I have a collection of news articles. I have performed tf-idf operations on them. I am using python as programming language so it was just the use of TF-IDF vectorizer function. I now have the document's tfidf representation. My aim is to cluster news articles using Particle swarm optimization.

As read in following research papers,

  1. https://www.researchgate.net/publication/4170865_Document_clustering_using_particle_swarm_optimization

  2. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0188815

  3. https://www.sciencedirect.com/science/article/pii/S1877750316305002

the input of the PSO algorithm should be an array representing document. In my case I am trying to use tf-idf as input. So, how can i use the result of tf-idf as input for Particle Swarm optimization? Particle Swarm optimization is a optimization algorithm. So while clustering, I am going to optimize or minimize distance between cluster's centroid and data points belonging to that cluster. Can anybody provide me a solution or hint towards something?

  • $\begingroup$ Particle swarm optimization is used for "optimizes a problem by iteratively trying to improve a candidate solution with regard to a given measure of quality". Can you edit the question to add a reference where PSO is used for clustering? $\endgroup$ – John L. Mar 20 '19 at 14:27
  • $\begingroup$ Let us continue this discussion in chat. $\endgroup$ – Annonymous programmer Mar 20 '19 at 18:49
  • $\begingroup$ You mention you've read some research papers where this is done. As Apass.Jack suggested earlier, I suggest you edit the question to provide references to the papers you read that do this, and summarize what you do understand about them and what you don't understand. I still suspect this is an XY problem. I suspect what you really want to do is cluster the documents, and there are many reasonable methods for clustering; it's not clear why you are so focused on PSO-based clustering as opposed to any other method for clustering. $\endgroup$ – D.W. Mar 21 '19 at 3:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.