点评详情
发布于:2018-10-8 17:05:17  访问:146 次 回复:0 篇
版主管理 | 推荐 | 删除 | 删除并扣分
E most likely path by way of the token lattice given some language
adapted Viterbi [12]).Figure 1 Tokenizer LGK974 Description components and data flow A diagram illustrating the tokenizer‘s elements and data flow by way of these elements.Barrett and Weber-Jahnke BMC Bioinformatics 2011, 12(Suppl 3):S1 http://www.biomedcentral.com/1471-2105/12/S3/SPage five ofFigure two A bounded lattice representing a sentence‘s segmentations An instance of a bounded lattice representing a sentence‘s segmentations.To construct a token lattice, a lattice constructor applies just about every transducer to every single character position within the input text. Require: graph , index , l, sequence if l = 0 then transducer could not identify token sequence LY2874455 order return end if currentNode index lastNode index + l if length ( sequence ) = 0 then empty edge if missingEmptyEdge ( graph , currentNode, lastNode addEmptyEdge ( graph , currentNode, lastNode)) thenendif return endif sequenceIndex 0 although sequenceIndex < length ( sequence ) do token sequence [sequenceIndex ] nextNode None if sequenceIndex + 1 length ( sequence ) then nextNode lastNode end if edge findEdge ( graph , currentNode, nextNode, token if edge = None then edge missing, create edge addEdge ( graph , currentNode, nextNode, token) edge could exist)end if sequenceIndex sequenceIndex + 1 currentNode = nextNode end whileAlgorithm 3 The following algorithm, cleanup, removes unneeded edges from the token lattice (required by Algorithm 1). Require: graph , node, isStart if isStart then beforeLastNode node - 1 for n = beforeLastNode to 0 do work backwards through the nodes cleanup ( graph , n , false ) end for else if node in graph AND outDegree ( graph , node ) = 0 then lattice path does not reach last node preds predecessors ( graph , node removeNode ( graph, node for p in preds do cleanup ( graph PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/27385778 , p, false ) finish for end if finish ifOf the 3 described software program elements, only the lattice constructor‘s pseudocode is presented. This really is due to token transducer code being certain to a token transducer‘s objective and due to existing documentation of a best-path selection algorithm (e.g.E most likely path by means of the token lattice offered some language model. An algorithm exists for ideal path selection (e.g. adapted Viterbi [12]).Figure 1 Tokenizer elements and info flow A diagram illustrating the tokenizer‘s components and facts flow via these elements.Barrett and Weber-Jahnke BMC Bioinformatics 2011, 12(Suppl three):S1 http://www.biomedcentral.com/1471-2105/12/S3/SPage 5 ofFigure two A bounded lattice representing a sentence‘s segmentations An instance of a bounded lattice representing a sentence‘s segmentations.To construct a token lattice, a lattice constructor applies every single transducer to every character position within the input text. The result of applying a lattice constructor on "The patient‘s 10mg tablet." is seen in Figure two. Provided: ? is a finite set of symbols referred to as the alphabet. ?S is the set of all finite strings more than . ?G := (V, E) is a directed graph consisting of a finite set of vertices as well as a finite set of labelled edges, E V ?S ?V. The token lattice G is constructed for some text s S as follows: ?Let L := i : i N0, 0 i PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/27324125 length(s). ?s is actually a slice of s; s := s[i : length(s)] offered an i L. ?vi V for i L.
共0篇回复 每页10篇 页次:1/1
共0篇回复 每页10篇 页次:1/1
我要回复
回复内容
验 证 码
看不清?更换一张
匿名发表 
当前位置
脚注信息

美博梁sir2010-2017版权所有