The Truth About Sentiment & Natural Language Processing - Synthesio

Synthesio currently uses human analysts for sentiment analysis but can ... we advance toward what is referred to as the singularity, it seems as though the best option .... analysis cannot understand sentiment in the context of your business.
2MB Sizes 0 Downloads 109 Views
 

                       

The  Truth  About  Sentiment  &  Natural   Language  Processing     By  Synthesio                    

Summary   Introduction  .2   Artificial  Intelligence’s  difficulties  with  sentiment  .3   Human  analysis  is  an  obligatory  step  when  analyzing  web  content  .5   Current  technological  advances  .5   The  future  of  semantic  technology  .8   .7.Conclusion  .10       Synthesio  –    The  Truth  About  Natural  Language  Processing  -­‐  March  2011  

 

1  

 

Introduction    

The   web   has   made   it   possible   for   brands   to   discover   what   people   are   saying   about   their   brands   online,   either   in   mainstream   media   like   online   newspapers  and  magazines,  or  on  social  media.  Consumers  now  search  for  opinions  online   before,  during,  and  after  a  purchase.  The  next  step  for  brands  is  finding  out  whether  people   are   talking   positively   or   negatively   about   their   brand,   and   why.   Some   online   ratings   provide   a  number  but  not  the  reasoning  behind  it,  and  may  only  present  half  of  the  story.   Numerous  companies  have  been  working  on  text  mining  for  close  to  30  years  in  some  cases,   thus   sentiment   analysis   is   not   a   new   area   but   it   has   become   a   hot   topic   thanks   to   social   media.   Social   media   monitoring   companies,   as   well   as   PR   practitioners,   and   digital   marketers   in   general,   have   waged   debate   over   whether   sentiment   should   be   analyzed   by   man   or   machine.   Synthesio   currently   uses   human   analysts   for   sentiment   analysis   but   can   add  natural  language  processing  capacities  on  a  case-­‐by-­‐case  basis.   Although  technology  is  quickly  advancing  to  catch  up  on  its  lag  behind  human  analysis,  as   we  advance  toward  what  is  referred  to  as  the  singularity,  it  seems  as  though  the  best  option   is  currently  combining  both  machine  and  man.                                         Synthesio  –    The  Truth  About  Natural  Language  Processing  -­‐  March  2011  

 

2  

 

Artificial  Intelligence’s  difficulty  with  sentiment     One  way  that  researchers  have  attempted  to  classify  sentiment  is  by  creating  a  “sentiment   lexicon”       Sentiment  is  not  analyzed  via  artificial  intelligence,  as  some  people  may  be  tempted  to  think.  Rather,  it  is  analyzed  via  a   systematic  process  that  involves  the  use  of  a  sentiment  lexicon.  This  lexicon  assigns  a  degree  of  positivity  or  negativity  to  a   word  by  itself  that  is  then  used  to  give  meaning  to  the  entirety  of  the  article.  This  is  a  way  of  analyzing  sentiment,  then,  by   considering   a   type   of   inherent   positivity   or   negativity   of   each   word   that   would   be   used   by   someone   to   talk   about   your   business  or  products.  For  example,  “happy”  would  be  deemed  a  positive  word,  as  well  as  “like”  and  “love”.  At  the  opposite   end  of  the  spectrum  we  can  see  words  like  “hate”,  “dislike”,  etc.       There  are  two  problems  with  this  methodology,  however.  The  first  problem  is  that  this  assigning  of  positive  and  negative   sentiment  evaluates  a  word  without   the  context   of   what   is   around  it.  The  dictionary   is   extremely   limited   in   the   number   of   words  that  will  always  attach  a  positive  or  negative  sentiment  to  an  expression.  The  second  problem  is  that  researchers   may  assign  different  degrees  of  positivity  to  a  word.  Particularly  in  the  case  of  ambiguous  expressions,  a  researcher  may  be   more  inclined  to  note  a  word  as  more  or  less  positive.      

Text  categorization  classifies  articles  by  topic1     Text  categor