Requirements

Let's refine the LinkedIn connection suggestion feature. Instead of a naive "mutual friends" approach, we want to build a system that provides more intelligent and relevant connection suggestions to users. This means taking into account factors like: * **Profile Similarity**: How similar are the job titles, skills, education, and location of two users? * **Network Proximity**: How many "degrees of separation" are there between two users? Are they directly connected, connected through a mutual connection, or further apart? * **Shared Group Memberships**: Are two users members of the same LinkedIn groups? This indicates shared interests. * **Common Company**: Did the users work at the same company at any point in time? The system should be designed to be extensible, allowing us to easily incorporate new factors in the future. It should also handle a large number of users and connections efficiently. Since connection suggestions are not critical real-time operations, we can allow for eventual consistency. # Requirements ## Functional Requirements 1. **Suggest Connections:** For a given user, the system should generate a list of suggested connections, ranked by relevance. 2. **Profile Similarity Calculation:** The system should calculate a similarity score between two user profiles based on job title, skills, education, and location. Implement Jaccard Index for skills matching. 3. **Network Proximity Calculation:** The system should determine the degree of separation between two users in the LinkedIn network. Consider a maximum degree of separation (e.g., 3) beyond which connections are not suggested. 4. **Shared Group Membership Calculation:** The system should identify users who are members of the same LinkedIn groups. 5. **Common Company Calculation**: The system should identify users who have worked at the same company. 6. **Relevance Ranking:** The system should combine the different factors (profile similarity, network proximity, shared group memberships, common company) into a single relevance score and rank the suggestions accordingly. Allow configurable weights for each factor. 7. **Exclusion List:** The system should exclude already-connected users and users who have explicitly declined a connection request from the suggested list. 8. **Scalability**: The system should be able to scale to handle a large number of users and connections. 9. **Configuration**: Weights for different factors influencing the connection suggestion should be configurable. ## Non-Functional Requirements 1. **Extensibility**: The system should be designed to easily accommodate new factors in the connection suggestion algorithm without requiring major code changes. (Open/Closed Principle). 2. **Modularity**: The system should be composed of loosely coupled modules, each responsible for a specific aspect of the connection suggestion process. 3. **Testability**: Each module should be easily testable in isolation. 4. **Thread Safety**: The system should be thread-safe, particularly when accessing and updating user profile and connection data. The connection suggestion generation process can be performed concurrently for different users. Read operations must be highly concurrent. Write operations must be carefully synchronized or use immutable data structures to avoid race conditions. 5. **Efficiency**: The connection suggestion algorithm should be efficient enough to generate suggestions in a reasonable amount of time (e.g., a few seconds). Use appropriate data structures and algorithms to optimize performance. 6. **Maintainability**: Code should be well-documented and easy to understand. 7. **SOLID Principles**: Adherence to SOLID principles is a must for clean, maintainable, and scalable code. ## Core Entities 1. **User**: Represents a LinkedIn user, including profile information (job title, skills, education, location), connections, and group memberships. 2. **Profile**: Encapsulates the profile information of a user (job title, skills, education, location). 3. **Connection**: Represents a connection between two users on LinkedIn. 4. **Group**: Represents a LinkedIn group. 5. **ConnectionSuggester**: The main class responsible for generating connection suggestions for a given user. 6. **SimilarityCalculator**: An interface for classes that calculate the similarity between two users based on different factors. 7. **NetworkProximityCalculator**: Calculates the network proximity between two users. 8. **SharedGroupCalculator**: Identifies users who share group memberships. 9. **CommonCompanyCalculator**: Identifies users who have worked at the same company. 10. **RelevanceRanker**: Ranks the suggested connections based on their relevance scores. 11. **SuggestionFilter**: Filters the potential suggestions based on exclusion criteria.

Design Linkedin Suggest connections feature.

Requirements

Interview Simulation

Functional Requirements

💡 Interview Tip

Non-Functional Requirements

Core Entity Identification

Premium Content