The third instalment of the adrenaline rush hackathon series is here. This week we challenge the data science community to predict the polarity of a text message or email in MachineHack’s Message Polarity Prediction: Weekend Hackathon #3.
The challenge will start on May 1st, Friday at 6 pm IST.
Problem Statement & Description
All of us receive a ton of messages and emails on a daily basis. Collectively that is a lot of data which can provide useful insights about the messages that each of us gets. What if you could know whether a certain message has brought you good news or bad news before opening the actual message. In this challenge, we will use Machine Learning to achieve this.
Given are 53 distinguishing factors that can help in understanding the polarity (Good or Bad) of a message, your objective as a data scientist is to build a machine learning model that can predict whether a text message has brought you good news or bad news.
You are provided with the normalized frequencies of 50 words/emojis (Freq_Of_Word_1 to Freq_Of_Word_50) along with 3 engineered features listed below:
- TotalEmojiCharacters: Total number of individual emoji characters normalized. (eg. :) )
- LengthOFFirstParagraph: The total length of the first paragraph in words normalized
- StylizedLetters: Total number of letters or characters with a styling element normalized
Target Variable: IsGoodNews
The data sets will be made available for download on Friday, May 1st 6 pm IST.
Data Description
The unzipped folder will have the following files.
- Train.csv – 947 observations.
- Test.csv – 527 observations.
- Sample Submission – Sample format for the submission.
Below are the file formats for the provided data
Train.csv
Test.csv
Sample_Submission.xlsx
Bounties
The top 3 competitors will receive a cool AIM goodie bag and a free pass to the plugin.
plugin, India’s largest virtual conference on AI, is a next-gen disruptive conference that brings AI professionals from around the world together in a virtual setting.
Rules
- One account per participant. Submissions from multiple accounts will lead to disqualification
- Participants may submit any number of times for this hackathon
- All registered participants are eligible to participate in the hackathon
- This competition will count towards your global ranking points on MachineHack
- You will not be able to submit once you click the “Complete Hackathon” button. You may ignore this feature
- We ask that you respect the spirit of the competition and do not cheat
- This hackathon will expire on 4th May, Monday at 7 am IST
Evaluation
The leaderboard is evaluated using the F1 Score for the participant’s submission.