News Articles Classification
14101,14108
Mtech. Computer Science
Part 1
DCST, Goa University
Outline
Aim
 To classify the news article's
according to the labels such as
Entertainment, Sports, Business,
Health and Technology.
 Categorize news articles into
predetermined Labels without
predefined feature vectors.
Procedure
 Data collection
◦ News articles are taken from pre labeled
groups such as Google News.
◦ Some pre-processing is also done on data.
 Training Phase
◦ Naive based classifier is used to build a
model in the training phase
 Testing Phase
◦ Given some unlabeled articles, naive based
classifier classifies it on the basis of learned
probabilities.
Clustering
 We use K-means clustering algorithm
to cluster the articles into groups.
Thank You

News articles classification

  • 1.
    News Articles Classification 14101,14108 Mtech.Computer Science Part 1 DCST, Goa University
  • 2.
  • 3.
    Aim  To classifythe news article's according to the labels such as Entertainment, Sports, Business, Health and Technology.  Categorize news articles into predetermined Labels without predefined feature vectors.
  • 4.
    Procedure  Data collection ◦News articles are taken from pre labeled groups such as Google News. ◦ Some pre-processing is also done on data.  Training Phase ◦ Naive based classifier is used to build a model in the training phase  Testing Phase ◦ Given some unlabeled articles, naive based classifier classifies it on the basis of learned probabilities.
  • 5.
    Clustering  We useK-means clustering algorithm to cluster the articles into groups.
  • 6.