Matrix Factorization (MF) is a common method for generating recommendations, where the proximity of entities like users or items in the embedded space indicates their similarity to one another. Though almost all applications implicitly use a Euclidean embedding space to represent two entity types, recent work has suggested that a hyperbolic Poincar\'e ball may be more well suited to representing multiple entity types, and in particular, hierarchies. We describe a novel method to embed a hierarchy of related music entities in hyperbolic space. We also describe how a parametric empirical Bayes approach can be used to estimate link reliability between entities in the hierarchy. Applying these methods together to build personalized playlists for users in a digital music service yielded a large and statistically significant increase in performance during an A/B test, as compared to the Euclidean model.
Personalized recommendation models (RecSys) are one of the most popular machine learning workload serviced by hyperscalers. A critical challenge of training RecSys is its high memory capacity requirements, reaching hundreds of GBs to TBs of model size. In RecSys, the so-called embedding layers account for the majority of memory usage so current systems employ a hybrid CPU-GPU design to have the large CPU memory store the memory hungry embedding layers. Unfortunately, training embeddings involve several memory bandwidth intensive operations which is at odds with the slow CPU memory, causing performance overheads. Prior work proposed to cache frequently accessed embeddings inside GPU memory as means to filter down the embedding layer traffic to CPU memory, but this paper observes several limitations with such cache design. In this work, we present a fundamentally different approach in designing embedding caches for RecSys. Our proposed ScratchPipe architecture utilizes unique properties of RecSys training to develop an embedding cache that not only sees the past but also the "future" cache accesses. ScratchPipe exploits such property to guarantee that the active working set of embedding layers can "always" be captured inside our proposed cache design, enabling embedding layer training to be conducted at GPU memory speed.
The harmful impacts of algorithmic decision systems have recently come into focus, with many examples of systems such as machine learning (ML) models amplifying existing societal biases. Most metrics attempting to quantify disparities resulting from ML algorithms focus on differences between groups, dividing users based on demographic identities and comparing model performance or overall outcomes between these groups. However, in industry settings, such information is often not available, and inferring these characteristics carries its own risks and biases. Moreover, typical metrics that focus on a single classifier's output ignore the complex network of systems that produce outcomes in real-world settings. In this paper, we evaluate a set of metrics originating from economics, distributional inequality metrics, and their ability to measure disparities in content exposure in a production recommendation system, the Twitter algorithmic timeline. We define desirable criteria for metrics to be used in an operational setting, specifically by ML practitioners. We characterize different types of engagement with content on Twitter using these metrics, and use these results to evaluate the metrics with respect to the desired criteria. We show that we can use these metrics to identify content suggestion algorithms that contribute more strongly to skewed outcomes between users. Overall, we conclude that these metrics can be useful tools for understanding disparate outcomes in online social networks.
Precise user modeling is critical for online personalized recommendation services. Generally, users' interests are diverse and are not limited to a single aspect, which is particularly evident when their behaviors are observed for a longer time. For example, a user may demonstrate interests in cats/dogs, dancing and food \& delights when browsing short videos on Tik Tok; the same user may show interests in real estate and women's wear in her web browsing behaviors. Traditional models tend to encode a user's behaviors into a single embedding vector, which do not have enough capacity to effectively capture her diverse interests. This paper proposes a Sequential User Matrix (SUM) to accurately and efficiently capture users' diverse interests. SUM models user behavior with a multi-channel network, with each channel representing a different aspect of the user's interests. User states in different channels are updated by an \emph{erase-and-add} paradigm with interest- and instance-level attention. We further propose a local proximity debuff component and a highway connection component to make the model more robust and accurate. SUM can be maintained and updated incrementally, making it feasible to be deployed for large-scale online serving. We conduct extensive experiments on two datasets. Results demonstrate that SUM consistently outperforms state-of-the-art baselines.
