Read this in other languages: Русский, Português
A bloom filter is a space-efficient probabilistic data structure designed to test whether an element is present in a set. It is designed to be blazingly fast and use minimal memory at the cost of potential false positives. False positive matches are possible, but false negatives are not – in other words, a query returns either "possibly in set" or "definitely not in set".
Bloom proposed the technique for applications where the amount of source data would require an impractically large amount of memory if "conventional" error-free hashing techniques were applied.
An empty Bloom filter is a bit array of m bits, all
set to 0. There must also be k different hash functions
defined, each of which maps or hashes some set element to
one of the m array positions, generating a uniform random
distribution. Typically, k is a constant, much smaller
than m, which is proportional to the number of elements
to be added; the precise choice of k and the constant of
proportionality of m are determined by the intended
false positive rate of the filter.
Here is an example of a Bloom filter, representing the
set {x, y, z}. The colored arrows show the positions
in the bit array that each set element is mapped to. The
element w is not in the set {x, y, z}, because it
hashes to one bit-array position containing 0. For
this figure, m = 18 and k = 3.
There are two main operations a bloom filter can perform: insertion and search. Search may result in false positives. Deletion is not possible.
In other words, the filter can take in items. When we go to check if an item has previously been inserted, it can tell us either "no" or "maybe".
Both insertion and search are O(1) operations.
A bloom filter is created by allotting a certain size.
In our example, we use 100 as a default length. All
locations are initialized to false.
During insertion, a number of hash functions,
in our case 3 hash functions, are used to create
hashes of the input. These hash functions output
indexes. At every index received, we simply change
the value in our bloom filter to true.
During a search, the same hash functions are called
and used to hash the input. We then check if the
indexes received all have a value of true inside
our bloom filter. If they all have a value of
true, we know that the bloom filter may have had
the value previously inserted.
However, it's not certain, because it's possible
that other values previously inserted flipped the
values to true. The values aren't necessarily
true due to the item currently being searched for.
Absolute certainty is impossible unless only a single
item has previously been inserted.
While checking the bloom filter for the indexes
returned by our hash functions, if even one of them
has a value of false, we definitively know that the
item was not previously inserted.
The probability of false positives is determined by three factors: the size of the bloom filter, the number of hash functions we use, and the number of items that have been inserted into the filter.
The formula to calculate probablity of a false positive is:
( 1 - e -kn/m ) k
k = number of hash functions
m = filter size
n = number of items inserted
These variables, k, m, and n, should be picked based
on how acceptable false positives are. If the values
are picked and the resulting probability is too high,
the values should be tweaked and the probability
re-calculated.
A bloom filter can be used on a blogging website. If the goal is to show readers only articles that they have never seen before, a bloom filter is perfect. It can store hashed values based on the articles. After a user reads a few articles, they can be inserted into the filter. The next time the user visits the site, those articles can be filtered out of the results.
Some articles will inevitably be filtered out by mistake, but the cost is acceptable. It's ok if a user never sees a few articles as long as they have other, brand new ones to see every time they visit the site.