1 The application of bigram information to the lattice is shown in fig 4.3.

2 Fig 4.11 outlines the way in which the bigram and matrices are used in conjunction.

3 During the training phase the probabilities of bigram and trigram transitions between grammatical tags are determined.

4 It uses a 997 word lexicon, and a bi-gram grammar extracted from 900 test sentence templates.

5 This method preserves the strong power of bigram to distinguish class, while reduces the extent of data sparseness.

6 And the syllable bigram language model to the recognition system will obtain the rate improvement.

7 We extend the bigram OOV seeds by the left and right (LR) neighbors on the basis of OOV border judgment. It helps to identify OOV with integrated meaning without length restriction.

8 In those texts, we select bigram as feature after Chinese word segmentation, deleting stop word and other process.

9 Found a large number of high-degree overlapped bigrams and high-degree biased bigrams existing in bigram feature set.

10 The method used is based on collecting and clustering bigram statistics using a rank correlation metric.

11 You already have a little information from the grammar: The bigram format calls for two switch statements, which are the minors nested inside the majors.

12 Secondly, this paper presents a hierarchical text filtering approach based on bigram in the off-line filtering module.

13 Thirdly, this paper proposes a new method to extract bigram as features based on illegal keywords.