当前位置:文档之家› A New Radical Based Approach to Offline Handwritten East-Asian Character Recognition

A New Radical Based Approach to Offline Handwritten East-Asian Character Recognition

A New Radical Based Approach to Offline Handwritten East-Asian Character Recognition
A New Radical Based Approach to Offline Handwritten East-Asian Character Recognition

A New Radical Based Approach to Offline Handwritten

East-Asian Character Recognition

Kumar Chellapilla Patrice Simard Microsoft Research kumarc@https://www.doczj.com/doc/3b6722598.html,

Microsoft Research patrice@https://www.doczj.com/doc/3b6722598.html,

Abstract

East-Asian characters possess a rich hierarchical structure with each character comprising a unique spatial arrangement of radicals (sub-characters). In this paper, we present a new radical based approach for

scaling neural network (NN) recognizers to thousands of East-Asian characters.

The proposed off-line character recognizer

comprises neural networks arranged in a graph. Each NN is one of three types: a radical-at-location (RAL)

recognizer, a gater, or a combiner. Each radical-at-location NN is a convolutional neural network that is designed to processes the whole character image and recognize radicals at a specific location in the character. Example locations include left-half, right-half, top-half,

bottom-half, left-top quadrant, bottom-right quadrant, etc. Segmentation is completely avoided by allowing each RAL classifier to process the whole character image. Gater-NNs reduce the number of NNs that need to be evaluated at runtime and combiner-NNs combine RAL classifier outputs for final recognition.

The proposed approach is tested on a real-world dataset containing 13.4 million handwritten Chinese character samples from 3665 classes. Experimental results indicate that the proposed approach scales well and achieves a low error rate.

Keywords : East-Asian handwritten character recognition, radicals, convolutional neural networks. 1. Introduction

Offline handwritten East-Asian (Chinese, Japanese, and Korean) character recognition is a challenging pattern-recognition problem. Several aspects of these character sets make the problem difficult for conventional machine learning techniques such as Bayesian methods, neural networks, support vector machines, template matching and nearest neighbor algorithms. The most prominent aspect that contributes to the difficulty is the large number of classes/characters. The number of unique CJK (Chinese, Japanese, and Korean) characters can range from 4,000 to 80,000 [1]. Effective classification approaches that use only a single neural network [6] exist for most Latin languages such as English, French, German, etc that have less than a few hundred classes. However, these single neural network based approaches do not directly scale to the thousands of East-Asian characters. Trying to learn more than a couple of thousand characters using stochastic

gradient descent (backpropagation) is a hard problem as the neural network weights do not converge even after several weeks of training.

Radical based approaches to CJK character recognition is not new [1-3]. Common approaches are:

a) extracting features holistically from the character image, or b) structurally decomposing characters into component

parts—usually strokes or radicals (groups of strokes). The latter is almost always preferred as conventional holistic approaches do not scale to the tens of thousands

of CJK characters. The decomposition schemes typically rely on explicit segmentation for strokes or radicals.

Traditionally, radicals are extracted either from the character skeleton or character strokes. Skeleton-based methods [4] first convert the character image into its skeleton graph. The skeleton graph is processed to discover radicals, their spatial relationships, and also character variations. Stroke-based methods [3,5] go one

step further and decompose radicals into primitive strokes, usually straight lines with a few possible slopes. Structural analysis is used to integrate the information contained in these strokes and recognize the character. Computationally, stroke-based methods are faster than

skeleton-based methods, but can be inferior due to ambiguities resulting from intersecting strokes.

In printed form (say from a font in printed documents) the character decomposition is very clear and

one can devise clever approaches to segment the characters into not only radicals (skeleton-based), but

also strokes (stroke-based). The hierarchical structure

can then be exploited in a bottom up manner to recognizer the character. Efficient dynamic programming

(Viterbi) algorithms can be designed to do the composition. Further, not all stroke combinations generate valid radicals, and not all radical compositions produce valid characters. This can be used to further improve the accuracy of the recognizer. Thus, stroke-based and skeleton-based approaches work well on machine printed CJK characters. However, in handwritten form several of the character strokes merge into continuous curves. This also extends to strokes from different radicals in a character. Such merging makes radical and stroke segmentation becomes difficult, if not impossible. This merging is similar to cursive handwriting in Latin based western languages. The merging of strokes and radicals is so

common place that, common handwritten character forms (for each character) have evolved that do not look like their printed counter parts, but allow for easy reading and writing. Hence, the traditional radical and stroke decomposition approaches perform poorly on cursive characters.

In this paper, we propose a new holistic method that exploits the hierarchical nature of East-Asian characters through a divide-and-conquer approach.

2. Radical Decomposition of CJK

Characters

The East-Asian character system is very hierarchical. Each character is made up of one or more strokes (as large as 20). Some of the constituent strokes form commonly occurring sub-characters (parts of characters) called radicals. The hierarchical decomposition has the following properties

a) Characters are made up of one or more radicals b) Constituent radicals occur in specific locations that are unique to the character (for example, left, top, or left-top quadrant, or as a band along the left and top parts of the character, etc) c) Radicals are made up of strokes

Some radicals are also characters, i.e., there exist some characters that comprise a single radical, but not every radical is a character. Two examples of Chinese radicals are presented below:

i) ??(tree) is a radical made up of 4 strokes and is also a character that represents a tree. The order of

strokes in the character is:

ii) ? (eye) is a radical made up of 5 strokes. It is also a character that represents an eye. The order

of strokes in the character is:

Three Chinese character examples are given below:

a) ? is the Chinese character for forest. It is made up of two tree radicals placed next to each other. b) ? is the Chinese character that means “each-other” or mutual. It is made up of a tree radical to the left and an eye radical to the right

c) ? is the Chinese character for “evil-spirit.” It is made up of three radicals ???, with ? showing up in the left-top corner, ? showing up in the right-top corner, and ? showing up at the bottom of the character.

In the current paper, the term radical will be used to refer to any collection of strokes (with one or more strokes) that repeatedly occur as parts of characters. Under this definition, each of the 214 KangXi radicals (defined in the UNICODE standard) is a radical, but it is possible for several other custom radicals to be defined and used.

3. Character layout

Each character is composed of one or more radicals arranged in a special layout. The arrangement takes on very few selective forms. This layout of a character can be effectively captured by a tree representation of the character. An example decomposition of the Chinese character (?) is presented in Figure 1.

(b)

Figure 1. (a) Tree Representation of an example Chinese Character (?) depicting the radical decomposition, and (b) three example radical paths.

This tree contains internal nodes (nodes that have children) and leaf nodes. Each of the leaf nodes is a radical. All possible selective forms can be represented using a few internal nodes called templates. These templates do not represent character pieces, but character layout. Interestingly, though there are tens of thousands of East-Asian characters, with several hundred radicals, the number of templates is only a handful (about a dozen or so). Here are 12 commonly used templates:

????????????

The first template defines the left and right parts of the character, the second defines the top and bottom parts of the character, etc. It is conceivable that other useful templates can be designed on an as needed basis. The tree representation provides interesting ways of

a) decomposing characters into radical-paths, and b) grouping characters based on their (radical, template-path) pairs

3.1. Radical-Paths and Template-Paths

We now define “radical-path” and “template-path” through examples. Three example radical paths in the character ? are presented in Figure 1(b). These three radical paths are represented as a) 0?\0?\? (green), ? b) 0?\1?\?(blue), and ? c) 1?\?(red)

Each node along the path is delimited by a back-slash (\). The different parts of each template are sequentially numbered starting at zero. For example, the template node, ?, contains two children corresponding to the top-part (0?), and bottom-part (1?).

Consider the green radical-path (Figure 1(b)) that the character ? belongs to. This radical-path is a particular instance of the template-path given by

0?\0?\?

where ? denotes a wild-card radical. This template path is a concrete way of capturing the property, “the top-left part of the character.” Note that the corresponding radical path obtained by replacing the wild-card ‘?’ with ? captures the property, “the top-left part of the character contains the radical ?.” The number of nodes in a template-path or a radical-path is its length. Here are a few example template paths of length 3

a) TOP-LEFT = 0?\ 0?\

b) TOP-RIGHT = 0?\ 1?\

c) BOTTOM-LEFT = 1?\ 0?\

d) BOTTOM-RIGHT = 1?\ 1?\

Using radical-paths one can decompose a character into its constituent radical paths. There exists one radical path for each radical present in the character.

3.2.Grouping Characters by Template-Paths

Given a radical path = (radical, template-path) pair one can ask the question: Which are all the characters that contain this (radical, template-path) pair? The answer would be a group of characters that share a particular radical in a given location in the character. A template path can then be viewed as a group of characters containing all possible radicals at a given location. This grouping of characters using (radical, template-path) pairs or simply a (template-path) is key to the radical based architecture of the East-Asian recognizer presented in this paper. Note that these template-path groups are not necessarily disjoint over different template-paths.

