Machine Learning Applications
Topics
- Computer vision
- Natural language processing
- Text and data mining
- Recommendation
- Applications of Machine Learning in Libraries and Information Organizations
Weekly Discussion: Reflecting on AI Techniques in Information Contexts
Please submit all graded work via Canvas. Participation requirements and grading details are provided in Canvas.
Over the past few weeks, we examined different approaches to AI, including symbolic and subsymbolic paradigms, as well as specific machine learning applications such as computer vision, natural language processing, text and data mining, and recommendation systems.
This discussion asks you to reflect on these techniques from your own perspective.
Step 1: Choose one AI approach or task
Select one technique discussed in this module.
You may choose:
- Symbolic AI (rule-based systems, knowledge graphs, ontologies)
- Subsymbolic AI (machine learning models)
- Computer vision
- Natural language processing
- Text and data mining
- Recommendation systems
- Supervised, unsupervised, or reinforcement learning
Briefly describe:
- What this technique does
- What kind of problem it is designed to address
- In what kinds of information contexts it might be used
Step 2: Your evaluation
You may consider:
- Which aspects were easiest for you to understand? Why?
- Which aspects felt abstract or difficult?
- Does the technique seem promising for libraries or information organizations?
- Does it raise concerns for you?
Step 3: Practical and ethical considerations
Consider the real-world implications of applying this technique.
You may address:
- What kinds of data would it require?
- What kinds of bias might appear?
- Would human oversight be necessary?
- Would interpretability matter in this case?
- Are there situations where a symbolic or rule-based approach might be preferable?
Step 4: Looking forward
Finally, briefly respond to one of the following:
- Which AI approach do you think will have the most impact in information organizations over the next five years?
- Which approach do you think is most misunderstood?
- Which technique do you think poses the greatest governance challenge?
Expected outcome
Your post should demonstrate:
- Understanding of at least one AI paradigm or task
- Ability to connect technical mechanisms to practical contexts
- Awareness of limitations, trade-offs, and governance considerations
Reading
- Martin Frické, Artificial Intelligence: Foundations of Computational Agents, Chapters 3, 8, 9
- Why Computer Vision Is a Hard Problem for AI
- AI unlocks ancient Dead Sea Scrolls mystery
- Digitizing the paper of record: Archiving digital newspapers at the New York Times
- Managing the risks of inevitably biased visual artificial intelligence systems
- Computer Vision Tagging the Metropolitan Museum of Art's Collection: A Comparison of Three Systems
- Unmasking the bias in facial recognition algorithms
- Simplilearn: Natural Language Processing In 10 Minutes
- GeeksforGeeks: Natural Language Processing (NLP) Tutorial
- GeeksforGeeks: Named Entity Recognition
- GeeksforGeeks: What is Sentiment Analysis?
- HathiTrust - Text and data mining resources
- HathiTrust - How to Use HathiTrust Data Resources
- Text & Data Mining resources @ UK Library
- Topic Modeling and Digital Humanities
- ProQuest's TDM Studio
- Google - Recommendation systems overview