning. You want to make the best possible decision. If your algorithm has incorrect human biases, you're losing money. ML naturally accounts for this stuff.
And the bias people are probably talking about is bias in the data, not bias in the "input", as you put it.
See http://arxiv.org/abs/1607.06520 for a counterexample to your point.
Do you know what I mean? There's truth as in what the data says, and there's truth as in what's underlying before social processes " corrupt " the truth. This paper examines ways to get at the latent truth.
Do you think that's a useful endeavor? It's dismissive to say it has anything to do with what the "author likes".
Ah, so you're defining "truth" as "whatever I imagine the world would be like if society worked the way I wished it did". There's another word for that; "fantasy".
This is rarely the case when working wtih real data, and thus inspecting whether our models are biased against protected classes is probably one of the most important things an ML practitioner should do.
There is no reason an ML algorithm would be biased against a "protected class". It doesn't know what those are. It's possible that the algorithm will uncover a truth that you don't like, like that there as differences in risk profiles across race or gender, but that doesn't mean the algorithm is incorrectly biased. It just means that reality is at odds with how you might want it to be.
Which data are you referring to? In most cases, the training data isn't human-generated, and if it is, we usually want to match human behavior as close as possible.
Read anything written by Solon Barocas: http://solon.barocas.org/
The issue of prejudice stems from refusing to treat AAVE as a valid input to an AI, mirroring the racism in American culture that similarly devalues and excludes it. Output can conform to some approximation of "standard" form without being problematic, following the pattern of interaction between standard and dialect speakers. As a speaker of the standard, it's gauche and derisive to imitate or refuse or comprehend a dialect speaker, not to respond in your own form.
What is really the difference between spelling a word wrongly and using a non-dictionary word to express something. We except one as "slang" but will criticize the other because they don't comply with established rules. Isn't that what slang is? A break from established rules.
Anyways, just a thought, because I agree with you; if you publish - proofread. I just find it interesting to throw a Socrates argument at myself from time to time.
Btw and somewhat of topic.. this was the Pinker interview that was most interesting:
https://www.youtube.com/watch?v=egU0dxzFKAQ
Programming a computer to perform a task is a powerful way to reveal hidden assumptions. What the problem of machine learning on human data reveals, is there is no objectivity. Our most 'objective' models will simply learn and then reinforce biases and prejudices and cause harm to people. We can't build objective, neutral, value-free models when it comes to human behavior, because humans change their behavior in response to the models. When we reject stereotypes we are imposing a set of values, just as much as when we embrace them. Machine learning forces us to confront the fact that this is exactly what we are doing.
There's a kind of philosophical crisis going on here. We need an entirely new way to think about the limits of objectivity in human sciences, and how to create ethical models in the presence of feedback between science and society. The language of objective physical science doesn't work.
https://www.google.com/search?q=white+man+white+woman&biw=13...
https://www.google.com/doodles/veterans-day-2015
http://www.cbc.ca/news/trending/google-doodle-juno-reaches-j...
Also: "approximately all of Google's most valuable users (or any other ad-based platform) are non-black"... "most valuable users"? Really? Putting aside whether that's even an accurate assessment in the financial terms in which it was presumably intended: Do ethics stop at treating well only the people with most profit potential? Is impact on business our only arbiter?
It's simply a well written and concise opinion (that's the opposite of a rant), completely relevant to the article in question, the contents of which, moderator 'dang' disagrees with. This is easily seen by his history of comments which exclusively censor and censure only one viewpoint.
And this is fine to hold an opinion as the expense of keeping an open mind. It is fine to use your power as a mod to advance your opinion. What is not fine is to hide your agenda underneath feigned impartiality. Your comment should read "White people can not complain about being snubbed by tech companies on this forum". It would be a far more direct and honest approach.
https://en.wikipedia.org/wiki/Whiteness_studies
On HN, we're hoping for thoughtful statements that come from a place of curiosity. Those aren't just two different kinds of discourse, they're incompatible, like forest fires are incompatible with forests.
