Since last week, when the current university semester officially ended, I’ve found a little time to work on the things I love. Learning data science and machine learning—itself an endless journey—helps me think about new problems and prepare possible approaches for them in my mind.
One class of these problems is the application of machine learning to Persian literature. Those of you familiar with machine learning surely know how many fascinating problems can be defined in this area—problems that help us understand and expand upon it.
Last week I decided to start one of the projects I’d been wanting to do and devote part of my between-semester break to it. This project was a study of Rumi’s poetry. I downloaded the data from the Ganjoor website and began working on the Divan-e Shams. Understanding the database structure took a little time, but thanks to its simple and elegant design, I quickly figured out how to find the verses. The initial results were very interesting to me.
One of the significant problems I ran into was the lack of libraries and trained models for the Persian language. I hope the number of these libraries and models keeps growing day by day. I kept thinking about how many interesting interdisciplinary projects could be defined at the intersection of Persian literature and data science.
Though you have cut yourself away from me, I am not without hope;
Though you have chosen another beloved over me,
As long as I have life, I will grieve for you—
For within despair there lie many hopes.
— Rumi
Before I forget, let me mention that Ganjoor is one of my favorite websites. I’m grateful to its founder and to everyone who works on it. May you endure.
Selected verse:
As long as you seek the jewel of the mine, you are the mine;
As long as you crave a morsel of bread, you are that bread.
If you grasp this hidden secret, you will know:
Whatever you are searching for—you are that.
— Rumi
