Unstructured-data-mart-sentimental-analysis
Developed a classification model to perform sentimental analysis on Amazon food reviews .
Performed text pre-processing techniques to build a vector space model using TF*ID in Python notebook Implemented topic modeling using latent dirichlet allocation to discover topics from text reviews Achieve an accuracy of 87% using logistic regression
ETL data into HDFS to build unstructured data mart using HiveQL to perform OLAP operations Developed dashboards for reports by Integrating Cloudera Big Data source into tableau to make informed decisions from sentimetal analysis Technologies used: sklearn, numpy, pandas, nltk, HDFS, HiveQL, MS Azure ML studio, tableau, predictive modeling