---
title: "Collaborative filtering concept"
subject: "Information System"
chapter: 7
key: "semester_8/information_system:collaborative-filtering-concept"
source: https://www.bctengineer.com/q/semester_8/information_system/collaborative-filtering-concept
site: BCT Engineering
---

# Collaborative filtering concept

Information System · Chapter 7: Web based information system and navigation

- Asked in: 2071 Bhadra, 5 marks
- Video: [Collaborative Filtering | Machine Learning | Recomendar Recommendation  System by Dr. Mahesh Huddar](https://www.youtube.com/watch?v=W30s-EoP3d0) — Mahesh Huddar
- Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14
- Page: https://www.bctengineer.com/q/semester_8/information_system/collaborative-filtering-concept
- Whole chapter: https://www.bctengineer.com/md/semester_8/information_system/chapter-7

## Question

What do you mean by a collaborative filtering? Discuss the concept of collaborative filtering with suitable practical example.

## Short answer

Collaborative filtering recommends items by collecting preferences from similar users without needing item content. It builds a user-item matrix, computes neighbor similarity, and predicts missing ratings. For example, if Ram and Sita both liked films A and B, and Sita liked C, then C is recommended to Ram.

## Answer

#### Meaning
**Collaborative filtering (CF)** is a recommendation method that suggests items to a user based on the opinions or behaviour of other people with similar interests. The idea is "people who agreed in the past will agree in the future". It does not need to know the content of the items.

#### Concept
- Data is stored as a **user-item matrix** of ratings or actions (most entries are empty).
- The system finds users (or items) that are **similar** using a measure such as cosine similarity or Pearson correlation.
- It predicts the missing rating as a weighted average of ratings by similar users, then recommends the highest predicted items.

#### Types
- **User-based CF:** find users similar to the target user.
- **Item-based CF:** find items similar to those the user liked (used by Amazon).
- **Model-based CF:** learn a model, such as matrix factorization.

#### Practical example
| | Film A | Film B | Film C |
|---|---|---|---|
| Ram | 5 | 4 | ? |
| Sita | 5 | 4 | 5 |
| Hari | 1 | 2 | 1 |

Ram's ratings are close to Sita's and far from Hari's, so Sita is his neighbour. Sita rated C = 5, so the system predicts a high rating for Ram and recommends film C.

#### Advantages and problems
- Advantages: works for any item type, finds unexpected items, improves with more data.
- Problems: cold start, sparsity, scalability, popularity bias and privacy.

## Check yourself

The answer's key points as multiple-choice questions, with the correct option and why.

1. What is the idea behind collaborative filtering?
   A. Items similar in content are always liked
   B. People who agreed in the past will agree in the future
   C. Everyone likes popular items
   D. Users never change taste
   Answer: B. CF uses opinions of similar people and needs no item content. (Key point: Idea)

2. How is a missing rating predicted?
   A. As a weighted average of ratings by similar users
   B. As a random number
   C. As the lowest rating given
   D. As the number of users
   Answer: A. Similar neighbours' ratings are combined. (Key point: Prediction)

3. In the table, why is Sita Ram's neighbour and not Hari?
   A. Sita has a longer name
   B. Hari rated more films
   C. Sita rated fewer films
   D. Ram's ratings are close to Sita's and far from Hari's
   Answer: D. Ram and Sita give 5, 4 to films A and B; Hari gives 1, 2. (Key point: Example)

4. Which type finds items similar to those a user liked (used by Amazon)?
   A. User-based CF
   B. Content-free CF
   C. Item-based CF
   D. Demographic CF
   Answer: C. User-based finds similar users; item-based finds similar items. (Key point: Types)

## Related in Information System

- [Cold start in collaborative filtering](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Acold-start-collaborative-filtering) — Information System, chapter 7, asked 2×
- [Short note: collaborative filtering](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Ashort-note-collaborative-filtering) — Information System, chapter 7

## Related in other subjects

- [Ideal and RC filtering of white noise](https://www.bctengineer.com/question.md?key=semester_6%2Fcommunication_systems%3Arc-filtering-white-noise) — Communication Systems, chapter 9
- [Content filtering vs packet filtering](https://www.bctengineer.com/question.md?key=semester_8%2Finternet_and_intranet%3Acontent-filtering-packet-filtering) — Internet and Intranet, chapter 5
- [Practical engineering profession; organizations](https://www.bctengineer.com/question.md?key=semester_8%2Fengineering_professional_practice%3Apractical-profession-organizations) — Engineering Professional Practice, chapter 3
- [Ideal and practical sources, conversion](https://www.bctengineer.com/question.md?key=semester_1%2Fbasic_electrical_engineering%3Aideal-practical-sources-conversion) — Basic Electrical Engineering, chapter 2