Differences

This shows you the differences between two versions of the page.

--- java:determinesentimentinspanish [2014/07/05 23:36] – created rlunaro
+++ java:determinesentimentinspanish [2014/07/07 13:37] – rlunaro
@@ Line 3: / Line 3: @@
 ===== Intro =====
-It's surprising that up today (07/05/2014), the search on internet of this information gives very few results. Sentiment determination is critical for mining the user's opinion information. In today's age, with the vast use of social networks, the users opinion and their emotions (anger, hapiness) and their sentiment (positive or negative) is very important to detect how users react -by example- regarding a particular brand, political party or administration.
+It's surprising that up today (07/05/2014), the search on internet of this information gives very few results((Search by "determinación de orientación de texto español" or by "identificación de emociones en un texto en español". In the first case, there are no "correct" results and in the second there are only five relevant results up to 07/07/2014)). Sentiment determination is critical for mining the user's opinion information. In today's age, with the vast use of social networks, the users opinion and their emotions (anger, hapiness) and their sentiment (positive or negative) is very important to detect how users react -by example- regarding a particular brand, political party or administration.
 I am starting this page as a trial to get:
@@ Line 13: / Line 13: @@
 http://www.programcreek.com/2012/12/how-to-make-a-web-crawler-using-java/
+===== My notes =====
+I think that the process of some textual information will be:
+  - De-abbreviate the sms language (salu2 -> saludos, k -> que, etc.)
+  - Clean ortographic errors and typos
+  - Entities detection (Barack Obama, New York, Microsoft...) names that are proper names
+  - Detect the sentiment of the text