The code smell is a sign of design and development flaws in a software system that reduces the reusability and maintainability of the system. Refactoring is done as an ongoing practice to remove the code smell from the program code. Among different code smells, the God class or Blob is one of the most common code smells. A god class contains too many responsibilities, violating object-oriented programming design's low coupling and high cohesiveness principles. This paper proposes an automatic approach to extracting a God class into multiple smaller classes with more specific responsibilities. To do this, we first construct a graph of methods (as nodes) for the concerning god class. The edge between any two methods is determined by their structural similarity, and the feature for each method is initialized using different semantic representation methods. Then, the variational graph auto-encoder is used to learn a vector representation for each method. Finally, the learned vectors are used to cluster methods into different groups to be recommended as refactored classes. We assessed the proposed framework using three different class cohesion metrics on sixteen actual God Classes collected from two well-known open-source systems. We also conducted a comparative study of our approach with a similar existing approach and found that the proposed approach generated better results for almost all the God Classes used in the experiment.
Recent advances in whole slide imaging (WSI) technology have led to the development of a myriad of computer vision and artificial intelligence (AI) based diagnostic, prognostic, and predictive algorithms. Computational Pathology (CPath) offers an integrated solution to utilize information embedded in pathology WSIs beyond what we obtain through visual assessment. For automated analysis of WSIs and validation of machine learning (ML) models, annotations at the slide, tissue and cellular levels are required. The annotation of important visual constructs in pathology images is an important component of CPath projects. Improper annotations can result in algorithms which are hard to interpret and can potentially produce inaccurate and inconsistent results. Despite the crucial role of annotations in CPath projects, there are no well-defined guidelines or best practices on how annotations should be carried out. In this paper, we address this shortcoming by presenting the experience and best practices acquired during the execution of a large-scale annotation exercise involving a multidisciplinary team of pathologists, ML experts and researchers as part of the Pathology image data Lake for Analytics, Knowledge and Education (PathLAKE) consortium. We present a real-world case study along with examples of different types of annotations, diagnostic algorithm, annotation data dictionary and annotation constructs. The analyses reported in this work highlight best practice recommendations that can be used as annotation guidelines over the lifecycle of a CPath project.
An increasing number and diversity of services are available, which result in significant challenges to effective reuse service during requirement satisfaction. There have been many service bundle recommendation studies and achieved remarkable results. However, there is still plenty of room for improvement in the performance of these methods. The fundamental problem with these studies is that they ignore the evolution of services over time and the representation gap between services and requirements. In this paper, we propose a dynamic representation learning and aligning based model called DySR to tackle these issues. DySR eliminates the representation gap between services and requirements by learning a transformation function and obtains service representations in an evolving social environment through dynamic graph representation learning. Extensive experiments conducted on a real-world dataset from ProgrammableWeb show that DySR outperforms existing state-of-the-art methods in commonly used evaluation metrics, improving [email protected]$ from $36.1\%$ to $69.3\%$.
With the increasing scale and diversification of interaction behaviors in E-commerce, more and more researchers pay attention to multi-behavior recommender systems that utilize interaction data of other auxiliary behaviors such as view and cart. To address these challenges in heterogeneous scenarios, non-sampling methods have shown superiority over negative sampling methods. However, two observations are usually ignored in existing state-of-the-art non-sampling methods based on binary regression: (1) users have different preference strengths for different items, so they cannot be measured simply by binary implicit data; (2) the dependency across multiple behaviors varies for different users and items. To tackle the above issue, we propose a novel non-sampling learning framework named \underline{C}riterion-guided \underline{H}eterogeneous \underline{C}ollaborative \underline{F}iltering (CHCF). CHCF introduces both upper and lower bounds to indicate selection criteria, which will guide user preference learning. Besides, CHCF integrates criterion learning and user preference learning into a unified framework, which can be trained jointly for the interaction prediction on target behavior. We further theoretically demonstrate that the optimization of Collaborative Metric Learning can be approximately achieved by CHCF learning framework in a non-sampling form effectively. Extensive experiments on two real-world datasets show that CHCF outperforms the state-of-the-art methods in heterogeneous scenarios.
Add to Chrome
Add to Firefox