If you have a programmatic way of determining the best match for an arbitrary long tail query, then haven't you already solved your search problem?
I suppose if you had a large enough (more or less hand crafted) sample set you could use machine learning techniques to optimize the weighting of each field, but I don't know if that would get you an answer any faster than tweaking human-designed weightings.
Crowdsourcing search result feedback might be another way to make this scale.
I guess that may have been a little predictable from TFA, but I'm regularly surprised by the folks who crop up at HN. I've been reading both your and VB's work for years, it is a funny coincidence to discover there is an relationship between you.
Here's how I'd look at it: Is the kind of processing you are doing more like the kind of thing you'd do in a text editor (where you need to move around a lot in the source file) or the kind of thing you'd do with "cat" or "tail" (where the last couple of lines is all the context you need).
* SAX is stream based. This is like "cat" or "tail" on a Linux system, there is no history besides what you yourself put together. If you don't need to walk up and down the tree (much) then this is by far more space and time efficient.
* DOM is tree based. This parses the whole document into memory and gives you an API for walking up and down it. If you need to jump around in the document, this may be the way to go, but note that parsing and storing the full document takes time and memory.
Regarding #1, in my experience this isn't a phenomenon limited to middle-aged women. That's pretty much how "normals" use the Internet (as much as I dislike that term). This is what regular people, and quite a few techies, do.
"The right to make limited use of the work" is different from publishing rights. This technology wasn't in common use when I went to school, but it seems really easy for the school to slip in something granting them the right to upload this to TurnMeIn into one agreement you sign or another.
For that matter, it's not entirely clear to me what the copyright issues are around work you do for your university. I mean you're already handing it over to the Professor with the exception that he'll do some things with it. Can he hand it out to future classes as an example of the kind of thing he's looking for? (This seems common.) Can he bundle it with other papers and use it as a textbook in future classes? (This seems uncommon.) I do think it is
common for schools to at least co-own the publishing rights to a Doctoral thesis. (And in fairness, if your advisers are doing a good job, they may have a legitimate claim to that.)
It certainly seems like in the sciences graduate students are frequently complaining about professors taking undue credit for the student's research. Is a paper different than that? How much of your work for a university falls under some sort limited "work-for-hire" style agreement?
I agree that visitors-to-conversions are is the core underlying metric.
But these bounces are a subset, and large one, of the visitors that don't convert. It seems that area of greatest opportunity should be within the 80% of visitors that bounce. Assuming I've got the conversion engine relatively well oiled, I'd really like to draw in those 80% of visitors that virtually walk out the door.
Wait, maybe I'm misunderstanding you, but you've found that having people explicitly searching for you increases your bounce rate? Wouldn't those people be less likely to bounce? Your site is literally what they were looking for (even if they are disappointed when they arrive).
For what it's worth, while there are some visitors explicitly looking for my site, most are coming in from long tail search queries (and those are the ones that tend to bounce).
I'm just wondering if 75% bounce rate is something to worry about. My gut feeling is that this is too high. In most other sites I've worked with this was more in the 35 to 45% range, but those were entirely different businesses.
> That's a 6,000x return on his capital in 5 years.
> I'll put it this way--if you made a $250 IRA contribution in 2005, it would have to be worth $1,500,000 today to match that return.
If you invested one penny that investment would now need to be worth $60 to match that return. Maybe it is just me, but $250 to $1.5M doesn't help me visualize a 6000x return at all. This analogy reminds me of an old Dennis Miller bit from Saturday Night Live where to illustrate the size of the national debt he takes out a dollar bill and says "it would take stack of these 3 trillion high to pay it back".
I'm not a big fan of this business but I take issue with the "Teachable Moment" image seemingly created for this post. How does what TurnItIn is doing constitute an "IP Rights" violation?
Surely in the contract that the school signed they gave TurnItIn permission to use these data in this way, and asserted that the teacher, by uploading the document, is ensuring that he or she has the right to grant that permission. At the very least this pushes the IP issues onto the school, as far as TurnItIn knows they have every right to use this content in that way.
It seems probable that the admissions paperwork for the school probably granted the school the right to make limited use of the work the student turns in. I know when I had to complete a "Bachelor Thesis" the school retained the right to use that content, at least for publication and probably for sale if they chose to. Is that uncommon?
Also, I think the problem TurnItIn is trying to solve is less about copyright violation and more about plagiarism/intellectual dishonesty in an academic context. You could readily buy a paper or essay from someone to obtain the copyright permissions, but if you submit that as your own work to the professor you've violated academic ethics.
You make this general point (correct me if I'm wrong, but I think your argument is that feminists care more about equality at the top end of the scale than the bottom) a couple of times in this thread. I'm not quite sure what you are trying to say.
If you mean that feminists don't pursue discrimination cases when it comes to low-pay/low-prestige/high-risk jobs, that's demonstrably false: There are many examples of "women's lib" folks organizing factories or migrant workers or whatever. Many of the related Supreme Court cases (e.g. Ledbetter, Chrapliwy) were about discrimination in blue collar jobs at factories. It is likely that these groups spend more time and money pursuing discrimination issues in blue-collar and low-wage workplaces than they do in the higher-salary, higher-prestige jobs, not least of which because (a) discrimination is probably more blatant and common there and (b) more people, men and women both, have below average (arithmetic mean) incomes than have above average salaries and (c) more women have low-pay, low-prestige jobs than men.
