Embedding Hierarchical Structures for Venue Category Representation

2022 ◽  
Vol 40 (3) ◽  
pp. 1-29
Author(s):  
Meng Chen ◽  
Lei Zhu ◽  
Ronghui Xu ◽  
Yang Liu ◽  
Xiaohui Yu ◽  
...  

Venue categories used in location-based social networks often exhibit a hierarchical structure, together with the category sequences derived from users’ check-ins. The two data modalities provide a wealth of information for us to capture the semantic relationships between those categories. To understand the venue semantics, existing methods usually embed venue categories into low-dimensional spaces by modeling the linear context (i.e., the positional neighbors of the given category) in check-in sequences. However, the hierarchical structure of venue categories, which inherently encodes the relationships between categories, is largely untapped. In this article, we propose a venue C ategory E mbedding M odel named Hier-CEM , which generates a latent representation for each venue category by embedding the Hier archical structure of categories and utilizing multiple types of context. Specifically, we investigate two kinds of hierarchical context based on any given venue category hierarchy and show how to model them together with the linear context collaboratively. We apply Hier-CEM to three tasks on two real check-in datasets collected from Foursquare. Experimental results show that Hier-CEM is better at capturing both semantic and sequential information inherent in venues than state-of-the-art embedding methods.

1998 ◽  
Vol 16 (1) ◽  
pp. 59-70 ◽  
Author(s):  
Tsion Avital ◽  
Gerald C. Cupchik

A series of four experiments were conducted to examine viewer perceptions of three sets of five nonrepresentational paintings. Increased complexity was embedded in the hierarchical structure of each set by carefully selecting colors and ordering them in each successive painting according to certain rules of transformation which created hierarchies. Experiment 1 supported the hypothesis that subjects would discern the hierarchical complexity underlying the sets of paintings. In Experiment 2 viewers rated the paintings on collative (complexity, disorder) and affective (pleasing, interesting, tension, and power) scales, and a factor analysis revealed that affective ratings were tied to complexity (Factor 1) but not to disorder (Factor 2). In Experiment 3, a measure of exploratory activity (free looking time) was correlated with complexity (Factor 1) but not with disorder (Factor 2). Multidimensional scaling was used in Experiment 4 to examine perceptions of the paintings seen in pairs. Dimension 1 contrasted Soft with Hard-Edged paintings, while Dimension 2 reflected the relative separation of figure from ground in these paintings. Together these results show that untrained viewers can discern hierarchical complexity in paintings and that this quality stimulates affective responses and exploratory activity.


Author(s):  
Masataka Yoshimura ◽  
Kazuhiro Izui

Abstract A large-scale machine system often has a general hierarchical structure. For hierarchical structures, optimization is difficult because many local optima almost always arise, however genetic algorithms that have a hierarchical genotype can be applied to treat such problems directly. Relations between the structural components are analyzed and this information is used to partition the hierarchical structure. Partitioning large-scale problems into sub-problems that can be solved using parallel processed GAs increases the efficiency of the optimization search. The optimization of large-scale systems then becomes possible due to information sharing of Pareto optimum solutions for the sub-problems.


F1000Research ◽  
2020 ◽  
Vol 9 ◽  
pp. 1246
Author(s):  
Ruizhu Huang ◽  
Charlotte Soneson ◽  
Felix G.M. Ernst ◽  
Kevin C. Rue-Albrecht ◽  
Guangchuang Yu ◽  
...  

Data organized into hierarchical structures (e.g., phylogenies or cell types) arises in several biological fields. It is therefore of interest to have data containers that store the hierarchical structure together with the biological profile data, and provide functions to easily access or manipulate data at different resolutions. Here, we present TreeSummarizedExperiment, a R/S4 class that extends the commonly used SingleCellExperiment class by incorporating tree representations of rows and/or columns (represented by objects of the phylo class). It follows the convention of the SummarizedExperiment class, while providing links between the assays and the nodes of a tree to allow data manipulation at arbitrary levels of the tree. The package is designed to be extensible, allowing new functions on the tree (phylo) to be contributed. As the work is based on the SingleCellExperiment class and the phylo class, both of which are popular classes used in many R packages, it is expected to be able to interact seamlessly with many other tools.


