Use sklearn or existing module to cluster text let each cluster belongs to multilabel -
I have such data
['The Art of Stactics in New York' ',' Actor Sword in New York. Actor Pretty Johnny Deep in New York ',' France, The Great Tower in Las Vegas, great hotel. ',' Empire State Building. This night is great! 'Empire State Building Wow! This night is so great! ',' The light of life you think about the future Love. ',' Do you think the road lights are yesterday where my road is? ',' Peeling a cat with a scroll on the ground ',' A wonderful woman in a dark rainy day. ',' The cat sleeps in the sun. ',' Three in one image is a cat's fragrance. ',' A woman is watching with a smile, falling leaves ',' a headache of a fat squirrel ',' a headache of yellow cat. 'A cat runs on ice and jumps too much. ',' The falling leaves and the appearance of a dog farming in the camera. ',' A cat points towards the camera, looking at one and the other side. ',' The long-dog dog plays on the grass. "A tragic dog wants to see the world outside the house." ] I want to cluster them and label each of them with more than one label. Is it possible to use existing equipment? I do not have enough time to implement myself Thank you very much!
You can try to use gensim : You can use the LDA, a method that gives you a distribution (subject) of tags for each of your documents (phrases). Then you can use any method to cluster based on the distribution of those topics - such as K-Means, for example, which will work
Please take a look:
And then you use the K-sense method from Skylare.
Comments
Post a Comment