01575nam a22001337a 4500999001700000100006600017245005500083260003000138300000900168500110000177700004201277942001101319952011101330  c65861d65858  aZahid Hussaina12MSIT19aSupervisor - Dr. Shah Zaman Nizamani  aEntity based sentiment analysis of electronic news  aNawabshah:bQUEST,c2019.  a37p.  aABSTRACT

News articles and electronic media provide us massive unstructured bulk information about day by day events. Every day millions of news and articles are published in newspapers. The news and articles give some information regarding physical or virtual entities. In the news or articles three categories of views are given. The view may be positive, negative or neutral. The view portrays by the news or articles is important for improving the performance of an entity. The aim of this research is to find the entity based sentiment analysis of the news. With the help of sentiment analysis polarity (Positive. Negative. Neutral) of any entity can be get. For this research a data crawler tool is developed in java language and sentiment analysis is done by using Stanford CoreNLP. For testing purpose the articles for taken from website of dawn newspaper. For sample data, news were downloaded continuously for seven weeks regarding some entities.

Keywords: Entity Based Sentiment Analysis, NLP (Natural Language Processing) Data Crawling, Electronic News articles, Stanford CoreNLP .  aDepartment of Information Technology   cTHESIS  00104070aQUESTCLbRESEARCHd2019-10-15l0oR/IMS-19pMP/53-652r2019-10-15 00:00:00w2019-10-15yTHESIS