Documentation
¶
Overview ¶
******************************************************************************
- DANIEL'S ALGORITHM IMPLEMENTAIONS *
- /\ | _ _ ._ o _|_ |_ ._ _ _
- /--\ | (_| (_) | | |_ | | | | | _>
- _| *
- CHINESE WORD SEGMENTATION *
- Features:
- 1. based on Hidden Markov Model
- 2. definition the states in B,E,M,S
- 3. solve the hidden states with viterbi algorithm. *
- http://en.wikipedia.org/wiki/Hidden_Markov_model
- http://en.wikipedia.org/wiki/Viterbi_algorithm * *****************************************************************************
Index ¶
Constants ¶
View Source
const ( SINGLE = 0 BEGIN = 1 MIDDLE = 2 END = 3 )
B,E,M,S
Variables ¶
View Source
var ( // the start probability of state // SP = [4]float64{0.23101714745, 0.76898285255, 0.0, 0.0} SP = [4]float64{0, 0, 0, 0} TP = [4][4]float64{ {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, } IGNORES = ",. ;。，；" )
View Source
var ( SPCount = [4]float64{0, 0, 0, 0} TPCount = [4][4]float64{ {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}, {0, 0, 0, 0}} )
Functions ¶
Types ¶
type WordEP ¶
type WordEP struct {
EP [4]float64 // the emission probability in each state
SC [4]uint32 // the count of appearence in each state
}
------------------------------------------------------ 单个字的观察
type Wordseg ¶
type Wordseg struct {
// contains filtered or unexported fields
}
func Default ¶
func Default() *Wordseg
---------------------------------------------------------- return default wordseg
func NewWordseg ¶
func NewWordseg() *Wordseg
Source Files
Click to show internal directories.
Click to hide internal directories.