undefined | Better HN

0 pointsfrozenport12y ago0 comments

Same thing.

I got an open ended problem that would take a measure of forever to solve: Specifically to write a program (in your favorite language!) that would do natural language processing and discern emotions and feelings about subjects from text. We were supposed to include our own lexical parser.

The author's strategy scales poorly with complex and interlocking problems. Where it is more important to show potential ability to solve the problem (puzzles), holistic understanding (talkin' about it), or prioir experience (specific questions about libraries).

This guy should address his core problem (stage anxiety?) instead of advocating a solution that is only tenable in a limited number of cases and represents far too much commitment from the participants.

0 comments

1 comments · 1 top-level

jarek12y ago

> I got an open ended problem that would take a measure of forever to solve: Specifically to write a program (in your favorite language!) that would do natural language processing and discern emotions and feelings about subjects from text. We were supposed to include our own lexical parser.

I know of a YC alumni company that asks for a weekend project on roughly that level and when rejected candidates ask if they can upload their work to Github, the company declines because they'd "like to keep an even playing field for future candidates."

Then you're left with choosing between having an awesome project you can't show off and being an inconsiderate jerk who uploaded code despite being asked not to.

Doesn't really make one enthusiastic for the next time you're asked to do an "exercise" or a "project."

j / k navigate · click thread line to collapse