This repo is a collection of scripts we use at Xeneta to qualify sales leads with machine learning. Read more about this project in the Medium article Boosting Sales With Machine Learning.
You can use this repo for two things:
- Try to beat our predictions using our data and your own algorithm
- Create a lead qualifier for your company, using your own data
Start off by running the following command:
pip install -r requirements.txt
You'll also need to download the stopword from the nltk package. Run the Python interpreter and type the following:
import nltk
nltk.download('stopwords')
We'd love to see more algorithms on the leaderboard, so send us a pull request once you've implemented one.
We've provided you with our vectorized and transformed data here. We can unfortunately not share the raw text data, as it contains sensitive company information (who our customers are).
To test our your own algorithm, simply add it the run.py file and run the script:
python run.py
Thanks to lampts for implementing the best performing algorithm so far, the SGDClassifier.
| Algorithm | Precision | Recall | F1 Score |
|---|---|---|---|
| SGD Classifier | 0.872 | 0.940 | 0.905 |
| Random Forest | 0.845 | 0.915 | 0.878 |
PS: We're also experimenting with a neural net (in TensorFlow) in the nn.py file.
To create your own lead qualifier, you'll need to get hold of company descriptions (to create your dataset). We currently use FullContact for this.
Note: We've added dummy data, so that you can run both scripts without getting errors, and to give you examples on how the sheets should look like.
This script trains an algorithm on your own input data. It expects two excel sheets named qualified and disqualified in the input folder. These sheets need to contain two columns:
- URL
- Description
Run the script:
python run.py
It'll dump three files into the qualify_leads project:
- algorithm
- vectorizer
- tfidf_vectorizer
You're now ready to start classifying your sales leads!
This is the script that actually predicts the quality of your leads. Add an excel sheet named data in the input folder. Use the same format as the example file that's already there.
Run the script:
python run.py
It'll output an excel sheet with a column named Prediction, where 1 equals qualified and 0 equals disqualified:
Got questions? Email me at per@xeneta.com.