Hierarchical Hyperedge Embedding-Based Representation Learning for Group Recommendation

Group recommendation aims to recommend items to a group of users. In this work, we study group recommendation in a particular scenario, namely occasional group recommendation, where groups are formed ad hoc and users may just constitute a group for the first time—that is, the historical group-item interaction records are highly limited. Most state-of-the-art works have addressed the challenge by aggregating group members’ personal preferences to learn the group representation. However, the representation learning for a group is most complex beyond the aggregation or fusion of group member representation, as the personal preferences and group preferences may be in different spaces and even orthogonal. In addition, the learned user representation is not accurate due to the sparsity of users’ interaction data. Moreover, the group similarity in terms of common group members has been overlooked, which, however, has the great potential to improve the group representation learning. In this work, we focus on addressing the aforementioned challenges in the group representation learning task, and devise a hierarchical hyperedge embedding-based group recommender, namely HyperGroup. Specifically, we propose to leverage the user-user interactions to alleviate the sparsity issue of user-item interactions, and design a graph neural network-based representation learning network to enhance the learning of individuals’ preferences from their friends’ preferences, which provides a solid foundation for learning groups’ preferences. To exploit the group similarity (i.e., overlapping relationships among groups) to learn a more accurate group representation from highly limited group-item interactions, we connect all groups as a network of overlapping sets (a.k.a. hypergraph), and treat the task of group preference learning as embedding hyperedges (i.e., user sets/groups) in a hypergraph, where an inductive hyperedge embedding method is proposed. To further enhance the group-level preference modeling, we develop a joint training strategy to learn both user-item and group-item interactions in the same process. We conduct extensive experiments on two real-world datasets, and the experimental results demonstrate the superiority of our proposed HyperGroup in comparison to the state-of-the-art baselines.

The Anti-pornography Civil Rights Ordinances, 1983–1991

The chapter analyzes attempted civil rights legislation against pornography (the “MacKinnon-Dworkin” ordinance). It delineates its underlying hierarchy theory: consciousness-raising, group representation, and intersectional legal analysis—a commanding approach to end oppression through civil lawsuits against producers and disseminators, avoiding criminal law. The contemporaneous critics’ charges of “rigidity” and “one-sidedness” are found wanting, inadequately apprehending hierarchy and subordination. The ordinances’ definitions are shown to target provably harmful material only, preventing overbreadth and vagueness. A legal argument is advanced that the ordinances are narrowly tailored to serve a compelling interest, the incidental restrictions on alleged First Amendment freedoms are no greater than is essential to further their interest, and the definitions are sufficiently analogous to other unprotected expressions (e.g., obscenity and group libel). The Seventh Circuit’s judicial invalidation in American Booksellers Association v. Hudnut (1985) is found based on ideology rather than law, political ideas rather than legislated rules.

Composition of place, minority vs. majority group-status, & contextualized experience: The role of level of group representation, perceiving place in group-based terms, and sense of belonging in shaping collective behavior

The current studies (N = 1,709) explore why demographic composition of place matters. First, this work demonstrates that relative level of group representation affects one’s experience of place in the form of self-definition (self-categorization), perceptions of place being representative or characteristic of factors that distinguish the group from others (place-prototypicality), and sense of belonging (place-identification; Studies 1a-1e; Studies 2a & 2b). Second, the studies illustrate that group representation within place shapes the way group member’s approach (i.e., expectations of group-based treatment and procedural justice; Studies 2a-2c), understand (i.e., attribution for group-based events, Study 2b; responsiveness to bias-reduction intervention, Study 4a; sense of solidarity, Study 4b), and behave (i.e., prejudice, Studies 3a & 3b; collective action, Study 4c). More broadly, I present a Social identity Paradigm for Contextualized Experience (SPACE) that provides an organizing framework for the study of the impact of characteristics of place on social identity-based contextualized experience and (in turn) collective behavior. Taken together, the findings provide evidence of distinct psychological experience and orientation as a function of minority versus majority-group status within place, as well as for a group-based approach to place. Implications for the study of collective and intergroup behavior are discussed.

Exploring the landscape for soft theorems of nonlinear sigma models

We generalize soft theorems of the nonlinear sigma model beyond the $$ \mathcal{O} $$ O (p2) amplitudes and the coset of SU(N) × SU(N)/SU(N). We first discuss the universal flavor ordering of the amplitudes for the Nambu-Goldstone bosons, so that we can reinterpret the known $$ \mathcal{O} $$ O (p2) single soft theorem for SU(N) × SU(N)/SU(N) in the context of a general symmetry group representation. We then investigate the special case of the fundamental representation of SO(N), where a special flavor ordering of the “pair basis” is available. We provide novel amplitude relations and a Cachazo-He-Yuan formula for such a basis, and derive the corresponding single soft theorem. Next, we extend the single soft theorem for a general group representation to $$ \mathcal{O} $$ O (p4), where for at least two specific choices of the $$ \mathcal{O} $$ O (p4) operators, the leading non-vanishing pieces can be interpreted as new extended theory amplitudes involving bi-adjoint scalars, and the corresponding soft factors are the same as at $$ \mathcal{O} $$ O (p2). Finally, we compute the general formula for the double soft theorem, valid to all derivative orders, where the leading part in the soft momenta is fixed by the $$ \mathcal{O} $$ O (p2) Lagrangian, while any possible corrections to the subleading part are determined by the $$ \mathcal{O} $$ O (p4) Lagrangian alone. Higher order terms in the derivative expansion do not contribute any new corrections to the double soft theorem.