As large-scale language models become the standard for text generation, there is a greater need to tailor the generations to be more or less concise, targeted, and informative, depending on the audience/application. Existing control approaches primarily adjust the semantic (e.g., emotion, topics), structural (e.g., syntax tree, parts-of-speech), and lexical (e.g., keyword/phrase inclusion) properties of text, but are insufficient to accomplish complex objectives such as pacing which control the complexity and readability of the text.In this paper, we introduce CEV-LM -a lightweight, semi-autoregressive language model that utilizes constrained edit vectors to control three complementary metrics (speed, volume, and circuitousness) that quantify the shape of text (e.g., pacing of content).We study an extensive set of state-of-the-art CTG models and find that CEV-LM provides significantly more targeted and precise control of these three metrics while preserving semantic content, using less training data, and containing fewer parameters. 1 Speed How quickly content changes Original: If you are in Austin, you have to take time and check out the place.It's a great brewery

Entity matching is the task of linking records from different sources that refer to the same real-world entity.Past work has primarily treated entity linking as a standard supervised learning problem.However, supervised entity matching models often do not generalize well to new data, and collecting exhaustive labeled training data is often cost prohibitive.Further, recent efforts have adopted LLMs for this task in few/zero-shot settings, exploiting their general knowledge.But LLMs are prohibitively expensive for performing inference at scale for real-world entity matching tasks.As an efficient alternative, we re-cast entity matching as a conditional generation task as opposed to binary classification.This enables us to "distill" LLM reasoning into smaller entity matching models via natural language explanations.This approach achieves strong performance, especially on out-of-domain generalization tests (↑10.85%F-1) where standalone generative methods struggle.We perform ablations that highlight the importance of explanations, both for performance and model robustness.Explain matching label class given the entity descriptions: Label: Match E_a: Nike Sportswear AF-1 488298-436 MN Navy.E_b: Air Force 1 [BRAND]

In recent times, deep learning methods have supplanted conventional collaborative filtering approaches as the backbone of modern recommender systems. However, their gains are skewed towards popular items with a drastic performance drop for the vast collection of long-tail items with sparse interactions. Moreover, we empirically show that prior neural recommenders lack the resolution power to accurately rank relevant items within the long-tail.

The quality variance in user-generated content is a major bottleneck to serving communities on online platforms. Current content ranking methods primarily evaluate text and non-textual content features of each user post in isolation. In this paper, we demonstrate the utility of considering the implicit and explicit relational aspects across user content to assess their quality. First, we develop a modular platform-agnostic framework to represent the contrastive (or competing) and similarity-based relational aspects of user-generated content via independently induced content graphs. Second, we develop two complementary graph convolutional operators that enable feature contrast for competing content and feature smoothing/sharing for similar content. Depending on the edge semantics of each content graph, we embed its nodes via one of the above two mechanisms. We also show that our contrastive operator creates discriminative magnification across the embeddings of competing posts. Third, we show a surprising result-applying classical boosting techniques to combine final-layer embeddings across the content graphs significantly outperforms the typical stacking, fusion, or neighborhood embedding aggregation methods in graph convolutional architectures. We exhaustively validate our method via accepted answer prediction over fifty diverse Stack-Exchange (https://stackexchange.com/) websites with consistent relative gains of over 5% accuracy over state-of-the-art neural, multi-relational and textual baselines.

We present InfoMotif, a new semi-supervised, motif-regularized, learning framework over graphs. We overcome two key limitations of message passing in popular graph neural networks (GNNs): localization (a k-layer GNN cannot utilize features outside the k-hop neighborhood of the labeled training nodes) and over-smoothed (structurally indistinguishable) representations. We propose the concept of attributed structural roles of nodes based on their occurrence in different network motifs, independent of network proximity. Two nodes share attributed structural roles if they participate in topologically similar motif instances over co-varying sets of attributes. Further, InfoMotif achieves architecture independence by regularizing the node representations of arbitrary GNNs via mutual information maximization. Our training curriculum dynamically prioritizes multiple motifs in the learning process without relying on distributional assumptions in the underlying graph or the learning task. We integrate three state-of-the-art GNNs in our framework, to show significant gains (3-10% accuracy) across six diverse, real-world datasets. We see stronger gains for nodes with sparse training labels and diverse attributes in local neighborhood structures.

