Doğal Dil İşleme

Text Representation

1. Bag-of-Words and TF-IDFDocuments as sparse term vectors weighted by how distinctive each term is.1–2 sa

Tasks

1. Text ClassificationSentiment, topic, and intent: linear models over text features, then fine-tuned encoders.1–2 sa
2. Sequence Labeling: Tagging and Named EntitiesAssigning a label to every token (part of speech, BIO entity spans), with context and label transitions.1–2 sa
3. Machine TranslationThe founding sequence-to-sequence task: alignment, attention, and automatic evaluation with BLEU and chrF.1–2 sa
4. Multilingual NLPOne model, many languages: shared vocabularies, cross-lingual transfer, and the costs borne by low-resource languages.1–2 sa