This article is about
There are many themes regarding SEO, but there are also many
false myths that circulate, both on the net and outside. In this study we will
consider only a few of them, namely those that normally occur more often or
that need further clarification.
In this article:
1. Understanding
Google
2. LSI
3. TF-IDF
4. TrustRank
5. RankBrain
1. Understanding
Google
Google is among the most used search engines in the world
and is the reference point in Europe and North America to bring the
predetermined public to the websites of the companies. It is no coincidence
that it is the focus of SEO campaigns and one of the many myths of this sector.
For this reason, it is preferable to learn as much as possible from the perspective
of the business analyst.
Fortunately, however, Google provides a lot of information
on how it works and has often provided sources to its users to learn more. For
example, there are very useful twitter accounts, such as those belonging to
John Mueller, Gary Illyes and Danny Sullivan, or Think with Google, Google devs
and Google AI. If you prefer blogs, then you can choose between Blog.google,
Webmasters.googleblog and AI.googleblog.com.
2. LSI
This acronym stands for Latent Semantic Indexing, which is
an indexing that works best with a narrow statistical dataset. This method was
patented by Bell Laboratories in the 1980s, well before the invention of the
World Wide Web. Specifically, it showed that whenever new information was added
to a body of data indexed using the LSI, that indexing would had to repeat.
Being precisely a technology prior to the web, it could not anticipate how to
index it.
Often this acronym appears in some Google patents, but not
necessarily to be the method used for the data of that index. However, there is
a new tool called LSI Keywords, but which does not exploit this technology nor
produce keywords, but rather related words that could be found on the same page
as the keyword already chosen for that page. In particular, with this method
Google will rewrite the query sought by users to show pages that it believes
may meet information or situational needs with the content that basically means
the same, which is a bit of the thought behind Google Hummingbird.
3. TF-IDF
Like the LSI, TF-DFI is an old indexing method invented
before the advent of the World Wide Web. Look at the frequency of a term in a
document and how often it appears in a corpus of indexed documents. This will
reveal whether a page is about a concrete term and how common or popular it is
in the same corpus. However, it does not include words that appear often and
have no meaning of their own, such as prepositions and articles.
However, this approach has been replaced by a more advanced
algorithm, known by the name of BM25. The TF-IDF is often referred to as part
of a process to identify query improvements that appear at the bottom of
Google's search results, but it is never associated with a part of how pages
are indexed on the web.
There are tools created for people who build pages for
websites. For example, these tools take the indexed query terms for a page and
pass them through a TF-IDF process in pages that rank first for those terms, so
as to create pages that compare well compared to those well classified pages.
4. TrustRank
The idea behind this process is to identify spam pages on
the web. The project was born out of a joint venture between Yahoo researchers
and Stanford University. Having first appeared on this institute's website,
many have mistakenly thought it was connected to Google, since Google was born
thanks to researchers and students of Stanford University.
TrustRank is not used to classify web pages, just as
Google's patented based approach has nothing to do with that invented by Yahoo.
Finally, she referred to reliability in her Quality Rater's Guidelines, but it
has nothing to do with this topic.
Also, be careful what you read about TrustRank. In fact,
often verifiable facts are mixed with hasty generalizations and with completely
disagreeing evidence to support the claim that Google uses this method to
classify pages (which not even TrustRank does).
5. RankBrain
Google has a long history regarding rewrite queries, dating
back to 2003. In this year they patented a method that used synonyms of queries
saved by the search engine, finding terms that could be substitutes or synonyms
of the original term that someone chooses for do a search. However, an update
called RankBrain was introduced.
The rewrite is based on a Word Vector technology developed
by the Google Brain team. However, the company has officially announced that
web pages cannot be optimized for this technology. Unfortunately, however, many
articles have appeared on the internet and many of the tips that are brought
are useful in themselves, but they are not useful for RankBrain.
Read Also: SEO NYC Company
Read Also: SEO Training In Delhi NCR
Comments
Post a Comment