The rapid proliferation of new users and items on the social web has aggravated the gray-sheep user/long-tail item challenge in recommender systems. Historically, cross-domain co-clustering methods have successfully leveraged shared users and items across dense and sparse domains to improve inference quality. However, they rely on shared rating data and cannot scale to multiple sparse target domains (i.e., the one-to-many transfer setting). This, combined with the increasing adoption of neural recommender architectures, motivates us to develop scalable neural layer-transfer approaches for cross-domain learning. Our key intuition is to guide neural collaborative filtering with domain-invariant components shared across the dense and sparse domains, improving the user and item representations learned in the sparse domains. We leverage contextual invariances across domains to develop these shared modules, and demonstrate that with user-item interaction context, we can learn-to-learn informative representation spaces even with sparse interaction data. We show the effectiveness and scalability of our approach on two public datasets and a massive transaction dataset from Visa, a global payments technology company (19% Item Recall, 3x faster vs. training separate models for each domain). Our approach is applicable to both implicit and explicit feedback settings.

Recent years have witnessed tremendous interest in understanding and predicting information spread on social media platforms such as Twitter, Facebook, etc. Existing diffusion prediction methods primarily exploit the sequential order of influenced users by projecting diffusion cascades onto their local social neighborhoods. However, this fails to capture global social structures that do not explicitly manifest in any of the cascades, resulting in poor performance for inactive users with limited historical activities.

Some social networks provide explicit mechanisms to allocate social rewards such as reputation based on users’ actions, while the mechanism is more opaque in other networks. Nonetheless, there are always individuals who obtain greater rewards and reputation than their peers. An intuitive yet important question to ask is whether these successful users employ strategic behaviors to become influential. It might appear that the influencers ”have gamed the system.” However, it remains difficult to conclude the rationality of their actions due to factors like the combinatorial strategy space, inability to determine payoffs, and resource limitations faced by individuals. The challenging nature of this question has drawn attention from both the theory and data mining communities. Therefore, in this paper, we are motivated to investigate if resource-limited individuals discover strategic behaviors associated with high payoffs when producing collaborative/interactive content in social networks. We propose a novel framework of Dynamic Dual Attention Networks (DDAN) which models individuals’ content production strategies through a generative process, under the influence of social interactions involved in the process. Extensive experimental results illustrate the model’s effectiveness in user behavior modeling. We make three strong empirical findings: (1) Different strategies give rise to different social payoffs; (2) The best performing individuals exhibit stability in their preference over the discovered strategies, which indicates the emergence of strategic behavior; and (3) The stability of a user’s preference is correlated with high payoffs.

This paper proposes a novel framework to incorporate social regularization for item recommendation. Social regularization grounded in ideas of homophily and influence appears to capture latent user preferences. However, there are two key challenges: first, the importance of a specific social link depends on the context and second, a fundamental result states that we cannot disentangle homophily and influence from observational data to determine the effect of social inference. Thus we view the attribution problem as inherently adversarial where we examine two competing hypothesis---social influence and latent interests---to explain each purchase decision. We make two contributions. First, we propose a modular, adversarial framework that decouples the architectural choices for the recommender and social representation models, for social regularization. Second, we overcome degenerate solutions through an intuitive contextual weighting strategy, that supports an expressive attribution, to ensure informative social associations play a larger role in regularizing the learned user interest space. Our results indicate significant gains (5-10% relative [email protected]) over state-of-the-art baselines across multiple publicly available datasets.

This paper studies the problem of learning latent representations or embeddings for users in social networks, by leveraging relationship semantics associated with each link. User embeddings are low-dimensional vector-space representations designed to preserve structural proximity indicated by the pairwise relationships. In social networks, the closeness (or proximity) between pairs of users is very different w.r.t. multiple social relationships and thus cannot be represented accurately using a single embedding space. Furthermore, social networks pose a unique challenge of relationship label sparsity that precludes the application of knowledge-graph embedding techniques.In this paper, we associate each observed link with multiple relationship types through relationship weights and learn projection matrices for each relationship type to model the social distance (or proximity) between users specific to each relationship. We propose a novel two-step mutual enhancement framework to iteratively (a) learn user embeddings preserving relationship-specific proximity, and (b) link-relationship weights capturing the role of each link in multiple relationship types. The first step learns user embeddings optimizing relationship-specific proximity, while fixing the relationship weights (or roles) for each link. In the second step, the user embeddings and corresponding projection matrices are assumed to be fixed, while the link-relationship weights are learned. We demonstrate that the relationship-aware user embeddings learned through this mutual enhancement framework, are more effective in representing the users and outperform representative baseline techniques in multi-label classification and relationship prediction tasks.

