ONLINE MARKETING > SEARCH ENGINE MARKETING > SEARCH ENGINE OPTIMIZATION PageRank Puzzle 3 YOU ARE HERE!

GOOGLE PAGERANK PUZZLE 3

The New Enhanced Google PageRank 3: More Missing Parts

Expert website optimization and promotion involve continual research, learning, and utilizing that knowledge to ensure the fulfillment of your website marketing plans. The recent New PageRank analyses, conducted by well known search gurus, are the most relevant sources for information on the New PageRank algorithm. Here is the overview of the basic data and influencing factors, utilized as input for most interesting PageRank expert analysis.

HILLTOP / EXPERT SYSTEM ALGORITHM

Hilltop is an algorithm utilizes an 'expert' system to rank pages, and compiles an index of expert web pages. The Hilltop's authors defined "an expert page as a page that is about a certain topic and has links to many nonaffiliated pages on that topic. When a search query is performed, "a lookup is done on the expert-index to find and rank matching expert pages."

Hilltop counts the number of incoming links from topic / theme / subject relevant pages (expert pages) to a content-rich web pages, and gives more weight to links referenced by larger authoritative web sites. "The best expert pages on the query topic" get the highest score, and are authority sites, therefore ranked the highest. This concept is often referred to as LocalScore and Local PageRank, and based on similar aspects of context/topic valued linking as Topic Sensitive PageRank.

Taking links referred from topic /theme /subject relevant webpages into consideration would obviously be of greater value to a searcher, but it also helps in determining websites interrelations, such as detecting affiliated, heavy cross-linking or similar website strategies.

Google has been contributing to Hilltop research project for a long time, and the author of Hilltop, Krishna Bharat, was employed by Google. Bharat filed for the Hilltop patent in January of 2001, with Google as an assignee. Hilltop's author and others at Google probably have significantly improved this algorithm since its original version in 2000. (Hilltop: A Search Engine based on Expert Documents - Krishna Bharat & George A. Mihaila)

TOPIC-SENSITIVE PAGERANK

Topic Sensitive PageRank (TSPR) concept is also based on topic / theme / subject evaluated linking and authoritative reranking.

Google PageRank™ system takes all quality links into the consideration, neglecting to take the topic of the page generating the link into account. That way, Google algorithm, based on the normal probability distribution, simulates a random searcher that follows random links, but finding more popular websites more likely.

In reality, searchers are not taking a random path - they are looking for information on a specific topic. Google has tried to overcome this limitation by implementing the link anchor text factor, but spammers utilize a bulk of keyword stuffed links all over the website, along with buying and trading links with topic unrelated sites, which makes finding quality results very hard.

Topic-Sensitive PageRank solves the above problem by attributing bias or preference, for certain kind of pages to random searcher, who is still most likely to follow relevant links from relevant pages, but topic-specific.

Topic-Sensitive PageRank has been researched, developed and presented by Stanford University computer-science researcher team in 2002 (several research papers, by Taher Haveliwala, Sepandar Kamvar and some more Stanford researchers, are available on Stanford Website). The researchers formed a company called Kaltix, acquired by Google in September 2003.

Kaltix technology also includes several other new techniques (BlockRank, Adaptive PageRank, Quadratic Extrapolation method), which enable up to five times faster computing of the PageRank and implementation of the Topic-Sensitive PageRank concept.

According to the experts, this is not enough to make personalized ranking possible, as PageRank would need to be about 10 times faster than now. But the Kaltiix enhancement "is good enough for a topic-based PageRank. And all of these algorithms are under a page of code long, so it could take a day to implement all three, if that", according to Kamvar, one of the authors.

APPLIED SEMANTICS' CIRCA TECHNOLOGY

Applied Semantics' AdSense technology is well known, as it's utilized by Google Adwords PPC system for delivering contextually based ads (the ads change based on the content of the page). But, their basic technology is CIRCA technology, initially a part of a customer relationship management (CRM) software:

"CIRCA Technology is based on a language-independent, highly scalable ontology that consists of millions of words, their meanings, and their conceptual relationships to other meanings in the human language."

Ontology is a large database that stores information, including synonyms, antonyms, common misspellings, alternative spellings, similarity, stemming, and other word relationship principles, and enables the interpretation of a text.

"The ontology, aided by sophisticated search technology, is the basis for a conceptual understanding of the multiplicity of word meanings, enabling computers to more effectively manage and retrieve information which results in improved knowledge discovery opportunities for searchers."

In short, CIRCA software applies ontological principles to a webpage to determine its context, and classifies a page (or website) into various categories, simulating the functioning of a human brain.

"Humans are searching for ideas, while automated systems are limited to searching for words. Applied Semantics is working towards bringing this "human approach" into the realm of an automated system, allowing individuals to use information more effectively." (Applied Semantics Technical White Paper)

Furthermore, as a scalable technology, the ontology incorporates self-learning algorithms, that enables the system to correct itself and learn by observing meanings used in context.

Expert analyses of the new, enhanced Google ranking algorithm, conducted by well known search engine gurus, and explained in laymen's terms. PageRank Puzzle 3 >>

User Oriented Design & Ethical Website Marketing Strategies by Distinctia, 2006