Questions tagged [recommender-system]
A recommendation engine tries to predict how much a user will enjoy certain goods (movies, books, songs, etc) and makes recommendations. They are often used by online vendors to suggest new purchases.
457 questions
0
votes
0
answers
8
views
Does it make sense using nDCG with implicit feedback?
I am comparing multiple offline recommender systems models on an implicit feedback dataset and reporting various metrics.
These models follow the same order with multiple metrics. The best model is ...
- 101
0
votes
0
answers
23
views
Detect paterns over time in multivariate dataset
I have a dataset representing the stock of a shop over several days. For each day, I have hourly inventories of the objects in the shop. Some products are sold, and others might temporarily disappear (...
0
votes
0
answers
19
views
Shallow better than wide deep learning embedding models?
I'm training a two-tower recommender embedding model where one tower represents users and another represents items. User and item embeddings should be close when a user clicked an item and far part ...
- 6,767
0
votes
0
answers
35
views
Top-N recommender system
Say an intermediary is using a two part recommender model that attempts to facilitate services between its clients and external vendors:
Model 1: Predict probability of vendor bidding on a given ...
3
votes
1
answer
217
views
What defines collaborative filtering? [closed]
What is the defining feature of collaborative filtering?
If you take two embedding vectors (one for a user and one for an item), you can do a dot-product and pass the result to a sigmoid function to ...
- 39
0
votes
0
answers
33
views
How can I integrate time in my Implicit Feedback dataset?
I'm working on a recommendation system based on Collaborative Filtering. Specifically, I've been looking at models such as NCF (Neural Collaborative Filtering) and SAR (Simple Algorithm for ...
- 1
0
votes
0
answers
15
views
How to develop shared bottom tower serving different tasks
I have two model classes both pyramid architecture.
Let's say first task is predicting user will buy something with architecture [feature_embedding_128, dense_1048, dense_512, dense_128, dense_1]
...
- 314
0
votes
0
answers
30
views
Recommender System for continuous predictors
I want to build a model that is able to predict the outcome of a user-client interaction.
I know that for categorical variables Factorization Machines are a good choice.
Imagine for example we are ...
- 133
0
votes
2
answers
77
views
The Impact of Vector Magnitudes in Recommendation Systems Matrix Factorization Models
I'm currently exploring latent factor models in recommendation systems, specifically focusing on the interaction between vector magnitudes and the angles between these vectors. While it's clear that ...
- 77
0
votes
0
answers
37
views
Statistics of algorithm AB testing on occurrences only
I have AB testing in place and I have to compare the data but I have no clue which statistical test to use.
I'll try to speak in dating app terms because I think it makes it easier to understand.
I ...
0
votes
0
answers
31
views
How to re-rank after candidate generation?
I am working on an app recommendation problem. I don't have any app features, but I have user features. I've tried different similarity based models and also using a multiclass classification model's ...
0
votes
0
answers
165
views
How to normalize data for weighted sum model
I'm building a simple weighted sum model for ranking.
$$
\text{Store Rank} = w_1 \cdot param_1 + w_2 \cdot param_2 \ldots + w_n \cdot param_n
$$
The problem here is that one of the parameters depends ...
- 101
1
vote
0
answers
20
views
RecSys model performance stalling at 47% AUC and F1-Score. Is the problem due to ratio of users to items in my dataset?
I'm having trouble with making my validation metrics go down for the binary_crossentropy and go up for the F1-score and AUC. I've tried tuning my hyper parameters such as the number of latent features ...
0
votes
0
answers
48
views
Optimization metrics for a single-item recommendation system
I'm working on a recommendation system that's a bit different from ones I've built before. In particular, this system shows only the top item to the user, and the user can either click on it, dismiss ...
- 113
1
vote
0
answers
33
views
How to choose k for MAP@K?
Scenario: We want to evaluate our recommender system, which recommends items to potential customers when visiting a product detail page.
Here are actual relevant items:
...
- 1,027
6
votes
2
answers
276
views
Recommend similar users (instead of items) with collaborative filtering
I'm learning about collaborative filtering, and all the resources I've found so far describes how to find items a user might like. However, how would I find the most similar users to a specific user, ...
