With  the  highly  expanding  usage  of 
social  media,  a  huge  amount  of  data  is  being 
collected  day by day.  Proliferation of movies and 
their  related  views  on  social  media  has  posed 
significant  challenges  in  discovering  the  most 
suitable movies.  Personalized recommenders that 
have been implemented  to recommend movies use 
only  a  source  of  data  and  largely  overlooked 
integration  of  several  big  data  sources  including 
tweets,  you  tube  comments,  posts  on  movies  and 
comments.  Moreover,  most  of  the  current 
approaches  focus  on  one  single  aspect  i.e.  either 
content-based, collaborative filtering. Overlooking 
the integration of multiple aspects with  many data 
provenances  caused  low  effectiveness  in  current 
approaches. Hence, to overcome these deficiencies 
a  novel,  hybrid  approach,  KLC  model  which 
integrated  user-based  collaborative  filtering, 
locality  sensitive  hashing  and  network-based 
approach  is  proposed.  The  New  KLC  model  has 
been  tested  with  10000000  data  items  it 
outperforms benchmark methods