Skip to main content

Subject to change as the Fall 2017 term progresses. ★ marks the required reading.

Lecture Topic Readings
August 22, 2017 Introduction
  • Introduction of social media and natural language processing research
  • Overview of the course
August 29, 2017 Twitter and Twitter API Tutorial [Quiz1 due]
  • Brief history of Twitter
  • Key features of Twitter
  • Hands-on instructions on obtaining Twitter data via APIs
September 5, 2017 no class [HW1 due]
  • instructor travels for conference (EMNLP and W-NUT)
September 12, 2017 Guest Lecture by Pravar Mahajan [Read1 due]
  • Python Numpy Tutorial
September 15, 2017 drop deadline
September 19, 2016 Language Identification and Naïve Bayes [Read2 due]
  • Domain/Genre Difference
  • Language Identification
  • Supervised Learning and Classification
  • Naïve Bayes Algorithm + feature selection (Information Gain)
September 26, 2016 Guest Lecture by Alan Ritter [Read3 due]
October 3, 2016 Paraphrase Data Sources [Read4 & Quiz2 due]
  • Overview of paraphrase research
  • WordNet, DIRT, MRPC (Microsoft Research Paraphrase Corpus), PPDB (Paraphrase Database), etc.
Octorber 10, 2017 Guest Lecture by Jeniya Tabassum [Read5 due]
Octorber 17, 2016 Paraphrase Identification and Linear Regression
  • Linear Regression
  • Cost Function
  • Gradient Descent
October 24, 2017 Tokenization, Normalization, POS/NE Tagging for Twitter
  • Tokenization, Emoticons
  • Noisy Text Normalization
  • Part-of-speech tagging
  • Named entity recognition
October 27, 2017 withdraw deadline
Octorber 31, 2016 Paraphrase Identification and Logisitc Regression [Read6 & Quiz3 due Oct 30]
  • Logistic Regression
  • Decision Boundary
November 7, 2017 Brainstorm Research Ideas [Slides due on Nov 6]
  • 2:20-2:50pm 1-minute student presentations
  • 2:50-3:10pm group discussions (round 1)
  • 3:10-3:30pm group discussions (round 2)
  • 3:30-3:40pm break and wrapping up
  • 3:40-4:10pm 5-minute group presentations
November 14, 2016 Vector Semantics [Quiz4 & Read7 due - last Quiz]
  • Unsupervised Learning
  • Class-based Clustering: Brown Clusters
  • Soft Clustering: Singular Value Decomposition (SVD)
  • Neural Word Embeddings: Word2vec (CBOW and Skip-gram)
November 17, 2016 (3:00-4:00pm) Distinghuished Guest Speaker Talk by Dr. Lyle Ungar
  • Dreese Labs 480
  • Measuring Psychological Traits using Social Media
November 20, 2017 .[HW2-part1 due]
November 28, 2016 (12:45-2:05pm) Clippers Talk by Wuwei Lan [Read8 due]
  • Jennings Hall 140
  • Learning Large-scale Paraphrases Continuously from Twitter
December 4, 2017 .[HW2-part2 due]
September 26, 2016 (Monday) Invited Talk by Jiwei Li (Stanford)
  • 4:00 -- 5:00pm, Dreese 480
  • Teaching a Machine to Converse
Octorber 12, 2016 Twitter Paraphrase and Latent Variable Model
  • Evaluation: Precision, Recall, F-measure
  • Multiple Instance Learning
  • Conditional Log-linear Model with Latant Variables
  • Crowdsourcing
Octorber 18, 2016 (Tuesday) Invited Talk by Margaret Mitchell (Microsoft Research)
  • 4:00 -- 5:00pm, Dreese 480
  • From Naming Concrete Objects to Sharing Abstract Thought: Vision-to-Language Begins to Grow Up
Octorber 19, 2016 Twitter Paraphrase and Latent Variable Model (cont')
Octorber 26, 2016 Automatic Summarization for Twitter and the PageRank Algorithm
  • SumBasic algorithm
  • PageRank algorithm
  • Graph visualization
  • Event-based summarization system
November 30, 2016 Deep Learning for NLP
  • Neural Network Basicsg: Neuron, Activation Function, Non-linearity, Learning
  • Recurrent Neural Network
  • Long Short-Term Memory Networks
  • Neural Machine Translation
  • Neural Conversation Generation
December 7, 2016 Review and Q&A
  • I will go over some course material (lecture slides or homework) where students have questions or where we didn't have enough time to go into details previously.
  • I will also have extended office hours that day for 1-on-1 meetings on research project review, Q&A, etc.