Measure Value of External Link (a technical approach)

You are here: Articles > Measure Value of External Link (a technical approach)

"Scrap"* or "Stub" by Carsten Cumbrowski, March, 2nd 2006

"I also believe, that a link to DMOZ is okay if you need to add an external link (that is not always necessary, depending on the Artilcle) until more suitable links are available. I know, that this is the basic issue, to determine what is suitable and what not. There is no definite answer to that (it's in its nature). You can improve the process though to get pretty decent results. "

Multiple "characteristics" must be considered and meassured based on "properties".

1. Intention

Probability that the Link was added for the purpose to add value (extended content and facts) or proof of a fact/reference/source Though one. This can only be determined via reducing the probability that a different intention was the trigger for the addition.

a) Search Marketing SEO

If done well, hard for a person that is not knowledgeable about the subject to detect, but the factor, which probability can be reduced the easiest by technical measures (se "no follow" and de-link proposals)

b) Self Promotion

Sometimes hard for anonymous additions, easier if done by user where some information were provided. Easy to detect by human if article in Wikipedia is generic, but site/page that was linked to is specific about a person or thing NOT referenced to and mentioned in the Article Text Itself.

c) Denunciation / Jest (e.g. Person A adds link to Person B's Home Page to Article pig)

Easy to detect

-> The probability of right intention could be increased significantly by technical measure. It can't be determined today thus pessimistic number is applied by editors (assuming a wrong intention by the person that adds the link rather than good). Other existing factors today are status of user in Wikipedia (Admin, Snr. Editor) and "trust" by other Editors (Editor proved his authority and intention to the Wikipedia over a period of time)

2. Quality

probability that linked to site has purpose and quality worth to visit (= no junk)

a) Who adds the link

Does the person that adds the link has knowledge about the subject or not. Hard to measure 100%. Indicators are editors past contributions to content in the same category activities or "Flag" of User as verified authority for general subject xyz (requires that editors identity is known and his real live background. Publicly accepted "Expert" for the Subject (not self proclaimed). Education verification (PhD etc.)

b) Who Removes it

Same methods as for a)

c) How often is it clicked compared to other external Links

Technical Property which could be tracked by Wikipedia

d) How long does user stay's on site? Does he returns to Article via Browser Back Button or following other Links on that Site

Google is tracking it already and improving the tracking. They work on making it an important factor for their Ranking. Hard for Wikipedia to do, but who knows the future

e) Removed because of 1. by editor with low 2 b) but gets re-added by defend 2 a) frequently

If Editor(s) is/are known not be an Expert for the Subject and removes an external reverence which gets re-added by different people over and over again, the chances are good that all the people who add the link are right and the editor(s) wrong

3. Relevance

probability that the linked to page/site adds value to the article and is beneficial to the user reading it (learn/find out more about it)

a) 2c) to 2e) can be applied

b) Statistical Comparison between Article , the Categories the Article is assigned to and other Articles in same Category as current Article and target Page and general Website content. Something that could be done via programming based on Search Engine Algorithms

This is very rough, but an Idea for starters and kick around.

< previous Article/Stub	<< Index	next Article/Stub >
Tips for Affiliate Marketing Newbie's - Is Pay Per Click a Viable Option?	Articles Index	PPC Conversion Tracking for Affiliates