Suppose we have predefined list of tags Tag #1
, Tag #2
, ..., Tag #N
and we want to assign tags to sample texts based on keywords or semantic analysis. A text can have multiple Tags. I don't want to consider it as classification problem assuming each tag as a separate class. What algorithm we can use here, Any suggestions would be of great help.
$\begingroup$
$\endgroup$
Add a comment
|
$\begingroup$
$\endgroup$
This is a multi-label classification task. Several labels can be assigned to each text. You can do this task using LSTM and sigmoid functions. The sigmoid functions are in the output layers and number of them is equal to number of your tags.