Sup BHW,
Computational linguist here, recently fascinated by text-to-text generation. It's an awesome field, but there aren't too many applications for it beyond text summarizers... or at least that's what I assumed before I came across Blackhat SEO. I'm still getting a hold of the whole idea, but I'm mainly interested in the "state of the art" of the tools used in this business. There's actually some pretty interesting stuff. Of course, I also think there's also a lot of area for improvement, heh.
I actually worked on spam filters for a short period of time (just 'opinion spam' though) and this got me interested in how well-equipped the "other side" was. I came across spinners, thought "I could do better than that" and started prototyping away, though this was before I figured out the actual difference between "SEO" versus "spam" and what its aims are, and before I had a look at some of the more interesting commercial tools advertised here.
I'm a hardcore Pythonist (unfortunately - because the only RL expectation of Python programmers is making Django websites), but I'm pretty good at making black boxes that take data in and do text processing magic with it. As an NLP guy, expect me to complain about everything not being data-driven enough and demand precision and recall statistics instead of "200% true user reviews" (...yes, that's just a joke). Also expect me to hopefully show off some stuff in the future.
SPEAKING OF WHICH: What are the exact rules on posting URLs by new members - are they totally banned, or are they allowed when obfuscated? Sorry, I seriously couldn't find anything about this in the rules/FAQ. I wanted to share something in Tools for my first real post, but yeah. I don't imagine I'll be discussing too much blackhat SEO itself, really.
(...and apparently this forum's spam/profanity filter is triggered by mere presence of certain words - that's not data-driven enough for me!)
Computational linguist here, recently fascinated by text-to-text generation. It's an awesome field, but there aren't too many applications for it beyond text summarizers... or at least that's what I assumed before I came across Blackhat SEO. I'm still getting a hold of the whole idea, but I'm mainly interested in the "state of the art" of the tools used in this business. There's actually some pretty interesting stuff. Of course, I also think there's also a lot of area for improvement, heh.
I actually worked on spam filters for a short period of time (just 'opinion spam' though) and this got me interested in how well-equipped the "other side" was. I came across spinners, thought "I could do better than that" and started prototyping away, though this was before I figured out the actual difference between "SEO" versus "spam" and what its aims are, and before I had a look at some of the more interesting commercial tools advertised here.
I'm a hardcore Pythonist (unfortunately - because the only RL expectation of Python programmers is making Django websites), but I'm pretty good at making black boxes that take data in and do text processing magic with it. As an NLP guy, expect me to complain about everything not being data-driven enough and demand precision and recall statistics instead of "200% true user reviews" (...yes, that's just a joke). Also expect me to hopefully show off some stuff in the future.
SPEAKING OF WHICH: What are the exact rules on posting URLs by new members - are they totally banned, or are they allowed when obfuscated? Sorry, I seriously couldn't find anything about this in the rules/FAQ. I wanted to share something in Tools for my first real post, but yeah. I don't imagine I'll be discussing too much blackhat SEO itself, really.
(...and apparently this forum's spam/profanity filter is triggered by mere presence of certain words - that's not data-driven enough for me!)