Edit: one privilege of HN moderation is getting devastating critiques of one's analogies from emailers. Ok, forest fires don't really make my point here. But there is an article called "When Analogies Fail" on the front page right now, so I'll leave it.
This is both bigoted and empirically false. Do you imagine suburban whites to be a particularly unsophisticated demographic? Imagine saying "you don't need machine learning to sell fried chicken to urban blacks".
https://mobile.twitter.com/MarkHamiIl/status/778141129564905...
I thought most people would be able to get written Scots dialect, but I did struggle with Walter Scott's Rob Roy where one of the characters speaks it phonetically.
There's also a guy doing a newspaper column in Scots in The National, causing quite a bit of controversy.
Spoken language is another matter. I've had to interpret between an Ulsterman and an Afrikaaner both of whom were nominally speaking English as a first language.
Is this long term really a problem though? Regression to the mean: No one cares about the specific slang of the 60s anymore. And the hip slang of today will be out of fashion in 15 years.
In a similar vain the extreme T9 keyboard text message abbreviations of the youth in the 2000s did go out of fashion because of automatic spell checking and speech recognition of modern smartphones.
How long term? What's "a problem"? We know with certainty that given the passage of enough time, the speech of "these kids today" will be completely unintelligible to you.
Definitely. Things are clearer when you write formally and strictly adhere to grammatical rules. When I'm speaking casually I omit words and (ab)use punctuation much differently, and while that is perfectly acceptable to do, it is more complex and ambiguous to parse.
As far as I'm aware the same pattern exists in a lot of languages and dialects, and it's possible to define formal AAVE too.
I don't know what you mean by formal black English (AAVE) vs. ... casual black English?
>I don't think there is anything "formal" about Standard English.
"Standard" English as she is spoke? No. But a3n was specifically talking about the formal version often called "Standard written English". It is very prescriptive and less flexible, which makes it easier to parse.
> Can you give an example of what you're talking about? Which grammatical rules do you have in mind?
Let's start with "every rule involving punctuation".
Another ambiguity is parse ambiguities like " I saw the girl with the telescope." What does "formal" English say about that?
Absolutely. It follows a strict set of formal rules and you can readily find sources to learn those rules. Heck, copyeditors regularly debate the specific nuances of how to consistently write Standard English.
Other dialects (and spoken language) are far less formalized. Where is the AP Stylebook or Strunk & White for AAVE?
I would guess the same problem occurs with other dialects which are sufficiently different from the standard.
Choose a student of any colour. When they need a text corpus for research, do you think they'll reach for a different, known, large text collection that matches their background?
And "more diverse datasets" isn't enough. I get that you need to train an AI on a set of old data, but any truly good NLP would be able to adapt to new terms and phrases as they're invented. Not saying it's an easy problem to solve, but many people I've spoken to don't seem to even be aware that language recognition is necessarily a moving target.
[1] http://www.theatlantic.com/technology/archive/2015/04/the-tr...
A large number of ImageNet classes are dogs and birds, so most of the representational power in CNNs trained on this set are devoted to distinguishing between similar-looking dog breeds and bird species.
Are we headed towards a time when deeply trained networks seem to work in an acceptably large number of cases, but how and why are unknown and susceptible to fatal flaws that rarely but catastrophically reveal themselves? I suspect that is the case with the current path of machine learning but hope I am wrong. It just seems to me that the inevitable results of certain efforts in machine learning are magic boxes that "work" but no one understands why or how, which smacks of the days of medicine prior to an understanding of the germ theory of disease where certain efforts at sanitation due to the belief in "ill humors" did improve health but were based on utterly false underlying theories, and those successes tended to reinforce other false solutions that did not improve health but in fact was harmful.
This blog post summarizes the paper "Machine Learning: The High-Interest Credit Card of Technical Debt": https://blog.acolyer.org/2016/02/29/machine-learning-the-hig...
[1] https://en.wikipedia.org/wiki/Oakland_Ebonics_controversy