4.Distribution of Character Layouts

4.1.Data set

We used a large database of Chinese handwritten characters collected using a tablet-PC for our experiments. The dataset contained a total of 13.4 million samples of 3665 frequently used characters from GB2312 (simplified Chinese) written by 6668 users. The dataset is not uniformly distributed over all 3665 characters, but contains a natural distribution of characters. Character samples were collected in both print and cursive forms. Print data comprised 88% (11.8 million samples) of the data while cursive data comprised 12% (1.6 million samples) of the data.

In our offline recognition experiments, the ink samples are converted to images by rendering them to 29x29 bitmaps using a rectangular guide box that was used during collection. The size choice of 29x29 (same as in [6]) is somewhat arbitrary. It is conceivable that CJK characters with several strokes might benefit from larger bitmap sizes. Figures 2 and 3 present Chinese character ink samples rendered as 29x29 bitmaps. Print characters are easy to recognize, the cursive forms are

much more difficult.

Figure 2

. Print character samples (neatly written).

Figure 3. Cursive character samples.

The dataset was split into three sets: train (80%), validation (10%), and test (10%). The train dataset was used for training recognizers, while the validation set was used to monitor progress and determine when to stop training. All accuracies reported are on the test dataset. https://www.doczj.com/doc/3b6722598.html,beling Radical-Paths

Each of the 3665 characters was manually labeled with its unique tree decomposition. Printed forms of the characters as present in popular UNICODE fonts were used. First, all radicals were identified along with their radical paths. These radical paths were then collectively represented through a depth-first-search traversal of the character layout (see Figure 1).

4.3.Distribution

The labeled radical-path data was used to compute the distribution of characters over possible template-paths. Tables 1 and 2 present their distribution over common template paths of length 1. Note that these observations don’t take into consideration characters that comprise only a single radical (template-paths of length 0). While Table 1 presents frequencies, Table 2 presents percentages. Each entry in the table has the form

{path}: (A,B) + ext(a,b)

where

{path} is a template-path of length one (root node),

A and a = number of radical paths,

B and b = number of characters containing radical

paths with a given root node

Each template path begins with a template node which can have 2 or 3 children. Each child node can be a terminal node or a sub-tree. The (A,B) numbers indicate the number of children that were terminal nodes, while the extension numbers ext(a,b) indicate the number of characters that had sub-trees. It is more common to have

terminal nodes (radicals) as children rather than sub-trees.

Table 1. Distribution of characters and radical over template paths of length 1 (root node of the tree).

0? (267,833)+ext(49,62) 1? (359,814)+ext(87,102) 0? (255,2195) + ext(33,36)

1? (776,2014) + ext(156,221)

0? (3,16) 1? (16,16)

0? (6,80)

1? (76,77)+ext(6,6) 0? (23,12) + ext(2,2) 1? (34,61) + ext(4,4) 2? (47,63) 0? (16,35) 1? (29,35)

2? (20,33)+ext(4,4) 0? (5,31)

1?(26,29) + ext(4,4) 0? (3,4) 1? (4,4) 0? (14,133)

1? (100,113) + ext(38,42) 0? (11,31)

1? (26,29) + ext(4,4) 0? (2,10) 1? (10,10)

0? (2,2)+ext(2,2) 1? (2,2)+ext(2,2)

Table 2. Normalized distribution of characters and radical over template paths with percentages rounded to the nearest integer

0? (11%,12%)+ext(2%,1%) 1? (14%,11%)+ext(3%,1%) 0? (10%,31%) + ext(1%,1%) 1? (31%,28%) + ext(6%,3%) 0? (0%,0%) 1? (0%,0%)

0? (0%,1%)

1? (0%,1%)+ext(0%,0%) 0? (1%,0%) + ext(0%,0%) 1? (1%,1%) + ext(0%,0%) 2? (2%,1%) 0? (1%,0%) 1? (1%,0%)

2? (1%,0%)+ext(4%,0%) 0

? (0%,0%)

1?(0%,0%) + ext(0%,0%) 0? (0%,0%) 1?

(0%,0%) 0

?

(1%,2%)

1?

(4%,2%) + ext(2%,0%)

0? (0%,0%)

1? (1%,0%) + ext(4%,0%) 0? (0%,0%) 1? (0%,0%)

0? (0%,0%)+ext(0%,0%) 1? (0%,0%)+ext(0%,0%)

Some interesting observations can be made from the percentages in Table 2:

1. The characters composed of left-right and top-bottom parts (marked in red) make up over 82% of the 3665 characters. Further, these parts contain mostly two radicals (one for each child node) as indicated by low ext(a,b) percentages.

2. The ratio (B/A) gives the average number of characters corresponding to each unique radical path.

B (#chars containing root template-node)A (#radical-paths)= Examining (B/A) we note that

a) For

?, if we know the right radical (and that the character is of ? type), then on average we reduce the search space from 3665 characters down to nearly 1.5 characters i.e., recognition is almost done!

b) For ?, if we know the left radical (and that the character is of ? type), then we reduce the search space from 3665 characters down to nearly 6 characters.

c) For

?, if we know the top or bottom radical (and that the character is of ? type), then we reduce the search space from 3665 characters to nearly 2 characters.

These observations strongly indicate that character-parts (radicals or strokes) that are important for recognition are not uniformly distributed over all regions of the character. Further, this distribution varies over which radicals are present in which locations. For example, knowing the right radical of a ? type of character is more useful than knowing the left radical. The above distribution was also computed using only 1000 of the 3665 characters. Very little change in the distribution of the template paths was found between the 1000 and 3665 characters. We conjecture that this distribution is likely to be similar, if recomputed over a much larger set of Chinese characters.

5. Radical based recognizer design

The end-to-end recognizer comprises neural networks arranged in a graph. Each NN is one of three types: a radical-at-location (RAL) recognizer, a gater, or a combiner.

5.1. Radical-At-Location (RAL) recognizer

Figure 4. An example template-path classifier.

Each Radical-At-Location classifier, such as the one shown in Figure 4, has a unique template-path associated with it. It examines the whole character, and recognizes the radical present at the location specified by its template-path. Its input is a character (image/ink) and its output is a list of radical scores. Passing the whole character image as input avoids the need for explicitly segmenting the character to find the radical. Note that since not all (radical, template-path) pairs are valid, only those radicals that can actually appear at the location specified by the template-path are present as output classes.

5.2. Template-node gater

The template-node gater is a classifier that determines which template node is present at a specific depth and along a specific path from the root node. In the simplest case, this could be the root template node itself (one of 12 or so possible template nodes or a subset thereof). For example, a root template node of ? would imply that the character has a left- and a right-part. It is conceivable that for characters with very deep hierarchies a template-path gater can be designed for internal nodes at higher depths. Figure 5 presents an example template-path gater for the root node. The root template-node gater examines the whole character to determine the root node template

for the character tree. Not all possible output templates (about 12 of them) need to be present – any subset of the

templates would also suffice.

Figure 5. An example template-node gater.

5.3. Combiners and End-to-End recognizer

The template-path combiner combines the outputs of two or more RAL recognizers. Figure 6 presents an example gater—RAL—combiner architecture for recognizing characters made up two halves (left-right or top-bottom). Each individual part (left, right, top, or bottom) is recognized by a different RAL. All four RAL outputs are combined to produce a final distribution of probabilities over the supported characters. Note that only those characters that contain a left- and a right-part show up at the output of the combiner. Depending on the settings the individual RAL classifiers can be further decomposed. For example, the right-part of the fourth output character ? can be further decomposed to two more levels.

6. Experiments

Two recognizers with slightly different architectures were trained to recognize 3,665 Chinese (CHS) characters. The first one, shown in Figure 7, does not use any gaters, while the second one, shown in Figure 8, uses an LR vs TB gater. 3,076 (84%) of the 3,665 characters have an LR (?, left-radical, right-radical) or a TB (?, top-radical, bottom-radical) form. The rest of the 589 characters are either base radicals or have one of the other 10 root template nodes. In both recognizers, a single CNN was built to recognize these 589 characters. In the figures, this classifier is labeled as the “Radical+Misc” Classifier. The four RAL classifiers have the following template-paths: a) L-Classifier = 0?\ b) R-Classifier = 1?\ c) T-Classifier = 0?\ d) B-Classifier = 1?\

Each RAL-NN was a convolutional neural network with two convolutional layers, one hidden layer and one output layer [6]. The first and second convolutional layers contained 5 and 50 nodes, respectively. The convolutional kernels was 5x5 in size with each layer down-sampling the convolved output by a factor of 2. The number of hidden nodes was 200. The number of output nodes varied with the RAL (Table 4).

