Sat Dec 29 01:03:46 PST 2018

180
Ranking Search Results by Reranking the Results Based on Local Inter-
Connectivity
That subheading probably sounds like a handful, but it is the name of a patent
Google filed. The patent is based on finding a good initial set of results (say the top
1,000 or so most relevant results
then reranking those results based on how well
sites are linked to from within that community.
If you have many links and have been mixing your anchor text but still can not
break into the top results, then you likely need to build links from some of the top
ranked results to boost your LocalRank. Just a few in community links can make a
big difference to where you rank. A site that has some authority but lacks the in
community links may get re-ranked to the bottom of the search results. A site that
has abundant authority, like Wikipedia, probably does not need many in
community links.
Topic-Sensitive PageRank (TSPR

Topic-Sensitive PageRank biases both the query and the relevancy of returned
documents based upon the perceived topical context of the query. The query
context can be determined based on search history, user-defined input (such as
search personalization?try Google Labs Search Personalization if you are
interested
or related information in the document from which the query came (if
people searched Google from a site search box, for example
.
Topic-Sensitive PageRank for each page can be calculated offline. Using an
exceptionally coarse topic set (for example, the top level Open Directory Project
categories
still allows Topic-Sensitive PageRank to significantly enhance relevancy
over using PageRank alone; however, TSPR can be applied more specifically as
well.
Since much of it is calculated offline, Topic-Specific PageRank can also be rolled
into other relevancy algorithms that are calculated in near real time.
I do not think it is exceptionally important for most webmasters to deeply
understand TSPR, other than to understand the intent of this algorithm. Instead of
grading the web on the whole, they would prefer to evaluate it based upon local
topical communities.
Latent Semantic Indexing (LSI

Latent semantic indexing allows machines to understand language by looking at it
from a purely mathematical viewpoint. Here is a brief description of how it works:
Latent semantic indexing adds an important step to the
document indexing process. In addition to recording which
keywords a document contains, the method examines the
document collection as a whole, to see which other documents
contain some of those same words. LSI considers documents
that have many words in common to be semantically close, and