参考文献：
- 拼写纠错功能实现
- 贝叶斯公式与拼写检查器
- big.txt
- 《数学之美》、《统计学习方法》

github源码：

https://github.com/hlk-1135/Dictionary

算法核心：贝叶斯算法：

运行效果：

当我们输入的单词有误时，贝叶斯算法开始派上用场了，对我们所输入的单词进行检查纠正，并进行一定的猜测。

public class SpellChecker {

    private static final char[] alphabets = "abcdefghijklmnopqrstuvwxyz".toCharArray();

    public void start() throws IOException {
        //1.构建语言模型
        String path = "E:\\big.txt";
        Map<String, Double> languModel = buildLanguageModel(path);
        Set<String> dictionary = languModel.keySet();

        while((input = reader.readLine()) != null) {
            input = input.trim().toLowerCase();
            if("bye".equals(input))
                break;
            if(dictionary.contains(input))
                continue;
            long startTime = System.currentTimeMillis();

            //3.在编辑距离内设置一个单词集，并删除字典中不存在的单词
            Set<String> wordsInEditDistance = buildEditDistance1Set(languModel, input);
            wordsInEditDistance.retainAll(dictionary);
            if(wordsInEditDistance.isEmpty()) {
                  wordsInEditDistance = buildEditDistance2Set(languModel, input);
                  wordsInEditDistance.retainAll(dictionary);
                  if (wordsInEditDistance.isEmpty()) {
                         System.out.println("Failed to check this word!");
                         continue;
                  }
            }
            // 4.计算所以可能的概率
            List<String> guessWords = guessRightWord(languModel, wordsInEditDistance);
            System.out.printf("Do you want to input %s and Cost time: %.10f second(s)\n",
                         guessWords.toString(), (System.currentTimeMillis() - startTime) / 1000D);
        }
    }

    /**
     * 读取语料库big.txt，构建模型
     * @param path
     * @return
     * @throws IOException
     */
    private Map<String, Double> buildLanguageModel(String path) throws IOException {
        Map<String, Double> languModel = new HashMap<String, Double>();
        BufferedReader reader = new BufferedReader(new FileReader(path));
        //去掉文档中除字母外的所有符号
        Pattern pattern = Pattern.compile("[a-zA-Z]+");
        String line;
        int totalCount = 0;
        while ((line = reader.readLine()) != null) {
            String[] words = line.split(" ");
            for(String word : words) {
                if(pattern.matcher(word).matches()) {
                    word = word.toLowerCase();
                    Double wordCount = languModel.get(word);
                    if(wordCount == null) {
                        languModel.put(word, 1D);
                    } else {
                        languModel.put(word, wordCount+1D);
                    }
                    totalCount++;
                }
            }
        }
        reader.close();

        for(Entry<String, Double> entry : languModel.entrySet())
            entry.setValue(entry.getValue() / totalCount);

        return languModel;
    }

    /**
     * 编辑距离为1的单词集合
     * @param languModel
     * @param input
     * @return
     */
    private Set<String> buildEditDistance1Set(Map<String, Double> languModel,String input) {
        Set<String> wordsInEditDistance = new HashSet<String>();
        char[] characters = input.toCharArray();

        // 删除：删除一个字母的情况，delete letter[i]
        for(int i=0;i<input.length();i++) {
            wordsInEditDistance.add(input.substring(0,i) + input.substring(i+1));
        }
        // 换位: 交换letter[i] and letter[i+1]
        for(int i=0;i<input.length()-1;i++) {
            wordsInEditDistance.add(input.substring(0,i) + characters[i+1] 
                    + characters[i] + input.substring(i+2));
        }
        // 替换: 将 letter[i]替换为a-z
        for(int i=0;i<input.length();i++) {
            for(char c : alphabets) {
                wordsInEditDistance.add(input.substring(0,i) + c + input.substring(i+1));
            }
        }
        // 插入: 插入一个新的字母 a-z
        for(int i=0;i<input.length()+1;i++){
            for(char c : alphabets) {
                wordsInEditDistance.add(input.substring(0,i) + c + input.substring(i));
            }
        }
        return wordsInEditDistance;
    }

    /**
     * 编辑距离为2的集合.通过editDistance1函数得到编辑距离为1的集合,
     * 该集合单词再通过editDistance1函数,就可以得到编辑距离为2的集合 
     * @param languModel
     * @param input
     * @return
     */
    private Set<String> buildEditDistance2Set(Map<String, Double> languModel,String input) {
        Set<String> wordsInEditDistance1 = buildEditDistance1Set(languModel, input);
        Set<String> wordsInEditDistance2 = new HashSet<String>();
        for(String editDistance1 : wordsInEditDistance1) {
            wordsInEditDistance2.addAll(buildEditDistance1Set(languModel, input));
        }
        wordsInEditDistance2.addAll(wordsInEditDistance1);
        return wordsInEditDistance2;
    }

    /**
     * 从语料库中获取正确单词
     * @param languModel
     * @param wordsInEditDistance
     * @return
     */
    private List<String> guessRightWord(final Map<String, Double> languModel,Set<String> wordsInEditDistance){
        List<String> words = new LinkedList<String>(wordsInEditDistance);
        //按照单词在字库中出现的频率大小排序，频率越大出现的可能性越大  
        Collections.sort(words, new Comparator<String>() {
            @Override
            public int compare(String word1, String word2) {
                  return languModel.get(word2).compareTo(languModel.get(word1));
            }
        }); 
        return words.size() > 5 ? words.subList(0, 5) : words;
    }
}