Bottom template-path Classifier

Identifies bottom-radical

Top template-path

Classifier

Identifies top-radical

Right template-path Classifier

Identifies right-radical

Root template-

template node

Classifier

Identifies left-radical

Figure 6. An example radical based classifier for

characters made up of two radicals.

Figure 7. An example recognizer without gaters.

Figure 8. The example recognizer with a gater.

Simple linear combiners with sigmoid nonlinearity (one layered neural networks) were used. Each combiner computed a weighted sum of its inputs and applied the nonlinearity to obtain each output. The large number of inputs (2,570) and outputs (3,665) result in a very large combiner weight matrix. However, since not all RAL output combinations are feasible, the weight matrix is very sparse and contains very few non-zero weights. For example, a dense weight matrix would have 2,570 x 3,665 = 9,419,050 weights. However, given that only 3,076 of these combinations result in actual characters, we expect only about 2 to 10 non-zero weights per output character, producing (2 to 10) x 3076 = 6,152 to 30,760 non-zero weights. Several combiners were built for combining the left and right RALs, the top and bottom RALs, all four RALs, and also the combiners shown in Figures 7 and 8.

The gater in Figure 8 has the same architecture as the RALs, but is much smaller with only 50 hidden and 2 output nodes. The gater is 4-5 times faster and can improve throughput by up to 40%, at a small decrease in accuracy (see Table 3). A gater with 3 outputs was also trained to tell the LR characters apart from the TB characters and also the rest of the 589 “Radical+Misc” characters. All networks were trained using stochastic gradient descent (backpropagation). For the combiners, weight decay was used as a simple regularization approach to ensure that most of these weights were zero or had very small values.

7.Results

Tables 3, 4, and 5 present the accuracies of the Gater, RAL, and Combiner classifiers, respectively. The two output gater achieves a high accuracy of 98.72% and is very successful at telling whether the character has a left-right or a top-bottom form. However, the three output gater has a much lower accuracy due to confusions between the (LR+TB) character set and the rest of the 589 (Radical+Misc) character set. Thus the LR vs TB gater was used in the final recognizers (Fig. 8).

Each of the RAL classifiers has less than 1000 outputs and thus is able to learn well. The (Radical+Misc.) classifier has the best accuracy of 94.48%, while the R-RAL classifier has an accuracy of only 85.80%. We conjecture that the simplicity of the characters produces such good results for the (Radical+Misc.) classifier while the large number of output classes and right radical variability makes recognition difficult for R-RAL. Overall, it is easier to recognize the left-, top-, and bottom-radicals in Chinese characters than the right-radical.

If the recognition errors of RALs were independent, then during combination we would expect the error rates to add together. For example, with an L-RAL error rate of 7.2% and an R-RAL error rate of 14.2%, a na?ve combiner would generate an error rate of over 21.4%. However, the LR-combiner produces an error rate of only 9.5%. Similarly, the TB-combiner produces a low error rate of 5.47%. Both of these results indicate that the radical decomposition approach has potential for scaling well by exploiting radical distribution patterns, like the ones shown in Tables 1 and 2. Lower error rates may be achieved if more powerful combiners are used.

The end-to-end classifier accuracies are presented in Table 6. The accuracy on print characters is over 95.5%. However, the cursive accuracy is only around 70%. Recognizing cursive characters appears to be a much more difficult as identified in several related research efforts [1-3]. From Table 3 we note that almost 82% of the characters are expected to contain LR or TB radical parts, while the remaining 18% of the characters make up the radicals and characters with other forms. Comparing the two end-to-end recognizers with and without the LR vs TB gater, we note that the gater produces an almost 40% increase in throughput with only a marginal reduction in accuracy.

Table 3. Gater accuracies.

Classifier #Chars

#Outputs

Accuracy

(%) LR vs TB Gater 3076 2 98.72

LR vs TB vs Rest

Gater

3665 3 74.95 Table 4. Radical-At-Location (RAL) accuracies.

Classifier #Chars

#Outputs

Accuracy

(%) L-RAL 2213 288 92.80

R-RAL 2213 932 85.80

T-RAL 863 316 90.68

B-RAL 863 446 90.47 Radical+Rest 589 589 94.48 Table 5. Classifier Combiner accuracies

Classifier #Chars

#Outputs

Accuracy

(%) LR-Combiner 2213 2213 90.50

TB-Combiner 863 863 94.53

LRTB-Combiner 3076 3076 89.50

Table 6. End-to-end Recognizer accuracies

Accuracy (%)

Classifier #Chars

Print Cursive With Gater (Fig. 8) 3665 95.56 69.13

No Gater (Fig. 7) 3665 95.81 70.64

8.Conclusion

We presented a new radical based approach for scaling neural network (NN) recognizers to thousands of East-Asian characters. Experimental results on a database of 13.4 million Chinese character samples from 3665 classes show that this method achieves good scalability by exploiting correlations between spatial components (radicals) in the characters.

References

[1]C-L Liu, S Jaeger, and M Nakagawa (2004), “Online

Recognition of Chinese Characters: The State-of-the-

Art,” IEEE PAMI, v. 26, no 2.

[2] D Shi, RI Damper, and SR Gunn (2003), “Offline

Handwritten Chinese Character Recognition by Radical

Decomposition,” ACM Trans. Asian Lang. Inf. Process.

2(1): 27-48.

[3]A-B Wang, K-C Fan (2001), “Optical recognition of

handwritten Chinese characters by hierarchical radical

matching method,” Pattern Recog. v. 34, no 1, pp. 15-35.

[4]WWS Ip, KFL Chung and DS Yeung (1995), “Offline

handwritten Chinese character recognition via radical

extraction and recognition”, ICDAR’95, pp. 185-189. [5]CW Liao and JS Huang (1990), “A Transformation

invariant matching algorithm for handwritten Chinese

character recognition”, Pattern Recognition, v. 23, no.

11, pp. 1167-1188.

[6]P. Y. Simard, D. Steinkraus, and J Platt, (2003), “Best

Practice for Convolutional Neural Networks Applied to

Visual Document Analysis,” in ICDAR’03, pp. 958-962.

数据结构第三章栈和队列3习题

第三章栈和队列试题 一、单项选择题 1.栈的插入和删除操作在()进行。 A. 栈顶 B. 栈底 C. 任意位置 D. 指定位置 2.当利用大小为n的数组顺序存储一个栈时,假定用top==n表示栈空,则向这个栈插入一个元素时, 首先应执行()语句修改top指针。 A. top++; B. top--; C. top = 0; D. top; 3.若让元素1,2,3依次进栈,则出栈次序不可能出现()种情况。 A. 3, 2, 1 B. 2, 1, 3 C. 3, 1, 2 D. 1, 3, 2 4.在一个顺序存储的循环队列中,队头指针指向队头元素的()位置。 A. 前一个 B. 后一个 C. 当前 D. 后面 5.当利用大小为n的数组顺序存储一个队列时,该队列的最大长度为()。 A. n-2 B. n-1 C. n D. n+1 6.从一个顺序存储的循环队列中删除一个元素时,需要()。 A. 队头指针加一 B. 队头指针减一 C. 取出队头指针所指的元素 D. 取出队尾指针所指的元素 7.假定一个顺序存储的循环队列的队头和队尾指针分别为front和rear,则判断队空的条件为()。 A. front+1 == rear B. rear+1 == front C. front == 0 D. front == rear 8.假定一个链式队列的队头和队尾指针分别为front和rear,则判断队空的条件为()。 A. front == rear B. front != NULL C. rear != NULL D. front == NULL 9.设链式栈中结点的结构为(data, link),且top是指向栈顶的指针。若想在链式栈的栈顶插入一 个由指针s所指的结点,则应执行操作()。 A. top->link = s; B.s->link = top->link; top->link = s; C. s->link = top; top = s; D. s->link = top; top = top->link; 10.设链式栈中结点的结构为(data, link),且top是指向栈顶的指针。若想摘除链式栈的栈顶结点, 并将被摘除结点的值保存到x中,则应执行操作()。 A. x = top->data; top = top->link; B. top = top->link; x = top->data; C. x = top; top = top->link; D. x = top->data; 11.设循环队列的结构是 #define MaxSize 100 typedef int ElemType;

高中语文说课稿子实用模板(共5篇)

