Question

Does there is any library or algorithm that can do auto detection of tags in a text (ignoring the usual words of the chosen language)?

Something like this:

string[] keywords = GetKeyword("Your order is num #0123456789")

and keywords[] would contain “order” and “#0123456789” …?
Does it exist? Or the user will select by himself all the tags of every document all the time? 😕

Answer 1

I think that a lot of different algorithms can be used. Some of them are simple another are super complex. I can suggest you the next basic way:

1. Split all text into array of words.
2. Remove stop words from the array. (Goole “stop words list” to get full list of stop words.)
3. Walk through the array and calculate count of each word.
4. Sort words in accordance with their ‘weight’ in the array.
5. Choose necessary amount of tags.