User personalization
- 2. 4R’s of Personalization • Recognize • Know your customer profile including demographics, interests and geographies. • Remember • Know your customer history meaning what they buy & browse. • Recommend • Reach them with right content, offer and product recommendations based upon their actions, preferences and interests • Relevance • Deliver the personalization within the context of user location, time and within the context of digital experience.
- 3. Collaborative filtering • CF recommends relevant content to users with similar “taste”, based on their reviews / purchases etc. • Two Categories to capture user preferences • Explicit Rating • Implicit Rating
- 4. Nearest Neighbourhood • The standard method of Collaborative Filtering is known as Nearest Neighbourhood algorithm. • There are two types: • User-based Collaborative Filtering • Item-based Collaborative Filtering
- 5. User-based collaborative filtering • We have an n × m matrix of ratings, with user uᵢ, i = 1, ...n and item pⱼ, j=1, …m. • Now we want to predict the rating rᵢⱼ if target user i did not watch/rate an item j. • The process is to calculate the similarities between target user i and all other users. • Select the top X similar users, and take the weighted average of ratings from these X users with similarities as weights.
- 6. Item-based collaborative filtering • we say two items are similar when they received similar ratings from a same user. • Then, we will make prediction for a target user on an item by calculating weighted average of ratings on most X similar items from this user.
- 7. Matrix Factorization • A matrix Factorization is a way of reducing a matrix into its constituent parts. • Example: • LU Matrix Decomposition Technique • A = L x U • The factors L and U are triangular matrices. The factorization that comes from elimination is A = LU. • L is the lower triangle matrix and U is the upper triangle matrix.
- 8. Content Filtering • Content-based filtering uses item features to recommend other items similar to what the user likes, based on their previous actions or explicit feedback. • For example, it can be movie attributes such as genre, year, director, actor etc., or textual content of articles that can extracted by applying Natural Language Processing.
- 9. Hybrid Recommender • Hybrid recommender systems combine two or more recommendation strategies in different ways to benefit from their complementary advantages
- 10. RNN based recommender • Recurrent Neural Networks are powerful tools for modeling sequences. They are flexibly extensible and can incorporate various kinds of information including temporal order. These properties make them well suited for generating sequential recommendations.
- 11. RNN captures the sequential information present in the input data i.e. dependency between the words in the text while making predictions: the output (o1, o2, o3, o4) at each time step depends not only on the current word but also on the previous words.
- 12. Practical Implementation • TensorFlow Recommenders • Sci-kit Surprise Kit • AWS Personalize
- 15. Thanks