Mongodb-获取字段匹配特定值的文档的不同计数

Question

这是我的示例文档：

{
"sessionId" : "98e9be38-3ff9-454a-aef4-625acaadbb62",
"userMessage" : "Liked It :)",
"hour" : "14"
},
{
"sessionId" : "98e9be38-3ff9-454a-aef4-625acaadbb62",
"userMessage" : "Hated It :(",
"hour" : "14"
},
{
"sessionId" : "98e9be38-3ff9-454a-aef4-625acaadbb62",
"userMessage" : "Just okay, could be better",
"hour" : "14"
}      
{
"sessionId" : "98e9be38-3ff9-454a-aef4-625acaadbb62",
"userMessage" : "Sample Message",
"hour" : "14"
},
{
"sessionId" : "98e9be38-3ff9-454a-aef4-625acaadbb62",
"userMessage" : "Sample message 2",
"hour" : "14"
} 

我想计算将userMessages匹配为'Liked It :)'或'Hated It :('或'Just okay，可能会更好'的文档数。

预期输出：

{[
   "Liked It :)":5,
   "Hated It :(":3,
   "Just okay, could be better":1
]}  

我尝试过的事情：

db.getCollection('messagelogs').aggregate([
{"$group" : {_id:"$userMessage", count:{$sum:1}}}
])  

但是上面的查询将所有'userMessage'字段分组，这不是我想要的，我只希望将'userMessage'字段分组为'Liked It :)'，'Hated It :('和'Just好吧，可能会更好'