There are several reasons behind the GoMeme concept which shed some light on why this is potentially interesting.
First of all, whenever a site links to an article they find on the Web, they are essentially helping to then promote that article to others. This means they are relevant to the content of that article in some way. The use of the Path List at the end of the GoMeme enables every site that helps to market the article to then get some of the credit for the article’s spread. The credit comes back to them in the form of links to them from other downstream sites that later post the article. This has the effect of increasing the each site in the path’s visibilty for the terms in the article as well as for the terms in the sites in the path list. If a site helps to syndicate content around the Web, doesn’t it make sense that they ought to appear in the history of that content as it spreads beyond them? Isn’t the path of information useful?
Consider an analogy — such as social networking sites like LinkedIn. They routinely utilize and share the path list for every message that they route. The path that the message took is in fact the key to determining whether the message is relevant at each node! Why shouldn’t the same thing take place with any piece of syndicated microcontent?
In the future every piece of content might be a GoMeme, and in that case, GoMemes would primarily spread only via sites that were explicitly relevant to their content (because if everything was a GoMeme, sites would have to choose which ones to post — it would no longer be a novelty, they would post them according to same criteria by which they post any article). Thus, the GoMeme would have the effect of spreading mainly to relevant sites, and thus the links back to the referrers in the Path List would be relevant links. This would effectively help to improve their rankings in search engines in proportion to their “timeliness” in finding “hot” articles early, and their “influence” in spreading them to other relevant sites. So sites that find things early, or spread things the most, would start to rise in the ranks. This makes intuitive sense — these sites are more expert than others or at least they are more “active.” This would enable sites that find things early and spread them widely to get preferential rankings in search engine results on relevant topics, which would make search engine results better in fact — you would see the sites that are more active first.
This method also provides more useful information to search engines. Google’s PageRank algorithm attempts to estimate the value of each site based on the number of other sites that point to that site and their estimated value; but because Path Lists do not exist for most articles on the Web it cannot see how influential a site is in the spread of a particular piece of syndicated content. As a result, as articles spread from site to site parties tend to either just cite the site they got them from, or to not cite anyone at all. This means that search engines cannot see how an article spread with very much resolution, if at all. As content is increasingly syndicated from blog to blog, this becomes a problem because the history (the path) of the content is essentially lost as it spreads. One reason is that search engine indexing is not frequent enough. Another reason is that no matter how frequent it is, a search engine cannot infer what site a given site found an article from , unless there is a citation attached. The GoMeme method provides an incentive for sites to forward the Path List for an article onwards — by doing so, sites can get credit for their roles in helping to distribute articles.
The GoMeme method is reminiscent of Hebbian Learning — a method for learning in neural networks that strengthens connections between neurons that interact. In this case the neurons are weblogs. The connections are strengenthed indirectly because by giving webmasters the ability to see the list of sites they get things from, and the list of sites that get things from them, beyond just 1 hop, they can then find new sites that are relevant to their interests to read and subscribe to. This has the effect of transforming indirect connections to direct connections. Sites that get things indirectly, can subscribe to get things directly. Sites can also subscribe to sites that get things from them. This improves the flow of information by enabling it to flow more directly among relevant sites.
Another point worth mentioning: The process of GoMeme propagation is not random – in fact every site that participates in hosting a given GoMeme MUST be co-relevant in some way. Think about it this way — if you find something and help to spread it, then it *is* relevant to you in *some way* otherwise you wouldn’t have picked it up. Likewise, if someone gets the meme from you, then they must have been reading your content, which means that your content is relevant to them in some way. This is transitive, so as it progresses further down the line relevance is preserved. If x is relevant to y, and y is relevant to z, then x is relevant to z. So the results are not actually irrelevant — even if there is no explicit expression of relevance on the individual pages in the sequence. The fact that they all participated in a sequence concerning the same article is what makes them relevant. In fact the meme may reveal relevance where none was apparrent before. The question is “how are they relevant?” — well they are all interested in the concept and possibly the content of the meme. They are also all connected via some set of sites that exchange information. So in fact this information can be used to discover relevance and connectedness among sites that was previously not necessarily explicit. This is additional information that can be useful to search engines.