SEMLogic Explained
The author's views are entirely their own (excluding the unlikely event of hypnosis) and may not always reflect the views of Moz.
I owe a big thank you to Mike Marshall from Fortune Interactive, who walked me through their SEM Logic product today (demo page here). The technology is very cool, and I'm thrilled that they're so open about sharing it with me and the SEO community. I'm a big believer (as you all know) in sharing your competitive intelligence and data - way to go Fortune!
The idea behind their product is to find the terms and phrases most semantically connected to the terms and phrases you want to target in your pages. The goal is to have an ordered list of the best terms to use on your page, in searching for linking partners and when creating anchor text. It's a scientific approach that I've been working on since I first read Dr. Garcia's work on the subject (yes, I know I point to this all the time).
Fortune's approach is more manageable and simple than what we've been attempting, though. Whereas we try to use the entire index of Google, Yahoo! or another search engine to power results, SEM Logic uses a document set created specifically for each term. They grab documents off the web from major competitors on the subject, high ranking sites and other places they feel are particularly relevant - a very smart way to get around the fact that pulling data from search engines is like pulling teeth.
Once they've created their training set, they build a term vector space around the collection and proceed to measure the semantic connectivity of each term to the target term and each other. Hence the 3D scatter graphs showing clusters of words and phrases that are grouped into certain areas. The idea is that the mathematical calculations of the software will produce the ordered list, showing most relevant terms and phrases to least.
From here, you can take the terms they spit out, use them on your page in relative frequencies (that match the levels from the tool's results) and also apply them to link building campaigns for anchor text and targeting the right link partners. Pretty sweet, huh?
Comments
Please keep your comments TAGFEE by following the community etiquette
Comments are closed. Got a burning question? Head to our Q&A section to start a new conversation.