- 167
0
votes
1
answer
347
views
Measuring perplexity over a limited domain in an LLM
Are there papers/a literature on measuring perplexity in using a Large Language Model such as ChatGPT/Flan over a limited domain?
I want to prompt an LLM to do movie recommendations/next job ...
- 113
0
votes
0
answers
133
views
Alternating Least Squares for Matrix Factorization with biases
I am attempting to use ALS for matrix factorization, using a loss function that includes user and item biases $c_i$, $d_j$,
$$
L = \sum_{(i,j) \in S} (r_{ij} - x_i\cdot y_j^\top - c_i - d_j)^2 + \...
2
votes
1
answer
277
views
Recommender system - for single user
I am building a recommender system, in which for one system there will be only one user. So we cannot use something like user-user data. Recommendation is an item which contains 10-15 attributes ie ...
- 33
1
vote
0
answers
14
views
Recommendation Model (students Recommended to a Job opening)
I am currently working on a Recommendation Model. Which takes a Job Openings and Computer Science students and recommend students to the job. Which Model should I use or any suggestions you can give?
...
- 11
1
vote
1
answer
2k
views
Is it ok to have low validation loss from the first epoch?
I'm trying to implement Neural Collaborative Filtering recommender system using Keras, the dataset I'm using is movielens-small. Whatever I do to hyperparameters or network, when training, the ...
- 11
0
votes
0
answers
17
views
What are the ways to use a LIST of features which are DYNAMIC (contents) in nature?
Any features which is represented as a list of 0 or more elements is what I call a Dynamic feature.
Let us suppose an example where there are 10 Million movies and ...
- 244
3
votes
1
answer
138
views
Evaluation of a recommendation system. How can I do that?
I'm learning to make a book recommendation system but I am facing some difficulties to evaluate the model. I chose the collaborative filtering item based strategy. The dataset is a matrix (book, user) ...
2
votes
1
answer
466
views
When adding batch norm layer do I need to added to all layers in DNN?
While developing deepfm model network I want to add batch norm layer because model seems to suffer from vanishing gradient. There are embedding layers, 2 layers a in deep model part and one dense ...
- 261
0
votes
0
answers
123
views
When the accuracy curve is U-shaped
I am currently working on MLP-based recommendation system.
During training, the model updates based on BCE loss function with train set, then shows the hit rate (rate of how ground truth item is in ...
- 97
6
votes
2
answers
969
views
ALS vs SGD in parallelization
So given the standard objective in matrix factorization for collaborative filtering of minimizing:
$$
L = \sum_{u,i \in S} (r_{ui}-q_i^Tp_u)^2 + \lambda(\sum_i||q_i^2||+\sum_u||p_u^2||)
$$
, where $r_{...
- 343
1
vote
1
answer
262
views
Recommender system with unique products
Task definition
I've been tasked to build a recommender system and I have to admit I'm a beginner in this field. Entities in my system are buyers and unique products that could be bought. You could ...
- 113
0
votes
1
answer
117
views
Predictions based on data with correlations within and between multiple sets of time series
I'm looking for a model to learn relationships within and between a set of partially observed time series in order to generate predictions for any timepoint in any of the set of time series.
More ...
- 1
0
votes
0
answers
101
views
How to evaluate the performance of recommender systems without having labeled data
I have a huge citation graph of research papers and datasets. So, there is an edge among two items when one of them cites another. So far I've used Node2Vec for creating a dataset recommender system ...
- 1
1
vote
0
answers
31
views
Content Based Recommendation System
how will a content-based recommendation system recommend after a sudden change in taste of user??
what will the system recommend to new users without any data about them available in content-based ...
- 11
3
votes
0
answers
47
views
asking humans to rank items
I have around 50 items and I need to ask human graders to rank them. Is there any good resource on how to design this crowd sourcing task? For example, it will be tiresome to ask humans to rank 50 ...
- 131
1
vote
0
answers
27
views
Recommender System based on item-item features evaluation method
The idea is to recommend books purely on the characteristics of it without any user's input (ratings) only getting top N books that are the most similar to a book.
I am implementing Euclidean distance ...
- 11
2
votes
0
answers
126
views
Transition matrix as a feature to feed machine learning algorithm
Currently, I am trying to replicate a paper to extend the research on the paper they create a transition matrix similar to this one:
my question is : how can I feed this information into my KNN or ...
