Comments on: Semantic Keyword Research with KNIME and Social Media Data Mining – #BrightonSEO 2015

By: Christian

Christian — Fri, 31 May 2019 09:58:19 +0000

Just read this, again.

Paul.. you are an inspiration.

One of the worlds best.

All good, Christian

By: Paul Shapiro

Paul Shapiro — Mon, 12 Sep 2016 14:25:33 +0000

In reply to Koji Kawano. Yeah, an alternative would to be to integrate the actual BoilerPipe library, which unfortunately isn't easy if you're not familiar with Java. I've written the code if you want to figure it out, it's on my GitHub.

By: Daisy Priya

Daisy Priya — Mon, 12 Sep 2016 08:39:07 +0000

Terrific writing and really impressive way to make things clear. Topical and intent based keyword research is very important these days and LSI keywords are playing major role after the Hummingbird update. But the newest change on Google keyword planner tool is making things difficult to select best relevant keywords depending on their average monthly search volumes. Really enjoyed reading the post and couple of valuable points are also noted.

By: Koji Kawano

Koji Kawano — Wed, 22 Jun 2016 21:20:46 +0000

Great post, Paul. Thank you very much for your insight. Unfortunately, I get lots of 403s that prevent me from extracting text. I am assuming BoilerPipe does not want to process a batch the way it is setup in your example. Please let me know if you have alternatives that were not mentioned in the original post.

By: Paul Shapiro

Paul Shapiro — Tue, 21 Jun 2016 18:09:18 +0000

In reply to Todd. Thanks! Once you're familiar, and you have stuff built out, the time investment is VERY minimal. Even if you don't have everything built-out, and you're familiar, it's amazing what you can accomplish in a short period of time.

By: Todd

Todd — Tue, 21 Jun 2016 17:54:44 +0000

Fantastic post, deep, and insightful. Very creative, low-cost solution to getting some deep data/research done!

I am curious, what’s the time commitment usually involved in this process once you’re familiar with KNIME from your experience?

Again, really enjoyed it.

By: Paul Shapiro

Paul Shapiro — Thu, 28 Jan 2016 18:31:24 +0000

In reply to AKi. Big fan of MarketMuse. I highly recommend it to anyone who finds this process too technically advance, or even anyone who would like to go even deeper (the technology is better).

By: Kai

Kai — Tue, 03 Nov 2015 11:02:03 +0000

In reply to Paul Shapiro.

Hello Paul, thanks for the reply and help provided. Indeed I had Version 3.0. I downgraded to yours to have the exact same setting. All nodes are there, but… the ‘Extract and Expand Tweet URLs’ workflow still has issues. I went through each node / process step, executed, checked the result table.
– String Manipulation: Part 1:t.co Extraction RegEx
– String Manipulation: Part 2:t.co Extraction RegEx
In these nodes I adapted the config to work on https which was http before and the result table looked fine afterwards.
– URLResolver
This simply does not seem to work anymore, the ‘Resolved URL’ column is exactly like the input column.
So unfortunately, at least so far, this is the furthest I could get. If you have any advise, highly appreciated.
Cheers, Kai

By: Paul Shapiro

Paul Shapiro — Wed, 28 Oct 2015 13:42:25 +0000

In reply to Kai. I'm running the latest stable version (2.1.2.0) and they're included. My guess is that you downloaded the Early Access 3.0 version? Or downloaded the version without all of the extensions? Another thing that's possible, is that they're included in the 32bit version and not the 64bit versions? Alternatively, these are all under the Community Nodes->Palladian section and you should be able to download them under Help->Install New Software... menu.

By: Kai

Kai — Wed, 28 Oct 2015 13:36:49 +0000

Paul, what an excellent article! I was searching for a practical approach to something like semantic keyword clouds and as far as I can tell, this is the sweet spot for it. However, after downloading & installing the complete KNIME and giving myself a try at the ‘extract and crawl twitter links’ workflow, I run into trouble with three nodes, that do not seem to be included in the node repository anymore
– HttpRetriver
– UrlExtractor
– UrlResolver
Does this also happen to you, or am I missing an additional node package I need to install, anything?
Thanks in advance.