Assignment - 5 Big Data
Assignment - 5 Big Data
Assignment - 5 Big Data
Nodes:
1. Users: In online social networks, nodes often represent users who are connected
through friendships, follows, or interactions.
2. Entities: Nodes can also represent entities such as organizations, businesses, or
groups that interact within a social context.
Edges:
Social networks can be classified into different types based on the nature of
relationships or interactions between nodes. Some common types include:
1. Users: The individuals for whom recommendations are generated. Users may
provide explicit feedback (ratings, likes, dislikes) or implicit feedback (browsing
history, purchase history) that informs the recommendation process.
2. Items: The products, services, or content that are recommended to users. Items
can vary widely depending on the application domain, including movies, books,
products, music tracks, news articles, or social connections.
3. Feedback Data: Data collected from users' interactions with items, such as
ratings, reviews, clicks, purchases, or social interactions. This data is used to
train recommendation algorithms and personalize recommendations for users.
4. Recommendation Algorithms: The algorithms or methods used to generate
recommendations based on user preferences and item characteristics. These
algorithms can be categorized into different types based on their approach,
including collaborative filtering, content-based filtering, hybrid methods, and
matrix factorization techniques.
Evaluation Metrics:
Recommender systems are evaluated using various metrics to assess the quality and
effectiveness of the recommendations, such as precision, recall, accuracy, coverage,
and diversity. These metrics measure aspects like the relevance of recommended items,
the proportion of relevant items retrieved, the overall accuracy of predictions, the
percentage of items recommended, and the variety of recommended items.