Search code examples
python-3.xnlpcluster-analysisgensimtopic-modeling

How to map detailed text to a unigram or a bigram


I am trying to figure out solution for requirement where in I am required to map long text to unigrams or bigrams. For example "Ability to motivate and manage team. You should be able to track the progress of the team and intervene to improve the progress". This long text should be mapped to "Team management". Basically I am trying to figure out communication/analytical skills from the long text seen in document like Job descriptions. I am struggling to figure out a solution for this. I do not want to hard code as the long text keep changing. Thanks for any help.


Solution

  • This seems like a multi-class multi-label classification problem. There is a lexical way of solving this problem. The article here will help you in solving the problem. One could skip the sentiment analysis part as the focus of this question is topic mapping.