篇一:高中语文说课模板通稿 各位评委老师:大家好,我是几号考生,我今天说课的题目是《》。下面,我将会从说教材、说教法、说学法、说教学过程等部分来进行我的说课。 一、说教材 教材简析:本篇课文是人教版高中语文课本必修几第几册第几单元第几课的课文,是一篇什么体裁的文章。文章主要叙述的容是什么,表现了作者什么样的思想感情。本篇课文具有什么样的地位,和什么特色。学习教材的目的是为了让学生能够触类旁通,举一反三,能将知识由课向课外延伸。通过对本篇课文学习,能够更好的培养学生的鉴赏能力和归纳总结能力。 教学目标: (一)诗歌:根据教学大纲的要求,诗歌教学要以诵读为主,在诵读中使学生领悟诗的思想感情,鉴赏文学作品能够感受形象,品味语言。同时,本篇诗歌通过典型意象创造出优美的意境,巧妙的将诗人的情感融入其中。确立教学目标如下: 我的知识目标是:品味诗的意象美,领悟诗的意境,体会诗人的真情。 能力目标是:培养学生鉴赏诗歌的能力。 德育目标是:体悟作者情怀,引发共鸣,逐渐养成良好的审美情趣,培养审美能力。 本节课的教学重点是:领反复诵读,鉴赏诗歌意象,体会诗歌的意境美,体悟诗情,获得情感体验。教学难点是:把握诗歌的思想感情,以及意象的体味与把握。 (二)文言文:根据教学大纲的要求,文言文教学要求培养学生阅读浅显文言文的能力,把握文章的思想容,积累语言知识。同时本课情节、结构、语言等的特点,确立教学目标如下: 知识与能力目标:了解作者的基本常识;理解文中所出现的生僻词以及意义特殊的词语;把握文中的主要思想感情,初步鉴赏文中的形象与个性。 过程与方法:通过诵读质疑、自主探究、合作解决,培养学生借助注释和工具书阅读浅显文言文,理解重点词句的含义以及初步的研究分析能力并积累文言文实词、虚词和句式,培养学生的自学能力和翻译能力。情感态度与价值观目标:理解作者在作品中所流露出的情感,培养正确的审美观/价值观/人生观。 我确定以上的三个目标是体现新课标所提倡的“知识与能力、过程与方法、情感与态度”并重的教学理念。教学重点:诵读课文,在把握文意的基础上理解词句,积累一些文言实词、虚词和句式。 教学难点:掌握文章的写作风格和特点;深入分析文章涵,并作出评价,激发学生的学习兴趣。 (三)现代文:新课标强调了要全面提升高中学生的语文素养,初步形成正确的人生观、世界观、价值观,并学会收集判断、处理信息,具有人文素养、创新精神与实践能力。同时本篇课文,感情真挚,感悟深刻,具有典型的人文性,结合本单元的教学目标以及本篇课文在语言,结构设置等方面的特点,确立本课的教学目标如下: 知识能力目标:领悟本文的写作手法,揣摩语言,体会语言的感情和特色,学习作者运用语言的技巧。同时学习本文对比喻、拟人等修辞手法的运用。 过程与方法目标:在教师恰当的提升引导下,培养学生“自主、合作、探究”的学习方式,并掌握一定的赏析语言的能力与方法。 情感与态度目标:理解作者在作品中所流露出的情感,培养正确的审美观/价值观/人生观。 我确定以上的三个目标是体现新课标所提倡的“知识与能力、过程与方法、情感与态度”并重的教学理念。我的教学重点是:理清本篇文章的思路,领会课文的思想感情。教学难点

第三章栈和队列习题_数据结构电子教案

习题三栈和队列 一单项选择题 1. 在作进栈运算时,应先判别栈是否(① ),在作退栈运算时应先判别栈是否(② )。当栈中元素为n个,作进栈运算时发生上溢,则说明该栈的最大容量为(③ )。 ①, ②: A. 空 B. 满 C. 上溢 D. 下溢 ③: A. n-1 B. n C. n+1 D. n/2 2.若已知一个栈的进栈序列是1,2,3,…,n,其输出序列为p1,p2,p3,...,pn,若p1=3,则p2为( )。 A 可能是2 B 一定是2 C 可能是1 D 一定是1 3. 有六个元素6,5,4,3,2,1 的顺序进栈,问下列哪一个不是合法的出栈序列?() A. 5 4 3 6 1 2 B. 4 5 3 1 2 6 C. 3 4 6 5 2 1 D. 2 3 4 1 5 6 4.设有一顺序栈S,元素s1,s2,s3,s4,s5,s6依次进栈,如果6个元素出栈的顺序是s2,s3,s4, s6, s5,s1,则栈的容量至少应该是() A.2 B. 3 C. 5 D.6 5. 若栈采用顺序存储方式存储,现两栈共享空间V[1..m],top[i]代表第i个栈( i =1,2)栈顶,栈1的底在v[1],栈2的底在V[m],则栈满的条件是()。 A. |top[2]-top[1]|=0 B. top[1]+1=top[2] C. top[1]+top[2]=m D. top[1]=top[2] 6. 执行完下列语句段后,i值为:() int f(int x) { return ((x>0) ? x* f(x-1):2);} int i ; i =f(f(1)); A.2 B. 4 C. 8 D. 无限递归 7. 表达式3* 2^(4+2*2-6*3)-5求值过程中当扫描到6时,对象栈和算符栈为(),其中^为乘幂。 A. 3,2,4,1,1;(*^(+*- B. 3,2,8;(*^- C. 3,2,4,2,2;(*^(- D. 3,2,8;(*^(- 8. 用链接方式存储的队列,在进行删除运算时()。 A. 仅修改头指针 B. 仅修改尾指针 C. 头、尾指针都要修改 D. 头、尾指针可能都要修改 9. 递归过程或函数调用时,处理参数及返回地址,要用一种称为()的数据结构。 A.队列 B.多维数组 C.栈 D. 线性表 10.设C语言数组Data[m+1]作为循环队列SQ的存储空间, front为队头指针,rear为队尾指针,则执行出队操作的语句为() A.front=front+1 B. front=(front+1)% m C.rear=(rear+1)%(m+1) D. front=(front+1)%(m+1) 11.循环队列的队满条件为 ( ) A. (sq.rear+1) % maxsize ==(sq.front+1) % maxsize; B. (sq.front+1) % maxsize ==sq.rear C. (sq.rear+1) % maxsize ==sq.front D.sq.rear ==sq.front

第三章栈和队列练习题

第三章栈和队列练习题 一、单项选择题 1.一个顺序栈一旦被声明,其占用空间的大小()。 A.已固定B.可以改变C.不能固定D.动态变化 2.链栈和顺序栈相比,有一个比较明显的缺点,即()。 A.插入操作更加方便B.通常不会出现栈满的情况 C.不会出现栈空的情况D.删除操作更加方便 3.用单链表表示的链式队列的队头在链表的()位置。 A.链头B.链尾C.链中D.任意位置 4.在解决计算机主机与打印机之间速度不匹配问题时通常设置一个打印数据缓冲区,主机将要输出的数据依次写入缓冲区中,而打印机则从缓冲区中取出数据打印,该缓冲区应该是一个()结构。 A.堆栈B.队列C.数组D.先性表 5.若已知一个栈的入栈序列是1,2,3,…,30,其输出序列是p1,p2,p3,…p n,若p1=30,则p10为()。 A.11 B.20 C.19 D.21 6.循环队列A[m] 存放其元素,用front和rear分别表示队头及队尾,则循环队列满的条件是()。 A.(rear+1)%m=front B.rear =front+1 C.rear=front D.(rear+1)%m-1=front 7.在一个栈顶指针为top的链栈中,将一个p指针所指的结点入栈,应执行()。 A.top->next=p; B.p->next=top->next; top->next=p; C.p->next=top; top=p; D.p->next=top->next; top=top->next; 8.在一个栈顶指针为top的链栈中删除一个结点时,用x保存被删结点的值,则执行()。 A.x=top;top=top->next; B.x=top->data;

人教版高中语文说课稿合集(必修一至必修五)139页)

目录 1、《荷塘月色》说课稿 (3) 2、《故都的秋》说课稿1 (6) 2、《故都的秋》说课稿2 (8) 3、《囚绿记》说课稿 (11) 4、诗经两首《氓》说课稿 (12) 5、《离骚》说课稿 (15) 6、孔雀东南飞说课稿 (17) 7《涉江采芙蓉》说课稿一. (18) 7、短歌行说课稿 (19) 8、归园田居其一说课稿 (20) 9、兰亭集序》说课 (23) 10、赤壁赋说课稿 (24) 11、《游褒禅山记》说课稿(2则) (26) 12、就任北京大学校长之演说 (30) 13、我有一个梦想说课稿 (35) 14、《在马克思墓前的讲话》说课稿 (38) 15、林黛玉进贾府说课稿 (39) 16、《祝福》说课稿 (43) 17、《老人与海》说课稿 (45) 18蜀道难说课稿 (46) 19、《秋兴八首》(其一)教案 (48) 20、杜甫诗三首》说课稿 (49) 21、琵琶行说课稿 (51) 22、李商隐诗2首 (53) 23、《寡人之于国也》说课稿 (55) 24、《劝学》说课稿 (58) 24《劝学》创新课说课稿 (59) 25、过秦论说课稿 (61) 26、师说》说课稿 (63) 27、动物游戏之谜 (65) 27-1动物游戏之谜 (67) 28、《宇宙的边疆》 (69) 13-1、宇宙的边疆 (70) 29、一名物理学家的教育历程》说课稿 (71) 30、《窦娥冤》说课稿 (75) 30-1、《窦娥冤》说课稿 (76) 30-2《窦娥冤》说课稿 (77) 31、《雷雨》说课稿 (79) 32、《哈姆雷特》说课稿 (81) 33、《望海潮》说课稿 (82) 34、《雨霖铃》说课稿 (83) 35、《苏轼词两首》说课稿 (84) 36、《念奴娇赤壁怀古》说课稿 (86) 37、《水龙吟》教案 (89)

