Covers: theory of Feature Extraction for Content Based RecSys

- Overview of algorithms to extract discriminative features using labelled data

Read section 4.3.4 of the textbook. This resource corresponds to Resource #3 in the shortlist, "Introduction to Content-Based Recommender Systems".

Charu Aggarwal

0 comment

Collaborators

- Objectives
- This list will teach how to represent items using TF-IDF method and two feature selection methods based on Mutual Information (MI) and Chi-square test.
- Potential Use Cases
- Content-based Recommender Systems are useful in
- Who is this for ?
- INTERMEDIATEThose who have gone through the "Introduction to Content Based Recommender Systems" Shortlist.

Click on each of the following **annotated items** to see details.

BOOK_CHAPTER 1. Supervised Feature Extraction and Weighing

- Overview of algorithms to extract discriminative features using labelled data

30 minutes

ARTICLE 2. Intuitive Guide to TF-IDF

- What is TF-IDF?
- Why is TF-IDF is good for item representation?

10 minutes

ARTICLE 3. Implementation of TF-IDF

- How can we use TF-IDF to recommend items?

30 minutes

ARTICLE 4. Video on TF-IDF

- Why use TF-IDF?
- How to calculate TF and IDF?

5 minutes

ARTICLE 5. Lecture Notes on TF-IDF and Content-Based RecSys

- How to calculate TF-IDF?
- How can the TF-IDF be improved?

15 minutes

ARTICLE 6. Short Video on Mutual Information

- What is mutual information?
- How to calculate mutual information?

10 minutes

ARTICLE 7. Mutual Information for Feature Selection

- How we solve for mutual information?

50 minutes

ARTICLE 8. Chi-square distribution

- What is Chi-square distribution?
- What is Chi-square test?

10 minutes

ARTICLE 9. Implementation of Chi-square Feature Selection

- How can we use Python to implement Chi-square feature selection?

40 minutes

0 comment