If you mean that feminists hold up the gender imbalance in high-wage, high-prestige jobs when advocating for their cause, of course they do. Everyone would. What middle class donor is going to pull out their wallet to ensure their daughters can get janitorial jobs? The aspirational jobs are simply more inspiring. Moreover, ensuring a gender balance in the lowest-wage, lowest-prestige job categories while leaving the top end alone isn't really achieving equality, is it?
By the way, I don't think your notion that women are under-represented in low pay/low prestige jobs holds up to scrutiny anyway. Women are over-represented in the lowest end of the scale, and under-represented in the highest end of the scale. Take a look at BLS or Census Bureau data on the highest and lowest paying jobs and the distribution of women and men in each. Here's an interesting table of unemployment rate by occupation and sex: http://www.infoplease.com/business/employment-rate-occupatio.... Note that the unemployment rate pretty closely correlates to what you'd expect from a gender bias: the unemployment rate in traditionally feminine occupations is higher for men than for women, the unemployment rate in traditionally masculine occupations is higher for women than for men.
I think the only claim in your line of thinking that holds up to the least bit of scrutiny is that men are over-represented in dangerous job categories, or more to the point that men account for a disproportionate share of workplace fatalities. This is true (although your numbers are a bit off). The breakdown of hours worked by sex is around 55% men to 45% women. The breakdown of fatal injuries is around 93% men and 7% women. But with only 4,300 workplace fatalities in the United States in 2009, we're talking about one fatality for every 59 million hours worked. Workplace fatalities are not the biggest labor problem in the US, probably not even for the most dangerous occupations.
I don't mean to flog this thread that is fairly dead at this point, but seriously, the level of largely unchallenged inaccuracy, "truthiness", and borderline misogyny in this thread is not only well below HN's typical level of quality, but frankly it is disturbing. This conversation needs some balance.
The parent comment here currently has 15 upvotes. Another comment on this page claiming there hasn't been gender based discrimination "for a very long time" has 3 upvotes. Many of the claims made in these posts are simply demonstrably false. They seem to represent some kind of knee-jerk, well, not misogyny, but something approaching it. Is there a term for "dislike of women" in place of "hatred of women"?
I'm male, but gender inequality and discrimination in the workplace is obvious to me, let alone to the women in my life. Gender isn't the dominant factor in predicting income, rank or career success. It may not even be among the top few factors. But it is certainly a factor, and not a negligible one at that. To be candid I'm both astounded and troubled this isn't obvious to more of the participants in this conversation.
> a woman can wear almost whatever she likes in a situation where a man can't
I'm not sure that's true. There's a lot more diversity in women's clothing. I think that probably makes it more difficult to dress for the occasion as opposed to less. For instance, a man can wear slacks and a dress shirt to work and then throw on a tie and jacket for a formal dinner. Most any single woman's outfit would be out of place in one environment or the other.
(In fact, men can get by in almost any situation with a pair of dress pants and a button down shirt.)
> Incidentally, women dress up to impress other women. Men simply don't notice eyelashes or nail varnish or whatever.
Agreed. That has been my observation as well. I think there may have even been scientific studies on this point.
> women haven't really been discriminated against in any serious, society wide way for a very long time
Wow, really? Upvotes even? I'm honestly surprised no one has pushed back on that statement yet.
Not to be condescending, but I'll guess that you are a man, and a young one at that.
In the west, there is certainly less officially sanctioned discrimination, although that is a relatively new phenomenon. (Some examples from the US: The Ledbetter Fair Pay Act was passed in 2009. The FMLA was passed in 1993. The Tailhook Scandal was 1991. Roberts v. United States Jaycees was 1984. Chrapliwy v. Uniroyal was 1982. Title IX was 1972. The EEOC and The Equal Pay act were 1963. Even 1920 isn't really that long ago.)
But if you honestly believe there is no "serious" or widespread gender based discrimination, you must either be very young, very privileged, or not paying attention. That statement just doesn't reflect the experience of people in the working world, whether white collar or blue. Google "gender discrimination study". Ask some professional women you know. Ask your significant other or women in your family. Look around your workplace and witness how people act toward and talk about women versus the way they act toward and talk about men.
It is getting better, but this isn't controversial: Gender discrimination exists.
There's no conspiracy or collusion (well, rarely, see Tailhook for example, or the wall street folks taking a business lunch at Hooters or taking a client to a strip club). It often isn't even malicious or intentional (I'm thinking of that study that showed people are more comfortable with and most likely to hire people like themselves, but my google fu fails me). But it certainly happens, in subtle and not-so-subtle ways, and this certainly has a "serious" impact on women's work life and careers.
(By the way there's also gender discrimination that happens in the other direction: try to become a nanny, day care worker or pre-school teacher as a man, or sue for custody of the children during a divorce.)
That's the saddest thing I've read in a while: a depressing combination of the mundane, touching and tragic. It's like Hemingway's 6 word story, there's so much drama packed into some of those phrases.
I wonder why many of the quotes have expletives masked as #, but then there's this one stuck in there: "That’s all guys! Fuck!"