Topics in Computer Science

General Information

Instructor

Jeanna Matthews

Office: Science Center 389

Phone: 268-6288

E-mail: jnm@clarkson.edu

Homepage: http://www.clarkson.edu/~jnm

Students

Saied Alshahrani

Overview

This is a directed study that is meant to approximate a regular 600 level research-based course. Students will do that following:

• Read 2-3 papers per week and document a schedule with links to all papers here on this page.
  Note: papers discussions starting in week 3 and there should be some make up work for weeks 1 and 2 scheduled.
• For each paper produce a writeup answering the following questions:
  1. What is the primary lesson(s) you took away from this paper (avoid abstract level summary)?
  2. Could I have done this work if I had the idea why or why not?
  3. Is there anything I could do to repeat or validate? (Make notes of materials available from the paper- data sets, source code)
  4. What is my best idea for follow on work that I could personally do?
  5. What is my best idea for follow on work that I'd like to see the authors or others do?
  6. Any logistical experimental lessons I learned (this is a little different than #1 here you are looking for techniques)?
  7. How does this compare to the other papers we read? Most similar? How different? Other comparisons?
  8. What is your biggest criticism of the paper?
  9. List 3 cited references or terms/concepts that you would be most interested in reading/learning more about.
• Assemble an individial portfolio of writeups for each paper.
• Propose midterm research deliverables and schedule a presentation and submit written materials.
• Propose final research deliverables and schedule a presentation and submit written materials.
• Students will attend weekly research meetings (Tues 6 pm).

Objectives

• Expose students to current and classical literature
• Give students an understanding of what it means to do research in computer science and specifically in computer security
• Teach students to critically evaluate research papers
• Tackle some challenging projects and write paper suitable for publication

Outcomes

• Students will learn to read and critique research papers
• Students will summarize and lead a discussion of research papers
• Students will be familiar with classic research literature
• Students will know where to look for and where to publish papers on computer security topics
• Students will make substantial contributions to a large project that can be submitted for publication

Exams

There will be a final research presentation and document rather than a final exam.

Tentative Grading Scheme

• 30% Portfolio of Writeups
• 30% Midterm Research Deliverables
• 40% Final Research Deliverables

Some Conferences To Follow

FAT *

AIES (AI, Ethics and Society)

FATES

NeurIPS

Schedule

Date	Main Paper/Activity	Other Notes/Related Papers
October 1	Mid-semester Research Writeup
December 1	End-semester Research Presentation
December 12	Final End-semester Research Writeup
December 12	Portfolio of Paper Writeups Due

Title	Authors	Venue and Year	Reading Date	Write-Ups
Generating Natural Language Adversarial Examples	Alzantot et al. (2018)	EMNLP 2018	09/12/2022
Black-box Generation of Adversarial Text Sequences to Evade Deep Learning Classifiers	Gao et al. (2018)	IEEE Symposiumon on Security and Privacy 2018	09/15/2022
HotFlip: White-Box Adversarial Examples for Text Classification	Ebrahimi et al. (2018)	ACL 2018	09/19/2022
Adversarial Examples for Natural Language Classification Problems	Kuleshov et al. (2018)	Rejected @ ICLR 2018 but good paper to read	09/21/2022
Pathologies of Neural Models Make Interpretation Difficult	Feng et al. (2018)	EMNLP 2018	09/26/2022
***** Progress Report of September 2022 *****			09/30/2022
Certified Robustness to Adversarial Word Substitution	Jia et al. (2019)	EMNLP 2019	10/03/2022
Toward Mitigating Adversarial Texts	Alshemali and Kalita (2019)	International Journal of Computer Applications 2019	10/05/2022
Combating Adversarial Misspellings with Robust Word Recognition	Pruthi et al. (2019)	ACL 2019	10/10/2022
TextBugger: Generating Adversarial Text Against Real-world Applications	Li et al. (2019)	NDSS 2019	10/12/2022
Adversarial Examples in Arabic (Supplementary Material)	Alshemali and Kalita (2019)	CSCI 2019	10/17/2022
Generating Natural Language Adversarial Examples through Probability Weighted Word Saliency	Ren et al. (2019)	ACL 2019	10/19/2022
BAE: BERT-based Adversarial Examples for Text Classification	Garg and Ramakrishnan (2020)	EMNLP 2020	10/24/2022
Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with Adversarial Examples	Cheng et al. (2020)	AAAI 2020	10/26/2022
***** Progress Report of October 2022 *****			10/31/2022
Is Bert Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment	Jin et al. (2020)	AAAI 2020	11/02/2022
TextAttack: A Framework for Adversarial Attacks, Data Augmentation, and Adversarial Training in NLP	Morris et al. (2020)	EMNLP 2020	11/07/2022
BERT-ATTACK: Adversarial Attack Against BERT Using BERT	Li et al. (2020)	EMNLP 2020	11/09/2022
Beyond Accuracy: Behavioral Testing of NLP models with CheckList	Ribeiro et al. (2020)	ACL 2020	11/14/2022
Word-level Textual Adversarial Attacking as Combinatorial Optimization	Zang et al. (2020)	ACL 2020	11/17/2022
Character-level Adversarial Examples in Arabic	Alshemali and Kalita (2021)	ICMLA 2021	11/20/2022
Contextualized Perturbation for Textual Adversarial Attack	Li et al. (2021)	NAACL 2021	11/24/2022
***** Progress Report of November 2022 *****			11/27/2022
##### End-semester Research Presentation ##### Title: Learning From Arabic Corpora But Not Always From Arabic Speakers: A Case Study of the Arabic Wikipedia Editions			12/07/2022
##### Final End-semester Research Writeup ##### Title: Learning From Arabic Corpora But Not Always From Arabic Speakers: A Case Study of the Arabic Wikipedia Editions Accepted at the Seventh Arabic Natural Language Processing Workshop (WANLP 2022 ), co-located with the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022), December 7-8, 2022.			12/08/2022
##### End-semester Project Presentation ##### Title: Aladdin Genies: Arabic Adversarial Text Normalization Attacks			12/12/2022