Search for: All records

Conversational recommender systems (CRS) dynamically obtain the users' preferences via multi-turn questions and answers. The existing CRS solutions are widely dominated by deep reinforcement learning algorithms. However, deep reinforcement learning methods are often criticized for lacking interpretability and requiring a large amount of training data to perform.In this paper, we explore a simpler alternative and propose a decision tree based solution to CRS. The underlying challenge in CRS is that the same item can be described differently by different users. We show that decision trees are sufficient to characterize the interactions between users and items, and solve the key challenges in multi-turn CRS: namely which questions to ask, how to rank the candidate items, when to recommend, and how to handle user's negative feedback on the recommendations. Firstly, the training of decision trees enables us to find questions which effectively narrow down the search space. Secondly, by learning embeddings for each item and tree nodes, the candidate items can be ranked based on their similarity to the conversation context encoded by the tree nodes. Thirdly, the diversity of items associated with each tree node allows us to develop an early stopping strategy to decide when to make recommendations. Fourthly, when the user rejects a recommendation, we adaptively choose the next decision tree to improve subsequent questions and recommendations. Extensive experiments on three publicly available benchmark CRS datasets show that our approach provides significant improvement to the state of the art CRS methods. 

Octopus-inspired switchable adhesives are integrated with sensing, processing, and control for robust underwater manipulation. 

Conductive adhesives are required for the integration of dissimilar material components to create soft electronic and robotic systems. Here, a heterogeneous liquid metal‐based conductive adhesive is developed that reversibly attaches to diverse surfaces with high stretchability (>100% strain), low modulus (<100 kPa), and strain‐invariant electrical conductivity. This SofT integrated composite with tacK through liquid metal (STICK‐LM) adhesive consists of a heterogeneous graded film with a liquid metal‐rich side that is embossed at prescribed locations for electrical conductivity and an electrically insulating adhesive side for integration. Adhesion behavior is tuned for adhesion energies > 70 Jm⁻²(≈ 25x enhancement over unmodified composites) and described with a viscoelastic analysis, providing design guidelines for controllable yet reversible adhesion in electrically conductive systems. The architecture of STICK‐LM adhesives provides anisotropic and heterogeneous electrical conductivity and enables direct integration into soft functional systems. This is demonstrated with deformable fuses for robotic joints, repositionable electronics that rapidly attach on curvilinear surfaces, and stretchable adhesive conductors with nearly constant electrical resistance. This study provides a methodology for electrically conductive, reversible adhesives for electrical and mechanical integration of multicomponent systems in emerging technologies.

 

Vaccination is the primary intervention for controlling the spread of infectious diseases. A certain level of vaccination rate (referred to as “herd immunity”) is needed for this intervention to be effective. However, there are concerns that herd immunity might not be achieved due to an increasing level of hesitancy and opposition to vaccines. One of the primary reasons for this is the cost of non-conformance with one’s peers. We use the framework of network coordination games to study the persistence of anti-vaccine sentiment in a population. We extend it to incorporate the opposing forces of the pressure of conforming to peers, herd-immunity and vaccination benefits. We study the structure of the equilibria in such games, and the characteristics of unvaccinated nodes. We also study Stackelberg strategies to reduce the number of nodes with anti-vaccine sentiment. Finally, we evaluate our results on different kinds of real world social networks. 

The devastating economic and societal impacts of COVID-19 can be substantially compounded by other secondary events that increase individuals’ exposure through mass gatherings such as protests or sheltering due to a natural disaster. Based on the Crichton’s Risk Triangle model, this paper proposes a Markov Chain Monte Carlo (MCMC) simulation framework to estimate the impact of mass gatherings on COVID-19 infections by adjusting levels of exposure and vulnerability. To this end, a case study of New York City is considered, at which the impact of mass gathering at public shelters due to a hypothetical hurricane will be studied. The simulation results will be discussed in the context of determining effective policies for reducing the impact of multi-hazard generalizability of our approach to other secondary events that can cause mass gatherings during a pandemic will also be discussed. 

Wearable electronics, conformable sensors, and soft/micro-robotics require conductive yet stretchable thin films. However, traditional free standing metallic thin films are often brittle, inextensible, and must be processed in strict environments. This limits implementation into soft technologies where high electrical conductivity must be achieved while maintaining high compliance and conformability. Here we show a liquid metal elastomeric thin film (LET) composite with elastomer-like compliance (modulus < 500 kPa) and stretchability (>700%) with metallic conductivity (sheet resistance < 0.1 Ω/□). These 30–70µm thin films are highly conformable, free standing, and display a unique Janus microstructure, where a fully conductive activated side is accompanied with an opposite insulated face. LETs display exceptional electro-mechanical characteristics, with a highly linear strain-resistance relationship beyond 700% deformation while maintaining a low resistance. We demonstrate the multifunctionality of LETs for soft technologies by leveraging the unique combination of high compliance and electrical conductivity with transfer capabilities for strain sensing on soft materials, as compliant electrodes in a dielectric elastomeric actuator, and as resistive heaters for a liquid crystal elastomer.

 

We report a pulsed laser annealing method to convert carbon fibers and nanotubes into diamond fibers at ambient temperature and pressure in air. The conversion of carbon nanofibers and nanotubes into diamond nanofibers involves melting in a super undercooled state using nanosecond laser pulses, and quenching rapidly to convert into phase-pure diamond. The conversion process occurs at ambient temperature and pressure, and can be carried out in air. The structure of diamond fibers has been confirmed by selected-area electron diffraction in transmission electron microscopy, electron-back-scatter-diffraction in high-resolution scanning electron microscopy, all showing characteristic diffraction lines for the diamond structure. The bonding characteristics were determined by Raman spectroscopy with a strong peak near 1332 cm −1 , and high-resolution electron-energy-loss spectroscopy in transmission electron microscopy with a characteristic peak at 292 eV for σ* for sp 3 bonding and the absence of π* for sp 2 bonding. The Raman peak at 1332 cm −1 downshifts to 1321 cm −1 for diamond nanofibers due to the phonon confinement in nanodiamonds. These laser-treated carbon fibers with diamond seeds are used to grow larger diamond crystallites further by using standard hot-filament chemical vapor deposition (HFCVD). We compare these results with those obtained without laser treating the carbon fibers. The details of diamond conversion and HFCVD growth are presented in this paper.