人教版职高语文基础模块上册教案

《永远的校园》 教学目的: 1.了解北大,走近北大,感受北大,思考北大 2. 借助演讲、辩论等形式培养学生思辨说理的能力 3.对比谢冕心中的北大与现实中北大的差异,并揣摩这种差异的根源,从而树立正确的大学观、人生观 教学难点: 如何看待理想中北大与现实中北大的差异 教法与学法: 1.课型:讲读课、公开课、研讨课、活动课 2.教法:质疑、讨论、辩论 3.教具及能力培养方向:多媒体课件(包括图片、音频、视频及文字材料);培养学生思辨能力 课时安排:2课时(第一课时主要诵读文章,感知大意;第二课时主要展开拓展、延伸的活动) 教学过程: 一、新课导入 2003年2月25日,在首都北京发生了二起震惊中外的爆炸案,先后发生在清华大学的荷园餐厅和北京大学的农园餐厅,虽然比不上恐怖份子驾机撞毁世贸大楼那样引人注目,但在很少发生恐怖流血冲突的中国大陆来说,这的确也算得上是惊天动地的事情了,最引人深

思的是:恐怖份子为什么要选择清华和北大进行作案呢?案情侦破之后,答案浮出水面,而这个答案竟让人哭笑不得,因为犯罪谷子声称:“制造爆炸案的目的是为了出名。” 对于这起事件,我不想多作分析了,我只是想说:为什么炸清华、北大就能出名呢?那是因为这两所大学太有名了,本身就是名望的象征。我们这就容易理解了:为什么从古至今那些杀人的罪犯很少留下姓名,而刺客荆轲的故事却被广为传诵,不仅仅是因为荆轲的大义凛然、高风亮节,还因为他刺杀的对象是权倾天下、赫赫有名的秦王。 的确,清华、北大太有名气了,不仅是每一位莘莘学子梦寐以求的地方,就连没有丁点儿文化常识的人可能也会对他们肃然起敬。江泽民同志作为一国主席,日理万机,我不知道他曾去过哪些大学,但我知道北大他是去过的,就在北大建校一百周年的校庆典礼上,还作了热情扬溢的讲话,因此,才有了我们课本上这篇文章。 大凡有名望的人、事、物都难得一见:比如卢浮宫珍藏的那些传世珍品;比如我们的局长、县长、市长或者更高的官员。于是我们容易对那些有名望的东西产生神秘感,认为他们与众不同,认为他们超凡脱俗,认为他们高不可攀,或许真的如此,或许未必竟然。那么,今天,我们就一起走进北大著名学者、中国当代著名诗歌理论家、朦胧诗派的倡导者谢冕先生用他的眼、他的手、他的心为我们搭建的北大校园,去感受北大的那一份脱俗和平凡吧! 二、感受北大

第3章-栈与队列习题参考答案

习题三参考答案 备注: 红色字体标明的是与书本内容有改动的内容。 一、选择题 1.在栈中存取数据的原则是( B )。 A.先进先出 B. 先进后出 C. 后进后出 D. 没有限制 2.若将整数1、2、3、4依次进栈,则不可能得到的出栈序列是( D )。 A.1234 B. 1324 C. 4321 D. 1423 3.在链栈中,进行出栈操作时(B )。 A.需要判断栈是否满 B. 需要判断栈是否为空 C. 需要判断栈元素的类型 D. 无需对栈作任何差别 4.在顺序栈中,若栈顶指针top指向栈顶元素的下一个存储单元,且顺序栈的最大容量是maxSize,则顺序栈的判空条件是( A )。 A.top==0 B.top==-1 C. top==maxSize D.top==maxSize-1 5.在顺序栈中,若栈顶指针top指向栈顶元素的下一个存储单元,且顺序栈的最大容量是maxSize。则顺序栈的判满的条件是( C )。 A.top==0 B.top==-1 C. top==maxSize D.top==maxSize-1 6.在队列中存取数据元素的原则是( A )。 A.先进先出 B. 先进后出 C. 后进后出 D. 没有限制 7.在循环顺序队列中,假设以少用一个存储单元的方法来区分队列判满和判空的条件,front和rear分别为队首和队尾指针,它们分别指向队首元素和队尾元素的下一个存储单元,队列的最大存储容量为maxSize,则队列的判空条件是(A )。 A.front==rear B. front!=rear C. front==rear+1 D. front==(rear+1)% maxSize 8.在循环顺序队列中,假设以少用一个存储单元的方法来区分队列判满和判空的条件,front和rear分别为队首和队尾指针,它们分别指向队首元素和队尾元素的下一个存储单元,队列的最大存储容量为maxSize,则队列的判满条件是(D )。 A.front==rear B. front!=rear C. front==rear+1 D. front==(rear+1)% maxSize 9.在循环顺序队列中,假设以少用一个存储单元的方法来区分队列判满和判空的条件,front和rear分别为队首 和队尾指针,它们分别指向队首元素和队尾元素的下一个存储单元,队列的最大存储容量为maxSize,则队列的长度是(C )。 A.rear-front B. rear-front+1 C. (rear-front+maxSize)%maxSize D. (rear-front+1)%maxSize 10.设长度为n的链队列采用单循环链表加以表示,若只设一个头指针指向队首元素,则入队操作的时间复杂度 为( B )。 A.O(1) B.O(n) C.O(log2n) D.O(n2) 二、填空题 1.栈是一种操作受限的特殊线性表,其特殊性体现在其插入和删除操作都限制在表尾进行。允许插入和删除 操作的一端称为栈顶,而另一端称为栈底。栈具有后进先出的特点。 2.栈也有两种存储结构,一种是顺序存储,另一种是链式存储;以这两种存储结构存储的栈分别称为顺序 栈和链栈。 3.在顺序栈中,假设栈顶指针top是指向栈顶元素的下一个存储单元,则顺序栈判空的条件是 top==0 ; 栈顶

第三章+栈和队列(参考答案)

第三章栈和队列 一、判断题 1、链栈的初始化是指开辟足够多的结点,然后置栈顶指针为 NULL。(×) 2、递归定义的数据结构通常不需要用递归的算法来实现对它的操作。(×) 二、填空题 1、向一个链式栈插入一个新结点时,首先把栈顶指针的值赋给新结点的指针域,然后把新结点的存储位置赋给___栈顶指针_____。 2、迷宫问题是一个回溯控制的问题,最好使用____栈______的方法来解决。 3、有如下递归过程: Void Print(int w) { int i; if (w!=0) { Print(w?1); for (i=1;i<=w;i++) printf(“%3d”,w); printf(“\n”); } } 调用语句print(4)的结果是__________。 1 2 2 3 3 3 4 4 4 4 4、假设用循环单链表实现队列,若队列非空,且队尾指针为R, 则将新结点S加入队列时,需执行下面语句:_ S->next=R->next _________;___ R->next=S _______;R=S; 三、选择题 1、设有4个数据元素a1、a 2、a3和a4,对他们分别进行栈操作或队操作。在进栈或进队操作时,按a1、a2、a 3、a4次序每次进入一个元素。假设栈或队的初始状态都是空。 现要进行的栈操作是进栈两次,出栈一次,再进栈两次,出栈一次;这时,第一次出栈得到的元素是 A 2,第二次出栈得到的元素是 B 4;类似地,考虑对这四个数据元素进行的队操作是进队两次,出队一次,再进队两次,出队一次;这时,第一次出队得到的元素是 C 1,第二次出队得到的元素是 D 2。经操作后,最后在栈中或队中的元素还有 E 2个。 供选择的答案: A~D:①a1 ②a2 ③ a3 ④a4 E:①1 ②2 ③ 3 ④ 0 2、栈是一种线性表,它的特点是 A 2。设用一维数组A[1,…,n]来表示一个栈,A[n]为栈底,用整型变量T指示当前栈顶位置,A[T]为栈顶元素。往栈中推入(PUSH)一个新元素时,变量T的值 B 2;从栈中弹出(POP)一个元素时,变量T的值 C 1。设栈空时,有输入序列a,b,c,经过PUSH,POP,PUSH,PUSH,POP操作后,从栈中弹出的元素的序列是 D 6,变量T的值是 E 4。 供选择的答案: A:①先进先出②后进先出③进优于出④出优于进⑤随机进出 B,C:①加1 ②减1 ③不变④清⑤加2 ⑥减2 D:① a,b ②b,c ③c,a ④b,a ⑤ c,b ⑥a,c E:① n+1 ②n+2 ③ n ④ n-1 ⑤ n-2 3、在做进栈运算时,应先判别栈是否 A 2;在做退栈运算时,应先判别栈是否 B 1。当栈中元素为n个,做进栈运算时发生上溢,则说明该栈的最大容量为 C 2。