This paper proposes an approach to learn robust behavior representations in online platforms by addressing the challenges of user behavior skew and sparse participation. Latent behavior models are important in a wide variety of applications: recommender systems; prediction; user profiling; community characterization. Our framework is the first to jointly address skew and sparsity across graphical behavior models. We propose a generalizable bayesian approach to partition users in the presence of skew while simultaneously learning latent behavior profiles over these partitions to address user-level sparsity. Our behavior profiles incorporate the temporal activity and links between participants, although the proposed framework is flexible to introduce other definitions of participant behavior. Our approach explicitly discounts frequent behaviors and learns variable size partitions capturing diverse behavior trends. The partitioning approach is data-driven with no rigid assumptions, adapting to varying degrees of skew and sparsity.

In recent times, deep neural networks have found success in Collaborative Filtering (CF) based recommendation tasks. By parametrizing latent factor interactions of users and items with neural architectures, they achieve significant gains in scalability and performance over matrix factorization. However, the long-tail phenomenon in recommender performance persists on the massive inventories of online media or retail platforms. Given the diversity of neural architectures and applications, there is a need to develop a generalizable and principled strategy to enhance long-tail item coverage.

A search query, being a very concise grounding of user intent, could potentially have many possible interpretations. Search engines hedge their bets by diversifying top results to cover multiple such possibilities so that the user is likely to be satisfied, whatever be her intended interpretation. Diversified Query Expansion is the problem of diversifying query expansion suggestions, so that the user can specialize the query to better suit her intent, even before perusing search results. In this paper, we consider the usage of semantic resources and tools to arrive at improved methods for diversified query expansion. In particular, we develop two methods, those that leverage Wikipedia and pre-learnt distributional word embeddings respectively. Both the approaches operate on a common three-phase framework; that of first taking a set of informative terms from the search results of the initial query, then building a graph, following by using a diversity-conscious node ranking to prioritize candidate terms for diversified query expansion. Our methods differ in the second phase, with the first method Select-Link-Rank (SLR) linking terms with Wikipedia entities to accomplish graph construction; on the other hand, our second method, Select-Embed-Rank (SER), constructs the graph using similarities between distributional word embeddings. Through an empirical analysis and user study, we show that SLR ourperforms state-of-the-art diversified query expansion methods, thus establishing that Wikipedia is an effective resource to aid diversified query expansion. Our empirical analysis also illustrates that SER outperforms the baselines convincingly, asserting that it is the best available method for those cases where SLR is not applicable; these include narrow-focus search systems where a relevant knowledge base is unavailable. Our SLR method is also seen to outperform a state-of-the-art method in the task of diversified entity ranking.

This paper studies the automated categorization and extraction of scientific concepts from titles of scientific articles, in order to gain a deeper understanding of their key contributions and facilitate the construction of a generic academic knowledgebase. Towards this goal, we propose an unsupervised, domain-independent, and scalable two-phase algorithm to type and extract key concept mentions into aspects of interest (e.g., Techniques, Applications, etc.). In the first phase of our algorithm we propose PhraseType, a probabilistic generative model which exploits textual features and limited POS tags to broadly segment text snippets into aspect-typed phrases. We extend this model to simultaneously learn aspect-specific features and identify academic domains in multi-domain corpora, since the two tasks mutually enhance each other. In the second phase, we propose an approach based on adaptor grammars to extract fine grained concept mentions from the aspect-typed phrases without the need for any external resources or human effort, in a purely data-driven manner. We apply our technique to study literature from diverse scientific domains and show significant gains over state-of-the-art concept extraction techniques. We also present a qualitative analysis of the results obtained.