2. The aim of this article is to propose a noble method to store visual, sound and memory data in such a
non linear way that it becomes very easy to get the requested data in any situation. This method
proposes to create many sections of data. Each section will have some data. Create some different sized
sections of the main data. Size of the sections will stay within a limited range. Insert data to all the
sections randomly. We will not know which data will reach to which section. We will also have some
stations which will help us to find a data. Each station will be connected to some sections. Each station
will have an individual ranking and also each station will have its own ranking for each of its connected
section. One section could be connected to many stations.
Initially, all stations individual ranking will be zero and it will have zero number of connected section.
When searching for a specific data, it will select any unchecked station randomly and search for the data
there. The station will send the request to many sections. Initially, it will send this request to all sections
by turn and when data will be found in a section, the station will create a connection with the section
and give it a ranking. When each station will have a decent connected network, it will only search for the
data in its own network. Failure to find the data will decrease the stations ranking and if the station finds
the data, its ranking will go up. So, any station will have many ranked connections. Not all the
connections will have same weights. If one section finds data more than others, its weight will be higher
than others. Each time same connection gets a data, ranking increases for that connection.
To increase its ranking, each station can withdraw some connection with low ranking and search for new
section. This advantage will be given to low ranked stations to bring up their ranking. When to give this
advantage to low ranked stations, that will be decided by the performances of higher ranked stations.
When higher ranked stations will not perform well, in that situation, lower ranked stations will get a
change. These rules we have to adopt based on the data set we have and nature of search.
If any section which forward request to its connected section and that section fails to get the data, that
section’s individual ranking will decrease. Any section able to get data in its network, that sections
ranking will go up. So, if a section finds a data, its individual ranking will go up. If station A sends a
request to section B and section B finds the data, then station A’s own ranking will go up and also station
A’s ranking for its connection to section B will go up as well.
So, any request for data will go to high ranked sections first and then to lower ranked sections. If many
sections are available as having same rank, chose one randomly from those. Data will be searched in
high ranked networks and then to lower ranked networks. So, any station will have many connections
with different ranking from different sections. Any station will have its own ranked connections
(connected network). Basically if any section has data which is asked frequently, then there will be many
4. receives very frequently. Even a complex visual or sound data becomes very tolerable or easily
understood as that is getting searched very quickly.
When people move to a new community, face data of that community gets search request frequently.
For that, the stations which have their data, their ranking goes up and people can recognize them
quickly and those which we are not seeing frequently, their ranking goes down. Whichever data gets
requested more time, its ranking goes up, otherwise goes down. This is very good, because that is how
brain is still able to do the frequently asked tasks quickly. And when time comes to change it, it can also
change it very easily.