高中语文《鸿门宴》说课稿模板

高中语文新课程标准教材 语文教案( 2019 — 2020学年度第二学期 ) 学校: 年级: 任课教师: 说课稿 / 高中说课稿 / 高中说课稿范文 编订:XX文讯教育机构

高中语文《鸿门宴》说课稿模板 教材简介:本教材主要用途为通过学习语文的内容,培养学生的阅读能力、交流能力,学习语文是为了更好的学习和工作,为了满足人类发展和实现自我价值的需要,本说课稿资料适用于高中语文科目, 学习后学生能得到全面的发展和提高。本内容是按照教材的内容进行的编写,可以放心修改调整或直接进行教学使用。 各位评委,各位同仁: 大家好!今天我说课的题目是《鸿门宴》。下面我将从:教材、教法、学法、课时安排、教学过程五个方面来进行我的说课。 一、说教材 1、教材简析及该课所处的地位作用: 《鸿门宴》是人教版新教材高一年级《语文》必修1第二单元 第6课,节选自《史记》。所选情节描写的是刘邦、项羽在推翻秦王朝后,为争夺天下而进行的一场惊心动魄的政治斗争。鸿门宴是项羽在新丰鸿门设下的一个富有杀机的宴会,是刘、项之间政治矛盾由潜滋暗长到公开化的表现,是长达五年的楚汉相争的序幕,名为宴会,实则是一场激烈的政治斗争。作者以娴熟的艺术技巧,生动地刻画了不同性格的人物形象,描绘了众多富有戏剧性的令人难忘的场面,情节波澜起伏,扣人心弦。鸿门宴一文充分展示了刘、项之间不可调和的矛盾,并预示了胜负的必然结局。因此具有很高的史学价值和

文学价值。如何通过本课激起学生学习的兴趣,树立学生学好文言文的信心和领悟我国传统文化的精髓,这是一个关键所在。 2、教学目标: 高中语文教学大纲对文言文学习的要求是:诵读古典诗词和浅易文言文,背诵一定数量的名篇。掌握文中常见的文言实词、虚词和句式,能理解词句含义,读懂课文,学习用现代观念审视作品的内容和思想倾向。 根据课文特点,结合大纲要求,我确立了如下教学目标:(课件展示) 板书一二三 (1)知识和能力目标:一个巨匠,一部巨著:了解作家作品及相关史实,积累文言词句知识,学习创造性阅读。 (2)过程和方法目标:两个英雄,两种性格:利用人物个性化语言,分析人物性格,学习作者运用对照的方法和在矛盾斗争中描写人物的方法,学习一分为二地评论历史人物,树立正确的读史观。 (3)情感和价值目标:三种认知态度: ①学以致用的态度,感受司马迁隐忍发愤的著书精神,从容应对学习生活中的各种困难和挫折 ②领悟态度:分析项羽悲剧的原因,到底“败”在哪里?领悟良好的性格决定一生的命运

高中语文说课稿范文

高中语文说课稿范文 篇一:高中语文说课稿范文(共7篇) 篇一:高中语文说课模板通稿 各位评委老师:大家好,我是几号考生,我今天说课的题目是。下面,我将会从说教材、说教法、说学法、说教学过程等部分来进行我的说课。 一、说教材 教材简析:本篇课文是人教版高中语文课本必修几第几册第几单元第几课的课文,是一篇什么体裁的文章。文章主要叙述的内容是什么,表现了作者什么样的思想感情。本篇课文具有什么样的地位,和什么特色。学习教材的目的是为了让学生能够触类旁通,举一反三,能将知识由课内向课外延伸。通过对本篇课文学习,能够更好的培养学生的鉴赏能力和归纳总结能力。 教学目标: (一)诗歌:根据教学大纲的要求,诗歌教学要以诵读为主,在诵读中使学生领悟诗的思想感情,鉴赏文学作品能够感受形象,品味语言。同时,本篇诗歌通过典型意象创造出优美的意境,巧妙的将诗人的情感融入其中。确立教学目标如下:我的知识目标是:品味诗的意象美,领悟诗的意境,体会诗人的真情。 能力目标是:培养学生鉴赏诗歌的能力。

德育目标是:体悟作者情怀,引发共鸣,逐渐养成良好的审美情趣,培养审美能力。本节课的教学重点是:领反复诵读,鉴赏诗歌意象,体会诗歌的意境美,体悟诗情,获得情感体验。教学难点是:把握诗歌的思想感情,以及意象的体味与把握。 (二)文言文:根据教学大纲的要求,文言文教学要求培养学生阅读浅显文言文的能力,把握文章的思想内容,积累语言知识。同时本课情节、结构、语言等的特点,确立教学目标如下:知识与能力目标:了解作者的基本常识;理解文中所出现的生僻词以及意义特殊的词语;把握文中的主要思想感情,初步鉴赏文中的形象与个性。 过程与方法:通过诵读质疑、自主探究、合作解决,培养学生借助注释和工具书阅读浅显文言文,理解重点词句的含义以及初步的研究分析能力并积累文言文实词、虚词和句式,培养学生的自学能力和翻译能力。情感态度与价值观目标:理解作者在作品中所流露出的情感,培养正确的审美观/价值观/人生观。 我确定以上的三个目标是体现新课标所提倡的“知识与能力、过程与方法、情感与态度”并重的教学理念。教学重点:诵读课文,在把握文意的基础上理解词句,积累一些文言实词、虚词和句式。 教学难点:掌握文章的写作风格和特点;深入分析文章内涵,并作出评价,激发学生的学习兴趣。

高中语文获奖说课稿(共5篇)

篇一:高中语文省级说课一等奖《李清照词两首》说课稿 《李清照词两首》说课稿 嘉峪关市一中张红丽 尊敬的各位评委、各位老师: 大家下午好! 我今天说课的题目是《李清照词两首》,这是高一第二学期后半学期的学习内容。下面我将从教材分析、学情介绍、教学目标、教学过程四个部分来进行说明。 第一部分——教材分析 《李清照词两首》是人教版必修四第二单元中的篇目,这一单元选编的是宋词,在新课标实验教材五个必修模块中,这是唯一的宋词单元。本单元选取的是柳永、苏轼、辛弃疾、李清照四位词人的作品,四者基本完整地体现了宋词的发展脉络和艺术价值。李清照的词《醉花阴》《声声慢》放在最后,属略读篇目,安排一课时完成教学。《醉花阴》是作者南渡前的作品,抒发“相思闲愁”,《声声慢》是作者南渡后的名篇之一,将重点赏析,写词人历遭国破家亡夫丧劫难后的愁苦悲戚,是词人晚年生活的真实写照,也是时代苦难的象征。本课的学习将引导学生深入熟悉诗词鉴赏的基本方法,为以后进行诗歌与散文系列的选修打下坚实的基础。 第二部分——学情介绍 通过本单元前面六首词的学习,学生对豪放派和婉约派的风格特点已有了大体认识,对词的文体常识和基本鉴赏方法已有所掌握,这一点对学生学习本课是很有利的,但李清照的这两首词如果不了解词人的人生遭遇就很难把握两首词中深层的感情内涵。 针对学生学习本课的不利因素,我安排学生在课前 1.回顾以前学过的李清照的作品 2.查找相关资料了解李清照生平 3.利用早读熟读成诵,初步感知两词。第三部分——教学目标 基于以上教材地位、学情特点以及新课标的要求,我确立了本课的三维教学目标: ①背诵这两首词,因为《语文课程标准》中明确要求学生要背诵一定数量的名篇。 ②通过反复诵读,对比探究,品味意象,体会“闲愁”与“悲愁”的区别,熟悉诗词鉴赏的基本方法,学会知人论世。因《语文课程标准》中要求学生“要加强诵读,在诵读中感受作品的意境和形象,领悟作品的丰富内涵和深层意义”,还提出要培养学生的探究意识,要学习鉴赏诗歌的基本方法,学会知人论世,了解与作品相关的作家经历、时代背景等,加深对作家作品的理解”,依据这两点所以我确立了教学目标二,同时这也是本课的教学重难点。 ③体会真情之美,培养学生对古诗词的热爱及审美情趣。《语文课程标准》中提出“高中语文课程应关注学生情感的发展,培养学生自觉的审美意识和高尚的审美情趣”,因此我确立了教学目标三。 教法与学法 依据教学目标,及《语文课程标准》中“语文教学要为学生创设良好的自主学习情境,倡导自主、合作、探究学习方式”的要求,本课将通过诵读法、情境 法、对比探究法来进行教学,学生通过诵读法、自主合作探究法来学习。 说完了教法与学法,紧接着我将要说的是“教学过程”,分五步走。 第四部分、教学过程 一、新课导入 课前播放《菊花台》歌曲,学生欣赏完我会问:为什么说“菊花残,满地伤”?菊花有什么象征意义?学生对《满城尽带黄金甲》这部电影比较熟悉,可能会想到铺满皇宫的菊花,宫廷军变的失败,生命的凋残消逝??我会顺势总结到:菊花在影片中象征①对已逝辉煌的伤感②对生命凋残的悲悼。“花”这个意象在中国古典诗词中有着深刻的文化内涵,古人有以花

