A message from our CEO about the future of Stack Overflow and Stack Exchange. Read now.

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.

195 questions with no upvoted or accepted answers
Filter by
Sorted by
Tagged with
6
votes
1answer
2k views

Proper way to use NDCG@k score for recommendations

Currently I am building a recommender system and using ranking metrics to verify its performance. I am using the NDCG@k score. Today I was experimenting and I realized that I might be calculating the ...
5
votes
0answers
2k views

How does XGBoost/lightGBM evaluate ndcg metric for ranking

I am currently running tests between XGBoost/lightGBM for their ability to rank items. I am reproducing the benchmarks presented here: https://github.com/guolinke/boosting_tree_benchmarks. I have ...
4
votes
1answer
184 views

How to Deal with Categorical Variables that Allow Selection of Multiple Values per Observation?

Say you are dealing with a movie database that has movies and their genres. Genre is a categorical variable but each movie can belong to more than one genre. For example, Movie A may be Comedy and ...
4
votes
0answers
45 views

Who will follow who based on tags?

Suppose users in a system like a social network are described by a number of tags. The number of tags can be assumed to be less than 10. Example John: funny musician geek professor Peter: skinny ...
4
votes
0answers
639 views

Stratified Cross-Validation with Collaborative Filtering

My dataset consists of binary preferences ($0$ or $1$) given by users on items like this: User-ID | Item-ID | Preference If a user has not given a preference to an item, then it is not in the ...
4
votes
0answers
76 views

Predicting user selections based on similar user

Lets say you give a set of users a set of polls, or give them a choice of foods to eat, or let them listen to a group of songs (guess like pandora). So looking at the choices that all the users make ...
3
votes
2answers
4k views

How to train a LSTM model for a next basket recommendation problem?

I try to use a LSTM model for a problem of next basket recommendation. I would like to apply the same approach as this article in Python using Keras : A Dynamic Recurrent Model for Next Basket ...
3
votes
0answers
94 views

Two Source Collaborative Filtering

This question is based on the now classical work: Collaborative filtering with implicit feedback. I'm mainly interested in finding references for the question below. Suppose we are building a ...
3
votes
0answers
469 views

Approach to find similarity between two different types of matrices

I'm using the movielens dataset to give recommendations to a user based on genre of the movies. I have two matrices, one contains the genre the user likes. We are considering user3 for now I also ...
3
votes
0answers
744 views

What's a good evaluation method for recommendation systems?

I'm using the 100k MovieLense dataset to build a recommendation system in R, using the recommenderlabpackage. From what I understood, the ...
3
votes
0answers
92 views

How to model and predict a user's preference by their click-through history?

I am trying to model my users' food preferences so that I can recommend restaurants which he/she might be satisfied. The following is some sample data: ...
3
votes
0answers
83 views

Can collaborative filtering be cast as a classical regression problem?

Having the Netflix challenge in mind: collaborative filtering is typically presented as a matrix dimension reduction. My question is how does the problem relate to classical regression (supervised ...
3
votes
0answers
245 views

Does it make sense to use Lasso regression for recommendation systems?

I'm really intrigued by Lasso Regression, and it seems like a potential candidate for estimating “objective” ratings for items based on the users who watched them. The user data is sparse, and the ...
2
votes
0answers
78 views

Row similarity in matrix vs in different factorizations

Suppose an arbitrary $m \times n$ matrix $M$ and the factorizations: Arbitrary: $M = U_a V_a^T$, where $U_a$ is $m \times k$, $V_a$ is $n \times k$ ($k < m,n$), and $rank(U_a)=rank(V_a)=k$. SVD: $...
2
votes
0answers
21 views

How does the “age” feature work in video recommendation systems?

In the paper Deep Neural Networks for YouTube Recommendations, it mentions that the “example age” feature helps recommending fresh contents in Section 3.3. Many hours worth of videos are uploaded ...
2
votes
0answers
50 views

Deriving Multiplicative Update Rules for Regularized NMF

After reading the following CrossValidated post, I cannot derived the correct multiplicative rules for regularized NMF from this paper. They obtain the coefficients $|I_u|$ and $|U_i|$ in the ...
2
votes
0answers
119 views

Does anyone know the rank of the Netflix Prize dataset?

I'm looking into the Netflix Prize at the moment. We model the dataset as an $n \times m$ matrix, where $n$ is the number of users and $m$ is the number of movies. Does anyone know the rank of the ...
2
votes
0answers
102 views

Is there a well established algorithm to match two documents on a semantic level?

I have a set of documents from a wide variety of topics and I would like to retrieve the ones that are more similar to a new document provided. A search based on common words is not good enough, so ...
2
votes
0answers
24 views

Pairwise recommender system

How can I model the following situation? I am looking for high level recommendations. if say a man with certain attributes (income, age, etc ...) rates a car with a given attributes (color, ...
2
votes
0answers
212 views

Why PMI + SVD works for similarities arithmetics?

Recently Julia Silge blogged here and here, quoting blog entry by Chris Moody, who suggested that the similarities arithmetic in word2vec can be approximated by using PMI indexes followed by SVD ...
2
votes
0answers
231 views

Robust Principal Component Analysis (RPCA) for collaborative filtering

Consider the robust completion problem of a matrix $X$ where $X_{ij}$ are observed matrix entries for $(i, j)$ ∈ $Ω_{obs}$ i.e, $\min_{L,S}$ $\text{rank}(L) + \mu ||S||_{0} $ subject to $X_{ij} ...
2
votes
0answers
940 views

Recommender Systems with Implicit Feedback Data - Modelling and Updating

I have a very large dataset of user-item interactions. It tells me how many times a user bought, viewed or liked an item - all these actions are in binary form and I don't have any explicit ratings. ...
2
votes
0answers
388 views

SVD based recommender system C#

I'm trying to reduce the number of dimensions in my dataset for a movie recommender system using SVD. I'm using the 'MovieLens 1M Dataset' from GroupLens.org. I've used the MathNet library for ...
2
votes
0answers
42 views

Trying to build a model to suggest related questions

I am trying to build a model which can recommend/predict a followup question based on the previous question. Like "Did you attend the college?" a followup question may be "What was your major at ...
2
votes
0answers
109 views

Combine Ranking matrix of User based and Item based

I am doing 1 assignment on Recommendation in R using RecommenderLab package. The data I have is of different customers. Each customer have its Demographic information like Age, Gender, Education, ...
2
votes
0answers
43 views

Recommendation/personlization algorithm conflict

I'm trying to build a recommendation engine for an e-commerce site. By using the common recomendation approach, I'm assuming that each product I recommend has the same value, so all I need to do is ...
2
votes
0answers
593 views

How to get TopN recommendations using Matrix Factorization e.g. libFM

Being very new to recommender-systems and Matrix Factorization i was wondering how to get topN recommendations for a given user. So far my strategy is to create all possible User/Item combinations, ...
2
votes
1answer
103 views

constant terms in stochastic gradient descent: when to apply them and how much of the constant gradient component?

in a derivation for the gradient of a collaborative filtering system (similar to Probabilistic Matrix Factorization), I got to the following expression for the gradient of a latent vector $\mathbf{u}...
2
votes
0answers
948 views

Cosine similarity between two types of user profiles

I have a movie dataset containing preference (ratings) of users on movies and also attributes of movies (genre, cast, director). I created two types of profiles for users based on the movie attributes ...
2
votes
0answers
106 views

Item based collaborative filtering when items have been available for different lengths of time

I am attempting to use item based collaborative filtering for product recommendation. The matrix is all 1s and 0s based on whether or not a buyer purchased an item, and I am using cosine similarity to ...
2
votes
0answers
143 views

What does it mean when Recommender Systems are making recommendations based on “static” data as opposed to “dynamic” recommender systems?

Much of new trends in recommender systems are based on giving recommendations taking into account changing user/services/items preferences or requirements. They mention, traditional recommenders make ...
2
votes
0answers
821 views

The name data sparsity in different applications

I am recently surveying the techniques or algorithms which handle the data sparsity problems in various fields. And I find quite similar name "data sparsity" or "sparse data" is used including the ...
2
votes
0answers
223 views

clustering in Item-based collaborative filtering

I read about item-based collaborative filtering in the paper from Sawar et al. I want to apply clustering on items to find the most similar items and then apply the prediction. Is this a good ...
2
votes
0answers
154 views

Given $n$ ratings from 0 to 1, how to calculate a weighted “average” estimating the “true” rating?

I've read How Not to Sort By Average Rating regarding how to average binary positive/negative ratings in a way that takes the number of ratings into account. The author uses the "lower bound of Wilson ...
2
votes
0answers
62 views

How to build event recommender based only on events' descriptions

I built a simple app that grabs events using eventbrite and meetup.com APIs and displays it based on your zip code. Now I'm trying to build a simple event recommender that will use events that you ...
2
votes
0answers
798 views

Matrix factorization and gradient descent for recommender systems; user bias?

I've been reading about using Matrix Factorization techniques to do collaborative filtering. A popular thing to do seems to be to add user and item biases into the ratings prediction. What I don't ...
2
votes
0answers
366 views

Appropriate threshold to map a similarity value to an edge in a graph

In order to cluster users given a user-item binary matrix data, I am planning to first find user's similarity (Jaccard) and then use graph theory to isolate clusters (communities). I need to map the ...
2
votes
0answers
417 views

Matrix factorization vs random walk with restart for recommender systems

Suppose I want to handle "friend recommendation" problem on a large social network graph. I came across random-walk-with-restart as one technique used. I was thinking of using matrix factorization as ...
2
votes
0answers
271 views

Evaluation methods for personalized recommendation

I am currently trying to verify and evaluate a personalized recommender system I am working on, which seems like a huge task. Evaluating a static recommender system is rather easy and can be done with ...
2
votes
0answers
253 views

Nearest neighbor information for recommendation engines

My friend and I are working on a project on distributed datastructures. We were wondering how much is nearest neighbor information used in modern recommendation systems and whether it would be ...
1
vote
0answers
19 views

Ranking metric that takes into account length of result list?

I would like to evaluate a problem where the user select an option from a list of variable length. The task is to provide a ranked list so that the item lower in rank is the most relevant. If I use a ...
1
vote
1answer
24 views

Concept drift in in user interaction data

concept drift usually refers to the change in the relationship between input and output data over time. I do have dataset of users' activity in an e-commerce website. Let's say we have a sequence of ...
1
vote
0answers
8 views

Recommendation Engine and Text Analytics

I am looking for a dataset on which I can use Collaborative filtering and Content based filtering along with Text Mining. Could anybody please suggest , is there any dataset on which I can apply ...
1
vote
0answers
16 views

How to validate Item-Based Collaborative Filtering?

So, I made an books recommender system engine (with R) which based on item-item matrix. I've made the whole system fully works the output will give 5 recommender system. But, the question is how can I ...
1
vote
1answer
83 views

FindSimilar items in a complex dataset

Im a MachineLearning newbie, but I want to learn more about this interesting topic using a practical example, on which I would appreciate any theoretical and practical help: I have a database of "...
1
vote
0answers
14 views

Deep item-based recommender objective function

I'm trying to understand the following paper written by researchers at eBay that uses deep learning to overcome the problem of making recommendations when you mostly have one-of-a-kind items. A ...
1
vote
1answer
59 views

Expected Value of Naive Recommender System

Let $k, n \in \mathbb{N}$, with $k \leq n$. Let $A = (a_1, a_2, ..., a_n)$ be an unordered finite sequence of real numbers. Let $(B_1, B_2, ..., B_k)$ be an unordered sequence of random variables such ...
1
vote
0answers
42 views

difference between factorization machines and collaborative filtering

Can factorization machines be considered as a collaborative filtering method ? If so, do they belong to user-based category or item-based category?
1
vote
0answers
75 views

Stochstic Gradient Descent for Collaborative Filtering

I am currently implementing a model-based Collaborative Filtering approach which relies on the matrix factorization technique. More precisely, I want to factorize the rating matrix ...
1
vote
0answers
16 views

What's the best recommender system for only first-visit recommendations?

I'm trying to build a recommender system that recommends items to users. However, users come only once and need to be accurately recommended on their first and unique visit, thus making cold start the ...