Social Media Analysis Platform

Sentiment Analysis

SUMMARY

The client needed a technological enhancement to their current sentiment analysis platform that could also be fine-tuned for efficient parallel processing. A dynamic reporting dashboard by persisting the sentiment of the market by harvesting the reviews by analyzing various social media platforms like Facebook, Twitter and several other online portals was achieved.

CLIENT PROFILE

The client is a market leader in providing complete social media management platforms for enterprises. The client helps enterprises leverage social media channels such as Facebook, LinkedIn, Twitter etc., for advertising, care, sales, research, and commerce, on one integrated platform.

BUSINESS CRITICAL CHALLENGES

The client works on the concept of extracting the sentiments out of the documents/reviews that are being fetched from online sites. Sentimental analysis is done by scrapping reviews from social media to divide them into parts for further analysis and yield different decision-making results/suggestions. The analysis will yield the result whether the sentiment is positive or negative or neutral. The reviews are continuously harvested from the online social media sites to showcase the sentiments and the analysis/result dashboard to be more dynamic as the sentiment analysis is more dynamic in nature. The current solution to extract the existing sentiment analysis platform has limitations in extracting the reviews from the online sites and the backend services were not tuned to optimize the best utilization of the available system resource. Out of various sentiment analysis methodologies, our client preferred to focus primarily on the 3 that give a better decision-making ability to organizations: attitude, opinion, and emotion.

OUR SOLUTION APPROACH

Parkar made noticeable improvements in all of these components by maintaining the existing solution for the sentimental analysis and made significant improvements to their current processes. Parkar re-architected the current solution by introducing various parallel computation and simultaneous execution of the process that helped the client to get real-time reviews from different social media sites for a real-time simulation and data on the reporting dashboards. The data is persisted in the database to make them available in various reporting formats in a single dashboard.

© 2017 Parkar Consulting Group

info@parkar.consulting

1844 8246 686