Java：对短语列表进行模糊搜索

Question

我有两个字符串Patterns和Queries列表，如下所示，并在不使用任何库的情况下执行模糊搜索。

   P = {"i would have", "they love", "i analyzed", "made decision"};
   queries = {"since this morning i have analyzed all the data from the company and made my decision",
            "they love the idea"};

A fuzzy search is that there can be at max extra one word in the phrase.
Fuzzy search all the phrases in the query string and the expected output search phrases
are -> [["i have analyzed", "made my decision"], ["they love"]]

我能够想出蛮力解决方案，它适用于上述用例

private static ArrayList<ArrayList<String>> fuzzy_search(String[] P, String[] queries) {
    ArrayList<ArrayList<String>> result = new ArrayList<>();
    for(String query : queries){
        ArrayList<String> queryTokens = new ArrayList(Arrays.asList(query.split(" ")));
        ArrayList<String> matches = new ArrayList<>();
        for(String pattern : P){
            String[] patternTokens = pattern.split(" ");
            ArrayList<Integer> ii = new ArrayList<>();
                for (String patternToken : patternTokens){
                    int i = queryTokens.indexOf(patternToken);
                    if(i >=0)
                        ii.add(i);
                    else {
                        ii.clear();
                        break;
                    }
                }
            if(!ii.isEmpty()){
                int n = ii.get(ii.size()-1) - ii.get(0);
                if(n==ii.size() || n == ii.size()-1){
                    StringBuilder sb = new StringBuilder();
                    for(int kk = ii.get(0); kk< ii.get(ii.size()-1)+1; kk++) {
                        sb.append(queryTokens.get(kk));
                        if (kk != ii.get(ii.size() - 1))
                            sb.append(" ");
                    }
                    matches.add(sb.toString());
                }
            }
        }
        result.add(matches);
    }
    return result;
}

但它不适用于像下面这样的复杂示例。 有人可以帮我找到适用于以下复杂场景的有效解决方案吗

{
  "queries": [
    "close box son add bad strong afford down bad middle",
    "business become for department chap offer business become for chap tell afford aware bad name business become for committee chap normal around egg expect learn prepare balance goodbye lot forget welcome"
  ],
  "phrases": [
    "balance goodbye lot forget",
    "health contact ever big",
    "business become for chap",
    "approach field",
    "else after blow inform",
    "around egg expect learn",
    "add bad",
    "increase accept",
    "close box",
    "afford bad"
  ],
  "solution": [
    [
      "close box",
      "add bad",
      "afford down bad"
    ],
    [
      "business become for department chap",
      "business become for chap",
      "afford aware bad",
      "business become for committee chap",
      "around egg expect learn",
      "balance goodbye lot forget"
    ]
  ]
}