The tricks of spammers are many and subtle. And I’m not always sure what their game is.

I was selling something on CraigsList recently. One of the inquiries I received asked me to e-mail the person’s personal e-mail, included in the mail message, directly (rather than hitting Reply); this was ostensibly to keep scammers and bots out. A bit odd, maybe, but somewhat plausible (at least in a misestimating-computer-capabilities kind of way), and CraigsList interactions are full of communications behavior that seems odd to me. I sold the item to someone else, but to close all my open loops on it, I e-mailed the address they specified to tell them that the item had been sold.

Big mistake. Since then, I’ve gotten at least a dozen e-mails, all with the same subject as that mail I sent and similar text about being interested in or attracted to me and my ad, asking me to sign up on a particular site (no fee, of course!) to send a private message. Occasionally, a picture is attached.

I’m not entirely sure what the game is with this. Phishing? Malware installation? Getting credit card info? I haven’t visited any of the linked sites, so I don’t know. But it’s yet another “spammers are doing what? and why?” moment.

Four bright spots in the American justice system this week:

• A federal judge has ruled that National Security Letters, at least accompanied by gag orders, are unconstitutional.
• DC District Court ruled that the CIA cannot reject out-of-hand the ACLU’s Freedom of Information Act request for information on the drone program. The lower court judge’s ruling effectively said that she thought the stuation absurd and unjust, but saw no way to compel the disclosure that was compatible with law and precedent. The district court disagreed with the no-compatible-way part, not the this-is-absurd part.
• The 9th Circuit set aside the 20-year-old capital conviction of Debra Jean Milke. Her conviction was based almost entirely on the testimony of a police officer that she had confessed, even though she and the other two other men convicted all denied her involvement. The officer had a history of lying under oath and mistreating defendants; he had also been ordered to record his interview with Milke but did not have a recording to back up his assertion that she confessed.
• Judge Otis Wright is laying down the smack on Prenda Law and its related lawyers, etc. for running what seems to be a massive scheme of extortion and fraud, using the courts and allegations of copyright infringement of pornographic videos to get people to pay settlements. Specious legal theories, extortion, forgery, a court hearing that would make Abbot and Costello proud, this case has it all.

The Guardian ran an article this weekend discussing predictive policing and its future. Read it, it’s worthwhile.

I greatly appreciated a number of the moral concerns Morozov raises, and he does an excellent job of connecting the issue to much of its surrounding social context. He also is quite balanced in his approach, urging caution while being cognizant of the real, on-the-ground benefits of the technology.

Unfortunately, he falls into the same trap as Eli Pariser (The Filter Bubble) in ascribing algorithmic deficiencies to questionable allegiances of their creators:

But how do we know that the algorithms used for prediction do not reflect the biases of their authors? For example, crime tends to happen in poor and racially diverse areas. Might algorithms – with their presumed objectivity – sanction even greater racial profiling?

I am sure there are people doing adversarial data mining and engaging in unscrupulous analytic activities, both in execution and intent. At RecSys 2010, for instance, the industry keynote discussed how one gambling operation uses customer modeling to predict when a patron is reaching a high risk of gambling to a level where they quit cold-turkey and intervening based on this knowledge, not to protect the gambler from their habit but to keep them coming back the next weekend.

But programmers building their algorithms inappropriately are not, in my opinion, the biggest threat. There is plenty of opportunity for racial profiling, religious bias, and other troubling bases for law enforcement to enter the equation without the need for complicit algorithmists.

First, the algorithm’s output is only as good as its input. The article acknowledges this to some extent, observing that police records are limited in their coverage and scope while the prospect of using data from sources such as Facebook could provide a much broader corpus to mine for crime signals. It does not, however, tie this in to the broader scope. Not only are the police records incomplete, they are tainted by the biases of current enforcement and law. If the police enforce laws disproportionately against certain groups (or perhaps worse, if the laws are written to have more impact on those groups), then those groups will be over-represented in the data; if the crime model learns this trait, it lends analytic support to closing the loop of a vicious cycle.

But suppose the data were perfect and the model fair. Suppose it then identifies “young male of Scandinavian descent”¹ as a trait predicting significantly² higher-than-normal probability for criminal involvement.

We are then faced with a profound moral question: is using a model that has identified such a feature, even from unbiased data, moral? Is it acceptable to increase police scrutiny (and likely hassling) of a certain ethnic, racial, or socio-economic group, even if we have data that says it is accurate?

Or is it better to sacrifice the accuracy of our crime modeling to sustain a democratic society with liberty and justice for all?

As data miners, model-builders, and members of society, we must think about these questions.

¹ For the record, I’m under 30 and half-Swedish.

² I use ‘significantly’ in the statistical sense, meaning that it is a robust predictor instead of noise or happenstance, not that it increases the probability by a substantial amount.

H1B’s. U-visas. J1’s, F1’s, guest worker permits, green cards, tourist visas.

There are too many types of visas. And they’re too temporary, too revokable, too dependent on the whims of a ‘sponsor’.

How about simplifying to just three groups of people:

• Citizens. Born or naturalized.
• Residents. Persons residing long-term or permanently in the country. They are entitled to all the rights and privileges of citizenry except voting, holding elected office, and possibly armed service. They can work, go to school, or whatever. Since continued residency is not required to maintain resident status, international commuters can classify as residents. Residents are eligible for citizenship after a reasonable waiting period (say 5 years, spending 54 of the last 60 months in the country), a civics test, and maybe an English proficiency test.
• Visitors. Persons in the country temporarily. They only reason this category exists is to let people into the country for business or pleasure without setting up a taxpayer number for them. The ‘temporary’ is not enforced. Visitors probably don’t qualify for many social services, except for health care (if/when we get public health care). If a visitor wants to stay long-term, they can go to their local immigration office, fill out a bit of paperwork, and get their taxpayer number.

Resident status is not contingent on an employer, family member, or school sponsoring you. Residents and citizens alike are subject to the same law. And if the law isn’t appropriate for one class, we should really question if it is appropriate for the others.

If we switch from income taxes to a consumption tax (fair tax, VAT, something like that), then we may be able to do away with the visitor/resident distinction as well.

Both visitor and resident status should be easy to get, but that’s a subject for another post.