IST 736 HW4

less than 1 minute read

HW4

MOST FINAL – sentiment MOST FINAL – deception

SENTIMENT

NEWEST ITERATIONS

HW 2 Pipeline – Sentiment

HW 2 Pipeline – Deception (lie_false = neg)

HW 2 Pipeline – Deception (lie_false = pos)

Iterations!

V1 CLEANED!

  • With export to csv
  • With export to corpus

Run with HW1 Sentiment Analysis Pipeline

Run with HW2 Sentiment Analysis Pipeline

DECEPTION

Run with Sentiment

Run with Deception

?!!?! I switched the ‘true’ and ‘false’ So instead of reading the “true” in the file as “this review is true” We’re reading it as “is lie? True” MUCH.BETTER.RESULTS.

Run with Deception

RESOURCES & READING:

great whitepaper

great presentation

great citations

ISSUES!

Issues importing csv

Had to change delimiter to

dirtyFile = pd.read_csv('deception_data_converted_final.csv',  sep='\t')

see here

Which caused more problems, of course :)

Updated: