Collaborative filtering models have undoubtedly dominated the scene of recommender systems. However, these methods do not take into account valuable item characteristics. On the other side, content-based algorithms only use this kind of information and may fail to generalize. Some collaborative filtering techniques have recently used side information about items, but they end up being huge models using thousands of features for modeling a single user-item interaction. In this paper, we present KGFlex, a sparse and expressive model based on feature embeddings. KGFlex studies which features are considered by each user when consuming an item. Then, it models each user-item interaction as a factorized entropy-driven combination of the only item features relevant to the user. An extensive experimental evaluation shows the approach's efectiveness, considering the recommendation results' accuracy, diversity, and induced bias.
The outstanding accuracy of collaborative filtering techniques has undoubtedly helped
recommender systems getting famous. However, these methods are based on the simple idea to
recommend certain items since "similar users have experienced those items", or "other users,
who have experienced the same items, have also experienced those items". On the contrary,
content-based recommendation algorithms aim to recommend new items that share the same
patterns of features of items liked in the past. The use of content features can make the model
interpretable [
In this work, we introduce KGFlex, a knowledge-aware recommendation system, that tackles this issue with a sparse and expressive model based on feature embeddings. KGFlex describes the catalog using features extracted from publicly available knowledge graphs, one of the most impactful and relevant sources for knowledge-aware recommender systems. Then, low-dimensionality embeddings are adopted to represent the semantic item features. Using an entropy-based strategy, KGFlex analyzes the users’ history to study the user-specific decision-making process of consuming or not consuming an item. Thus, the subsets of item features relevant to the user in her decision-making process are adopted to model the user-item interaction.
To evaluate the performance of KGFlex, we conduct extensive experiments on two diferent publicly available datasets. We evaluate the accuracy and diversity of recommendation results and analyze whether the algorithm produces biased recommendations. The results show that KGFlex has competitive accuracy performance and, at the same time, generates highly diversified recommendations with a low induced bias.
KGFlex exploits the knowledge encoded in a knowledge graph as side information to characterize both items and users. One of the main assumptions is that users decide to enjoy an item based on a subset of its characteristics, implying that not all the item features are equally important. In the following, we show how KGFlex describes each user and item with a set of features. Taking a cue from information theory KGFlex exploits the notion of information gain to measure the relevance of a feature for a user in deciding to consume or not an item.
From to Decision-Making. A knowledge graph can be represented as a set of triples where entities are linked to each other by binary relations. Each connection in is then a triple →− , where is a subject entity, is a relation (predicate), and is an object entity. If we consider chains of predicates that connect two entities at a higher depth, a -hop predicate can be defined as = ⟨ 1,..., ⟩ if →− 1 1 →− 2 ... →− ∈ . For convenience, ℎ( ) = for : →− ∈ denotes the depth of the predicate chain. When no confusion arises, from now on we will use →− to denote a generic chain with ℎ( ) ∈ {1,...,}.
Given a collection of items ℐ and a knowledge graph , we assume each element in ∈ ℐ has a mapping to a corresponding entity in . Under this assumption, an item can be explored, at depth , to identify the set ℱ() of the semantic features describing it: ℱ() = {⟨, ⟩ | → − ∈ ,ℎ( ) ∈ {1,...,}}. (1) We describe each user ∈ with the set ℱ = ⋃︀∈ℐ ℱ(), i.e. all the features representing the items ℐ ⊆ ℐ enjoyed by . Finally, we define the overall set of features in the system as ℱ () = ⋃︀∈ℐ ℱ() of features in the system: In the following, the () superscript is omitted whenever it is not relevant in the context.
Once items and users have been associated with their set of features, we use the notion of information gain to measure the importance of each feature for a user in deciding whether to consume or not consume an item, i.e. in distinguishing positive from negative items in the dataset. Indeed, given a dataset with a certain extent of entropy (uncertainty) on the target attribute, the information gain (,) measures the expected reduction in information entropy obtained from the observation of the value of the -th attribute of a sample x. To this aim, we build, for each user , a balanced dataset with all the consumed items from ℐ and the same amount of negative items randomly picked up from ⋃︀∈,̸= ℐ ∖ ℐ. For each of these positive and negative items, is provided with a sample whose attributes correspond to the features in ℱ and indicate the presence ( = 1) or the absence ( = 0) of the corresponding feature in ℱ. Therefore, the attribute provides an information gain in distinguishing positive from negative samples equal to (, ) = 1− (| = 1)− (| = 0).
We finally associate a weight = (, ) to each pair of user and feature to represent the influence of a feature —in the view of the user— in the prediction of user-item interactions. Sparse Embeddings. KGFlex models the features in ℱ as collaboratively learned embeddings in a latent space. Since KGFlex promotes the idea of having user fine-tuned versions of the same model, we have both a global representation of the features in ℱ and a personal view, for each user , of the features in ℱ ⊆ ℱ . Notably, the model is structured into two distinct parts. On the one hand, KGFlex keeps a set of global trainable embeddings and biases shared among all the users, with = {(g ∈ R , ∈ R), ∀ ∈ ℱ }. On the other hand, each user in KGFlex also has his/her personal representation of the features he/she interacted with, i.e., the features in ℱ. These embeddings are collected within the set , defined as = {p ∈ R , ∀ ∈ ℱ}. Then, the inner product between the personal representation p and the global representation g , plus a bias value , estimates the afinity of user to feature . The sum of such afinities for all the features in ℱ = ℱ ∩ℱ, weighted according to the pre-computed entropy-based coeficients, estimates the interaction ˆ between user and item : ˆ = ∑︁ (pg + ).
∈ℱ (2) Eq. (2) encodes the strategy KGFlex exploits to handle the features: it takes advantage of user profile to involve only a small subset of them in the estimate of the user-item afinity.
To learn the model parameters, KGFlex adopts Bayesian Personalized Ranking (BPR), the most common pair-wise Learning to Rank strategy, that, given a training set = {(,+,− ) | + ∈ ℐ ∧ − ∈ ℐ∖ℐ, ∀ ∈ }, optimizes the loss = ∑︀(,+,− )∈ ln (ˆ+ − ˆ− ), with the assumption that a user prefers a consumed item + over a non-consumed item − .
Experimental Setting. The evaluation of the performance of KGFlex is conducted on two
wellknown datasets: Yahoo! Movies and Facebook Books. The datasets have been binarized, retaining
ratings of 3 or higher, and have been preprocessed with iterative 10-core and 5-core, respectively.
The semantic features have been retrieved through a 2-depth exploration of the DBpedia ,
removing some useless features [
We compare KGFlex with BPR-MF [
We have measured the recommendation accuracy with nDCG [
Main Results. Table 1 depicts the evaluation outcome for the aforementioned metrics with a cutof of 10. For Yahoo! Movies, KGFlex is outperformed exclusively by kaHFM and MF, but continues to show acceptable accuracy results. It is noteworthy that KGFlex significantly outperforms BPRMF, albeit both are learned with a pair-wise BPR optimization, hence underlining the beneficial role of the extracted knowledge. Moreover, examining the item coverage and Gini values, we note the high degree of personalization provided by KGFlex. We link this result to the personalized view of the knowledge granted by the framework. Moreover, in KGFlex the collaborative signal on explicit user interests ensures to recommend diverse items among the ones sharing characteristics of interest for the user. The aforementioned behavior is not confirmed in Facebook Books. Indeed, the accuracy results seem to remain below the performance of other approaches. However, the diversity results show how BPR-MF, MF, and NeuMF may have been flooded by popularity signal, which led them to perform poorly regarding the item coverage and Gini metrics. Instead, KGFlex does not suffer from this problem and approaches the superior performance of Item-kNN in terms of diversity.
Oftentimes, recommender systems fail to recommend unpopular items, which tend to remain
underrepresented [