Turkish Song Lyrics Analysis with NLP and Song Generation with Deep Learning

Songs are a crucial part of a culture. It represent the language of emotions which makes this dataset interesting for answering many questions with NLP techniques.

Inspiration

This dataset is a great resource for applying NLP techniques. Many different methods from basic analysis like wordcloud to lyrics generation with LSTM neural networks can be applied.

Web Scraping

The dataset is scraped from lyrics websites using BeautifulSoup library. Disclaimer : The file might contain missing songs, lyrics in a different language, or incorrect lyrics.

Singers

• Ahmet Kaya
• Barış Manço
• Büyük Ev Ablukada
• Can Bonomo
• Candan Erçetin
• Cem Adrian
• Cem Karaca
• Ceza
• Duman
• Ezhel
• Feridun Düzağaç
• Kayahan
• Mabel Matiz
• Mazhar Fuat Özkan (MFÖ)
• Mor Ve Ötesi
• Mustafa Sandal
• Müslüm Gürses
• Nazan Öncel
• Orhan Gencebay
• Pinhani
• Sagopa Kajmer
• Serdar Ortaç
• Sertab Erener
• Sezen Aksu
• Tarkan
• Teoman
• Yaşar
• Yeni Türkü
• Zeki Müren
• Zülfü Livaneli
• Şebnem Ferah

NLP TASKS

• Better tokenization pipeline with Turkish Stemmer and Lemmatizer
• Bigrams, Trigrams
• TF-IDF
• Topic Modelling with NMF or LDA
• Naive Bayes Classifier
• Song classification by Lyrics
• Sentiment Analysis
• NER
• Lyrics Generation with LSTM

Loading…