Artificial Intelligence (AI)

Here’s some useful resources on Artificial Intelligence, categorized by topic.

Natural Language Processing Fundamentals

• Word embeddings
  • fastText by Facebook Research (GitHub, paper).
  • GloVe
  • Word2Vec
  • Polyglot by Rami Al-Rfou.

Machine Learning Fundamentals

• Visualization of optimizer algoritms & which optimizer to use by Sebastian Ruder. TL;DR: “adaptive learning-rate methods, i.e. Adagrad, Adadelta, RMSprop, and Adam are most suitable and provide the best convergence for these scenarios. (don’t use vanilla SGD)”

Machine Learning for Natural Language Processing

• Oxford Course on Deep Learning for Natural Language Processing. [GitHub] [YouTube playlist (unofficial)]
  Very good free course with exercises (Python) covering from the basics of both machine learning and natural language processing, then how to apply it.

Natural Language Processing (NLP) Tools & Software (mostly in Python)

• NLTK. Natural Language Toolkit. tokenizer, POS tagging, identify named entities, … Book: Natural Language Processing with Python.
• AFNER (Named Entity Recognition).
• Pretrained word vectors for 294 languages (from Wikipedia) by Facebook Research.
• Snowball (Word Stemmer).

Text Corpus in Indonesian Language

• Wikipedia Dumps. For Indonesian, use idwiki. My preferred format is “Articles, templates, media/file descriptions, and primary meta-pages, in multiple bz2 streams, 100 pages per stream”, so you can process in smaller chunks.
• Kumpulan Korpus Bahasa Indonesia by Herry Sujaini (2013)
• Are there any Indonesian corpora available online? – Quora
• freq-dist-id by Jim Geovedi/Ardwort. Repositori data yang digunakan dalam makalah Perbandingan distribusi frekuensi kata bahasa Indonesia di Kompas, Wikipedia, Twitter, dan Kaskus.
  • Lanin, I., Geovedi, J., & Soegijoko. W. (2013). Perbandingan distribusi frekuensi kata bahasa Indonesia di Kompas, Wikipedia, Twitter, dan Kaskus. In Proceedings of Konferensi Linguistik Tahunan Atma Jaya Kesebelas (KOLITA11) (pp. 249-252).
• Indonesian Wordlist by Jim Geovedi. Only useful for password cracking, not linguistic research.
• Tagged text corpus with Named Entity Recognition by Yohanes Gultom. (named entity recognition)
• LAPOR.go.id (Bachelor thesis paper by Chyntia Megawati)
• Korpus Plagiarisma Indonesia (2016) by Felik Junvianto.
• Indonesian Manually POS Tagged Corpus (2016) by Fam Rashel.
• SEAlang Library Indonesian (not downloadable).
• Artikel: Korpus Daring Bahasa Indonesia by Wahyu Adi Putra Ginting (2010).

Natural Language Processing for Indonesian Language

• Pelabelan Corpus Bahasa Indonesia / Indonesian POS Tagging – INACL
• Kumpulan thesis, paper, dan artikel tentang NLP (Natural Language Processing) Bahasa Indonesia (2014) by Andy Librian (maker of Sastrawi)
• Word stemmer
  • Sastrawi (2017). Stemmer PHP for Indonesian language. Python version by Hanif Amal Robbani (pip, slower?).
  • Porter Stemmer for Indonesian language (2015) by Adinda Praditya. In Ruby. (paper)
• Pretrained word vectors in Indonesian (from Wikipedia) by Facebook Research.