第三章栈与队列 练习题

第三章栈与队列练习题 一、选择题 1、栈结构通常采用的两种存储结构是( A )。 A、顺序存储结构和链表存储结构 B、散列和索引 C、链表存储结构和数组 D、线性链表和非线性存储 2、设栈ST用顺序存储结构表示,则栈ST为空的条件是(B) A、ST.top-ST.base<>0 B、ST.top-ST.base==0 C、ST.top-ST.base<>n D、ST.top-ST.base==n 3、向一个栈顶指针为HS的链栈中插入一个s结点时,则执行() A、HS->next=s; B、s->next=HS->next;HS->next=s; C、s->next=HS;HS=s; D、s->next=HS;HS=HS->next; 4、从一个栈顶指针为HS的链栈中删除一个结点,用x保存被删除结点的值,则执行(C) A、x=HS;HS=HS->next; B、HS=HS->next;x=HS->data; C、 x=HS->data;HS=HS->next; D、s->next=Hs;Hs=HS->next; 7、一个队列的入列序列是1,2,3,4,则队列的输出序列是(B )//尾插入元素,头删除元素。 A、4,3,2,1 B、1,2,3,4 C、1,4,3,2 D、3,2,4,1 9、循环队列SQ采用数组空间SQ.base[0,n-1]存放其元素值,已知其头尾指针分别是front和rear,则判定此循环队列为满的条件是(C)//不懂啊!!! A、Q.front==Q.rear B、Q.front!=Q.rear C、Q.front==(Q.rear+1)%n D、Q.front!=(Q.rear+1)%n 11、用单链表表示的链式队列的队头在链表的(A)位置 A、链头 B、链尾 C、链中 12、判定一个链队列Q(最多元素为n个)为空的条件是( A) A、Q.front==Q.rear B、Q.front!=Q.rear C、Q.front==(Q.rear+1)%n D、Q.front!=(Q.rear+1)%n 14、在一个链队列Q中,删除一个结点需要执行的指令是(C) A、Q.rear=Q.front->next; B、Q.rear->next=Q.rear->next->next; C、 Q.front->next=Q.front->next->next; D、Q.front=Q.rear->next; 15、用不带头结点的单链表存储队列,其队头指针指向队头结点,队尾指针指向队尾结点,则在进行出队操作时(D) A、仅修改队头指针 B、仅修改队尾指针 C、队头尾指针都要修改 D、队头尾指针都可能要修改。 16、栈和队列的共同点是(C) A、都是先进后出 B、都是先进先出 C、只允许在端点处插入和删除元素 D、没有共同点 18、设有一顺序栈S,元素s1,s2,s3,s4,s5,s6依次进栈,如果6个元素出栈的顺序是s2,s3,s4,s6,s5,s1,则栈的容量至少应该是(B) A、2 B、 3 C、 5 D、 6 20、设有一顺序栈已经含有3个元素,如图3.1所示元素a4正等待进栈。下列不可能出现的出栈序列是(A) 0 maxsize-1

数据结构第3章栈与队列习题

第3章栈与队列 一、单项选择题 1.元素A、B、C、D依次进顺序栈后,栈顶元素是,栈底元素是。 A.A B.B C.C D.D 2.经过以下栈运算后,x的值是。 InitStack(s);Push(s,a);Push(s,b);Pop(s,x);GetTop(s,x); A.a B.b C.1 D.0 3.已知一个栈的进栈序列是ABC,出栈序列为CBA,经过的栈操作是。 A.push,pop,push,pop,push,pop B.push,push,push,pop,pop,pop C.push,push,pop,pop,push,pop D.push,pop,push,push,pop,pop 4.设一个栈的输入序列为A、B、C、D,则借助一个栈所得到的序列是。 A.A,B,C,D B.D,C,B,A C.A,C,D,B D.D,A,B,C 5.一个栈的进栈序列是a,b,c,d,e,则栈的不可能的输出序列是。 A.edcba B.decba C.dceab D.abcde 6.已知一个栈的进栈序列是1,2,3,……,n,其输出序列的第一个元素是i,则第j个出栈元素是。 A.i B.n-i C.j-i+1 D.不确定 7.已知一个栈的进栈序列是1,2,3,……,n,其输出序列是p1,p2,…,Pn,若p1=n,则pi的值。 A.i B.n-i C.n-i+1 D.不确定 8.设n个元素进栈序列是1,2,3,……,n,其输出序列是p1,p2,…,p n,若p1=3,则p2的值。 A.一定是2 B.一定是1

C.不可能是1 D.以上都不对 9.设n个元素进栈序列是p1,p2,…,p n,其输出序列是1,2,3,……,n,若p3=1,则p1的值。 A.可能是2 B.一定是1 C.不可能是2 D.不可能是3 10.设n个元素进栈序列是p1,p2,…,p n,其输出序列是1,2,3,……,n,若p3=3,则p1的值。 A.可能是2 B.一定是2 C.不可能是1 D.一定是1 11.设n个元素进栈序列是p1,p2,…,p n,其输出序列是1,2,3,……,n,若p n=1,则p i(1≤i≤n-1)的值。 A.n-i+1 B.n-i C.i D.有多种可能 12.判定一个顺序栈S为空的条件为。 A.S.top= =S.base B.S.top!= S.base C.S.top!= S.base+S.stacksize D.S.top= = S.base+S.stacksize 13.判定一个顺序栈S为栈满的条件是。 A.S.top-S.base= =S.stacksize B.S.top= = S.base C.S.top-S.base!=S.stacksize D.S.top!= S.base 14.链栈与顺序栈相比有一个明显的优点,即。 A.插入操作方便B.通常不会出现栈满的情况 C.不会出现栈空的情况D.删除操作更加方便 15.最不适合用作链栈的链表是。 A.只有表头指针没有表尾指针的循环双链表 B.只有表尾指针没有表头指针的循环双链表 C.只有表尾指针没有表头指针的循环单链表 D.只有表头指针没有表尾指针的循环单链表 16.如果以链表作为栈的存储结构,则退链栈操作时。 A.必须判别链栈是否满B.判别链栈元素的类型 C.必须判别链栈是否空D.对链栈不作任何判别

高中语文优秀说课稿范文3篇

(封面) 高中语文优秀说课稿范文3篇 授课学科: 授课年级: 授课教师: 授课时间: XX学校

高中语文优秀说课稿范文3篇 高中语文优秀说课稿范文1 一.关于教材: 我所说课的篇目是两千年前的爱国主义诗人屈原的代表作《离骚》,它是高中二年级的教学内容,具体编排在第三册第一单元(古典诗歌) 的第二篇。该单元所选中国古代古体诗8首,其中《诗经》中3首,汉 魏晋五言诗3首,《孔雀东南飞》并序以及《离骚》节选。在该单元中,《离骚》节选占有及其重要的位置。一方面,以屈原的作品为代表的楚 辞是继北方黄河流域出现《诗经》之后,在南方长江流域出现的具有独 特艺术魅力的新诗体,其本身在中国诗歌发展史上具有重要地位;另一 方面,楚辞作品,尤其是屈原的《离骚》从词句的角度看,地方特色鲜明,学生理解的难度较大。 从作品的内容看,《离骚》是一首规模宏伟的政治抒情长诗,全诗373句2400余字,既具有诗人自传的性质,又具有某些幻想性的浪漫主 义成分,整首诗歌感情回环激荡,撼人心魄。课文节选了长诗的第一节,内容主要是诗人自述身世、品质、政治理想以及自己的忠诚不为君王理 解的苦闷之情,表现了屈原坚持“美政”、至死不渝的高尚情操,抒发 了诗人忧国忧民、献身理想的爱国情感。 二.关于教学方法 结合作品特点,按照课程标准对高中阶段在古典诗歌方面阅读能力 的要求,我把《离骚》节选的教学目的确定为以下两点: 一是了解屈原的生平,理解诗人忧国忧民的高尚品质,培养学生的