- 21
3
votes
1
answer
336
views
Matrix Factorization and Overfitting
I recently came accross the algorithm of Matrix Factorization for a recommendations system.
One of the tutorials I followed can be found here.
According to it given the initial matrix $R$ and the ...
- 131
0
votes
0
answers
80
views
How to choose the best recommender system? What evaluation metrics to use?
I want to build a recommender system to suggest similar songs to continue a playlist (similar to what Spotify does by recommending similar songs at the end of a playlist).
I want to build two models: ...
- 1
0
votes
1
answer
68
views
Recommend items to complete a set
I'm trying to predict/recommend items to add to an incomplete basket. In this case, the basket is a project — e.g., you detect a customer trying to fix the AC in their car, hence you suggest things ...
- 109
0
votes
1
answer
424
views
weighted average in collaborative filtering
I studied on collaborative filtering recently and found that for ratings at last most methods applied were weighted average, no matter what they had proposed, similarity, time, etc.
Thus I am ...
- 31
1
vote
1
answer
52
views
What is the general right call to make from a marginal difference in A/B test results in recommender system?
This was one of the business-related questions from my technical interview last week for a data science position in a recommender system team at a search engine company focusing on advertisement ...
- 221
1
vote
0
answers
172
views
Statistical test for recommendation algorithms
I've developed 2 different image based recommendation systems for an e-commerce company and now it's time to evaluate them using some statistical tests. Both algorithms work in a way in which if fed ...
- 363
3
votes
1
answer
72
views
Sample size calculation for precision at $k$: a/b-testing a recommender system
I would like to conduct an online experiment to compare two different versions of a recommender system. The system returns a list of $r$ ranked recommendations. I would like to evaluate the ...
- 536
0
votes
0
answers
555
views
Funk SVD for binary data - product like or dislike
Assume the following situation:
you have a user-item sparse matrix. However, instead of the usual 1 to 5 rating scale, items can only receive a positive (1) or negative (-1) feedback.
Thus, the matrix ...
- 143
1
vote
0
answers
51
views
Content-based filtering recommendation system for a group
Is there any example of content-based filtering recommendation for a group of observations? For example, user can choose a group of 10 movies they like the most, and the recommendation engine will ...
- 21
0
votes
1
answer
245
views
Sequential recommendation: how to effective encoding output item?
Now I am learning about sequential recommendation - session based recommendation. I have understood that User-item interactions may be viewed as sequential action (first I clicked item A, then click ...
- 150
2
votes
1
answer
608
views
Is there one splitting strategy for both K-NN and Matrix Factorization recommender systems?
I am researching several different recommender systems, some of which are based on a user K-Nearest Neighbour algorithm and some of which are based on a matrix factorization algorithm. My dataset is ...
2
votes
1
answer
821
views
What's the difference between a recommender system and a decision tree?
We learned in Machine Learning that both of those techniques try to predict an output (whether person A likes a specific product or whether person A has a high default risk) based on data of other ...
- 111
0
votes
0
answers
38
views
How many variables for a recommender system?
I'm creating a survey for 60 people. I'd like to measure their music, tv-series and purchasing tastes.
The survey will show a number of songs. People will be asked to say if they like it (1) or not (0)...
0
votes
0
answers
492
views
Next best offer models in recommendation system
I'm trying to so some research on "Next Best Offer" or "Next Best Action" models used in recommendation systems. Searched on google but didn't find any article with detailed ...
- 331
0
votes
1
answer
80
views
Search, rank and recommend in large text datasets
Imagine you are Spotify and you have billions of songs. Assume that each of these songs are transcribed into text. How do you design your search and recommendation pipeline such that when somebody ...
- 307
0
votes
0
answers
108
views
Constructing a User Profile for Music Taste
My goal is to construct User profiles based on positive (and maybe also negative) interactions with songs. A User has the option to like a song. This would give me a list of likes for each user. With ...
- 31
0
votes
0
answers
80
views
Why does the RMSE stay the same regardless of the algorithm that I use?
I have a dataframe with users, items, and ratings that are either 0 or 1. There are more items than users, some users might rate lots of common items, and some not any common items at all. Here is a ...
