Polyacrylamide (PAM)-based microspheres are commonly used as water plugging and profile control agents, but the poor mechanical strength and few studies on the dispersion stability, both of which are closely related to the profile control performance, limit the application of microspheres. Herein, we synthesize nanoscale PAM-based copolymer hydrogel microspheres with an inverse microemulsion copolymerization of acrylamide (AM) and 2-methyl-2-acrylic amide propyl sulfonic acid (AMPS) in the presence of vinyl-functionalized silica nanoparticles (VSNPs). The results show that a small amount of VSNPs (1.0 wt %) increases the compressive strength of the hydrogel by 0.6 times. The swollen nanoscale PAM/silica hydrogel microspheres show good dispersion stability. VSNPs significantly improve the elasticity of the hydrogel microspheres, and their dispersion stability under high temperature and high-salinity conditions. The simulation evaluation of core plugging suggests that the plugging rate of PAM-based polymer/silica hybrid microspheres with addition of 0.7 wt % VSNPs increases from 80% to 92% compared to neat polymer microspheres. This work provides a novel design of nanoscale cross-linked microspheres for deep profile control in different geological environments.

Given an entity in a source domain, finding its matched entities from another (target) domain is an important task in many applications. Traditionally, the problem was usually addressed by first extracting major keywords corresponding to the source entity and then query relevant entities from the target domain using those keywords. However, the method would inevitably fails if the two domains have less or no overlapping in the content. An extreme case is that the source domain is in English and the target domain is in Chinese.In this paper, we formalize the problem as entity matching across heterogeneous sources and propose a probabilistic topic model to solve the problem. The model integrates the topic extraction and entity matching, two core subtasks for dealing with the problem, into a unified model. Specifically, for handling the text disjointing problem, we use a cross-sampling process in our model to extract topics with terms coming from all the sources, and leverage existing matching relations through latent topic layers instead of at text layers. Benefit from the proposed model, we can not only find the matched documents for a query entity, but also explain why these documents are related by showing the common topics they share. Our experiments in two real-world applications show that the proposed model can extensively improve the matching performance (+19.8% and +7.1% in two applications respectively) compared with several alternative methods.

No abstract

No abstract