查询企业名称的分词问题

我要搜索比如：北京希尔顿酒店、北京金茂威斯汀大饭店、北京乐多港万豪酒店。
当用户输入： 希 或 希尔 或 希尔顿  的时候都能收到。
 
那么我现在的思路就是：将 企业名称（希尔顿）配词， 并在建立索引的时候，把 北京希尔顿酒店 拆分成：
北、京、北京、希、尔、顿、希尔顿、酒、店、酒店。
 
 
分词思路：
    1）    希尔顿配词
    2）    先用ik_max_word 分词   (分词为)==> 北京、希尔顿、酒店
    3）    使用es自带filter对 ik分的词进一步细化。  比如：ngram  （但ngram并不理想）
 
 
问题： 我想问下大家，有没有更好的思路。 或者  有没有es自带的filter能达到这个效果？
    
 
 
 
参考：
{
    "analysis":{
        "filter":{
            "pinyin_filter":{
                "keep_joined_full_pinyin":"true",
                "lowercase":"true",
                "keep_original":"true",
                "keep_none_chinese_together":"true",
                "remove_duplicated_term":"true",
                "keep_first_letter":"true",
                "keep_separate_first_letter":"false",
                "type":"pinyin",
                "keep_none_chinese":"true",
                "limit_first_letter_length":"16",
                "keep_full_pinyin":"true"
            },
            "ngram_filter":{
                "type":"ngram"
            }
        },
        "analyzer":{
            "ik_pinyin_analyzer":{
                "filter":[
                    "lowercase",
                    "ngram_filter",
                    "pinyin_filter",
                    "unique"
                ],
                "type":"custom",
                "tokenizer":"ik_max_word"
            },
            "suggest_search_analyzer":{
                "filter":[
                    "lowercase"
                ],
                "tokenizer":"keyword"
            },
            "pinyin_analyzer":{
                "tokenizer":"my_pinyin"
            }
        },
        "tokenizer":{
            "my_pinyin":{
                "keep_joined_full_pinyin":"true",
                "lowercase":"true",
                "keep_none_chinese_in_joined_full_pinyin":"true",
                "keep_original":"true",
                "keep_none_chinese_together":"true",
                "remove_duplicated_term":"true",
                "keep_separate_first_letter":"false",
                "type":"pinyin",
                "keep_none_chinese":"false",
                "limit_first_letter_length":"50",
                "keep_full_pinyin":"false"
            }
        }
    },
    "number_of_shards":"1",
    "number_of_replicas":"1"
}