CIKM AnalytiCup 2017: Lazada Product Title Quality Challenge Forum

Go back to competition Back to thread list Post in this thread

> Getting Started

Welcome to the Competition!

This challenge tests your data science skills to advance the state-of-the-art in analyzing product titles in e-commerce and assess their quality. Win your share of SGD15,000 in cash prizes, plus an invitation to present your solution on November 6 at CIKM 2017.

The challenge will run from 1 May to 31 July 2017. Don’t lose time, start working on your solution now! To get started, please first review all the information. If you have further clarification, please post your questions on this forum, so the other participants can both contribute and benefit as well.

We have also provided a few baselines on the leaderboard. BaselineRandom consists of random predictions. The other three are BaselineDecisionTree, BaselineNaiveBayes, and BaselineLogisticRegression. These are meant for reference points. The implementation codes of these non-random baselines are also made available by their respective authors in the following forum postings. These codes are provided "as is" and without warranty of any kind. A quick way to get started is to download and extend one of these baselines. You are of course strongly encouraged to develop your own solutions.

Wish you all the best!

Posted by: hadylauw @ April 28, 2017, 5:33 a.m.

Hey guys,

Here is the code for BaselineNaiveBayes: . It's a multinomial NB model, which uses the title tokens only. The code is in Python (scikit-learn).

Cheers and Good Luck!

Posted by: mtkachenko @ April 28, 2017, 9:54 a.m.
Post in this thread