8.5 Huffman Code

Objective: Compress text data
Approach:
1. Assume each letter is a single-node tree assigned a weight represented by the frequency of the letter in the text
2. Repeatedly merge the trees with the lowest weight
3. Assign to each letter the binary string showing the path to the letter from the root
Codes that are not prefixes of one another are called prefix codes
The algorithm provides optimal prefix codes

o
---30 ----
-0--- --1-
18o-- --12o
|| ||| ||||
---0|| |1 --0| 1|-|
a:10| |o e:7| b:5|
----- |8|| ---- ----
0|| |1
---|| ||o
-c:4-| 4|
0|| 1
---|| ||-|
d:3-| f:1-

a:00, b:11, c:010, d:0110, e:10, f:0111

[prev] [prev-tail] [front] [up]