爱国忧民的思想感情;二是反复诵读,抓住关键词语,疏通诗句,培养学生初步评价鉴赏古典诗歌的能力。学生是学习的主体,每当我们向自己的学生教授新知识的时候,自然要考虑到我们的教学对象在学习本课知识时的原有基础、现有困难及某些学习心理特征,从而有针对性地确定学习的重点、难点及教学对策。高中二年级学生,具有一定的古代诗歌的评价能力,但对于时代久远的楚辞作品,评价的前提――读懂诗歌的内容方面,仍存在困难,根据大纲要求结合学生的实际情况以及本文特点,我确立本文重点及难点如下: 教学重点: 鉴赏重要的语句,感受屈原高尚的人格美。 教学难点: 关键词句的疏通。如何突破重点、难点,有效完成本课的教学任务呢?我决定从教材特点和面临的具体学情出发,首先要把握两个原则:第一,注意高视角、低起点,先把基础知识落到实处,再循序渐进进行教学;第二,重视素质教育,有意识地对学生进行人文素质的培养。 具体教学措施如下: 第一,充分预习。要求学生课下预习,早读检查(内容包括:A作家作品;B写作背景;C掌握生字生词;D熟读课文),逐步培养学生的认知能力和自学能力。 第二,分解重点、难点。本课将分两课时进行,把教学难点,即词句的疏通安排在第一课时中,把教学重点,即作品思想内容的理解安排在第二课时,以确保学生真正做到一课一得。

高中语文说课稿

高中语文说课稿 导读:高中语文优秀说课稿范文1 一.关于教材: 我所说课的篇目是两千年前的爱国主义诗人屈原的代表作《离骚》,它是高中二年级的教学内容,具体编排在第三册第一单元(古典诗歌)的第二篇。该单元所选中国古代古体诗8首,其中《诗经》中3首,汉魏晋五言诗3首,《孔雀东南飞》并序以及《离骚》节选。在该单元中,《离骚》节选占有及其重要的位置。一方面,以屈原的作品为代表的楚辞是继北方黄河流域出现《诗经》之后,在南方长江流域出现的具有独特艺术魅力的新诗体,其本身在中国诗歌发展史上具有重要地位;另一方面,楚辞作品,尤其是屈原的《离骚》从词句的角度看,地方特色鲜明,学生理解的难度较大。 从作品的内容看,《离骚》是一首规模宏伟的政治抒情长诗,全诗373句2400余字,既具有诗人自传的性质,又具有某些幻想性的浪漫主义成分,整首诗歌感情回环激荡,撼人心魄。课文节选了长诗的第一节,内容主要是诗人自述身世、品质、政治理想以及自己的忠诚不为君王理解的苦闷之情,表现了屈原坚持“美政”、至死不渝的高尚情操,抒发了诗人忧国忧民、献身理想的爱国情感。 二.关于教学方法 结合作品特点,按照课程标准对高中阶段在古典诗歌方面阅读能力的要求,我把《离骚》节选的教学目的确定为以下两点:

一是了解屈原的生平,理解诗人忧国忧民的高尚品质,培养学生的爱国忧民的思想感情;二是反复诵读,抓住关键词语,疏通诗句,培养学生初步评价鉴赏古典诗歌的能力。学生是学习的主体,每当我们向自己的学生教授新知识的时候,自然要考虑到我们的教学对象在学习本课知识时的原有基础、现有困难及某些学习心理特征,从而有针对性地确定学习的重点、难点及教学对策。高中二年级学生,具有一定的古代诗歌的评价能力,但对于时代久远的楚辞作品,评价的前提――读懂诗歌的内容方面,仍存在困难,根据大纲要求结合学生的实际情况以及本文特点,我确立本文重点及难点如下: 教学重点: 鉴赏重要的语句,感受屈原高尚的人格美。 教学难点: 关键词句的疏通。如何突破重点、难点,有效完成本课的教学任务呢?我决定从教材特点和面临的具体学情出发,首先要把握两个原则:第一,注意高视角、低起点,先把基础知识落到实处,再循序渐进进行教学;第二,重视素质教育,有意识地对学生进行人文素质的培养。 具体教学措施如下: 第一,充分预习。要求学生课下预习,早读检查(内容包括:A 作家作品;B写作背景;C掌握生字生词;D熟读课文),逐步培养学生的认知能力和自学能力。

最新整理职高高一语文教案

最新整理职高高一语文教案 职高高一语文教案教学目标 1.引导学生理解课文蕴涵的哲理; 2.借鉴课文的写作方法,学习精彩语言; 3.培养学生的思维能力,使学生成为独立的思考者。 教学过程 一、导入新课 1.同学们,大自然中栖息着千姿百态的鸟儿,你见过黄鹂这种鸟吗?黄鹂不仅是一种美丽的鸟,而且几乎成为一种文化的象征,在古代诗人的笔下多有描写,大家能吟出几句吗? 有关“黄鹂”的古代诗句: 两个黄鹂鸣翠柳,一行白鹭上青天。(杜甫) 几处早莺争暖树,谁家新燕啄春泥(白居易) 千里莺啼绿映红,水村山郭酒旗风。(杜牧) 留连戏蝶时时舞,自在娇莺恰恰啼。(杜甫) 这些诗句通过描写黄鹂的啼鸣、飞舞表现黄鹂娇好美丽的形象,抒发对“杂花生树,群莺乱飞”的美好春天的热爱之情。今天我们要学习现代著名作家孙犁的一篇散文《黄鹂》,那么在孙犁的这篇散文里,黄鹂又是怎样的形象?作者抒发的又是怎样的感情呢?让我们来学习这篇课文。 2.作者简介 孙犁:现代著名作家。河北省安平县人,原名孙树勋。代表作品有长篇小说《风云初记》、中篇小说《铁木前传》、短篇小说《荷花淀》。其在抗日战争和解放战争初期创作的小说、散文结集为《白洋淀纪事》。其小说充满诗情画意,有“诗体小说”之称。是文学史上“荷花淀派”的代表作家。 二、整体感知 1.朗读 找四位同学分段落朗读课文,尽量把文章的思路体现出来。 2.(师)概括课文内容 通过刚才读课文,我们熟悉文章的内容,概括地说,作者叙述“我”和

黄鹂的几次相遇,其中有抗战时期的黄鹂,在青岛养病时见到的黄鹂,鸟市上的黄鹂,还有江南的黄鹂,通过这四次写黄鹂,作者表达自己的思想。 三、学习文章的思想内容(哲理) 读完这篇文章你有什么收获呢?先说思想上的收获,文章使你明白什么道理。 学生可能列举的内容与教师的分析引申: (1)要爱护鸟类,爱护动物,保护环境。 分析引申:可以这样理解。黄鹂在文中是美的化身,作者反抗人类对美的暴力破坏。 (2)各种事物都有极致。在一定的环境里,才能发挥这种极致。 分析引申:文中举哪些例子说明各种事物都有极致?黄鹂的极致是什么?黄鹂的禁锢和限制又是什么? 学生分析讨论 明确:说明各种事物都有极致的例子:虎啸深山,鱼游潭底,驼走大漠,雁排长空。 黄鹂的极致:它们的啼叫,是要伴着春雨、宿露,它们的飞翔,是要伴着朝霞和彩虹的。在这种环境下,黄鹂达到它们的极致。 黄鹂的禁锢和限制:①抗战时期,炮火连天的环境,美丽的黄鹂只是一闪而过,不能达到美的极致;②青岛养病时,病友史同志举枪瞄准,黄鹂成惊弓之鸟,从此一去不返,人类的射杀,是黄鹂不能达到美的极致的禁锢和限制;③逛鸟市时见到的黄鹂,被系在一根木棍上,一会儿悬空吊着,一会儿被拉上来,这成它们的禁锢和限制。(第三点学生一般能找出,第一点和第二点需要教师的引导) (3)(师)小结 所以,黄鹂只有在大自然的怀抱中自由飞翔,自由生活,才能活出它的本性,才能现出“美的极致”。而且,作者的思索由黄鹂转向大千世界,茫茫宇宙。既然黄鹂的啼叫、飞翔要伴随着春雨、宿露,朝霞、彩虹,那么,虎、鱼、驼、雁这些更加壮观的美的极致呢?难道不需要深山、潭底、大漠、长空这样的美的极致?美的极致,只有在一定的环境里,才能发挥:这是从平凡生活中升华

相关主题
文本预览
相关文档 最新文档