2020 ◽  
Vol 10 (22) ◽  
pp. 8003
Author(s):  
Yi-Chun Chen ◽  
Cheng-Te Li

In the scenarios of location-based social networks (LBSN), the goal of location promotion is to find information propagators to promote a specific point-of-interest (POI). While existing studies mainly focus on accurately recommending POIs for users, less effort is made for identifying propagators in LBSN. In this work, we propose and tackle two novel tasks, Targeted Propagator Discovery (TPD) and Targeted Customer Discovery (TCD), in the context of Location Promotion. Given a target POI l to be promoted, TPD aims at finding a set of influential users, who can generate more users to visit l in the future, and TCD is to find a set of potential users, who will visit l in the future. To deal with TPD and TCD, we propose a novel graph embedding method, LBSN2vec. The main idea is to jointly learn a low dimensional feature representation for each user and each location in an LBSN. Equipped with learned embedding vectors, we propose two similarity-based measures, Influential and Visiting scores, to find potential targeted propagators and customers. Experiments conducted on a large-scale Instagram LBSN dataset exhibit that LBSN2vec and its variant can significantly outperform well-known network embedding methods in both tasks.


2004 ◽  
Vol 126 (2) ◽  
pp. 217-224 ◽  
Author(s):  
Masataka Yoshimura ◽  
Kazuhiro Izui

A large-scale machine system often has a general hierarchical structure. For hierarchical structures, optimization is difficult because many local optima almost always arise, however genetic algorithms that have a hierarchical genotype can be applied to treat such problems directly. Relations between the structural components are analyzed and this information is used to partition the hierarchical structure. Partitioning large-scale problems into sub-problems that can be solved using parallel processed GAs increases the efficiency of the optimization search. The optimization of large-scale systems then becomes possible due to information sharing of Pareto optimum solutions for the sub-problems.


Catalysts ◽  
2021 ◽  
Vol 11 (11) ◽  
pp. 1388
Author(s):  
Youhe Wang ◽  
Zhihong Li ◽  
Chang Dai ◽  
Ningning Du ◽  
Tingting Li ◽  
...  

A unique method to prepare Zn-P co-modified hierarchical ZSM-5 zeolites was developed. The ZSM-5 zeolite was directly synthesized by a dry gel conversion without adding any templates or seeds. Afterwards, the hierarchical structure was endowed to the ZSM-5 zeolite by the sequential desilication-dealumination. Zn and P species were then introduced into the hierarchical ZSM-5 zeolites by the impregnation method and their activity in methanol to aromatics process was investigated. It was found that the Zn-P co-modified hierarchical ZSM-5 zeolites possessed more Zn-related Lewis acid sites, and the ratio of Zn(OH)+/ZnO was increased. The catalytic evaluation results revealed that the benzene, toluene and xylene (BTX) and aromatics selectivity were significantly improved from 20.59% and 29.41% of pristine ZSM-5 zeolite to 28.12% and 41.88% of Zn-P co-modified hierarchical counterpart (1.5Zn0.3P/HZSM-5), respectively. Owing to the introduced highly stable Zn-P co-modified hierarchical structures, the lifetime (conversion not less than 99%) of ZSM-5 zeolite during methanol to aromatics reaction was increased from 6 h to 18 h.


2017 ◽  
Vol 9 (2) ◽  
pp. 168781401668335 ◽  
Author(s):  
Hua-Yi Hsu ◽  
Bo-Ting Lin ◽  
You-Ren Hsu

Nanopost arrays are generally used in applications of reflection gratings and in changing material surface wettability. Nanopost arrays can be used as a passive component to induce dendritic self-organized hierarchical architectures. In this study, through the use of a phase-field model, we performed a three-dimensional numerical simulation to demonstrate that nanopost structures affect the expanding speed of the surface of a dendritic self-organized structure in the growing path of a hierarchical structure. Additionally, we demonstrated that the nanopost array arrangement on the surface changed the hierarchical structure branching. Finally, introducing an externally applied force to the system enabled the use of a nanopost as an active component. Nanopost surroundings were determined to significantly affect the final distribution of dendritic structures and induce hierarchical structures after an external force was introduced to the system.


2018 ◽  
Vol 36 (3) ◽  
pp. 700
Author(s):  
Tiago Peres da Silva SUGUIURA ◽  
Omar Cléo Neves PEREIRA ◽  
Waenya Fernandez de CARVALHO ◽  
Isolde Terezinha Santos PREVIDELLI

Data sets with complex structures is increasingly common in dental research. As consequences, statistical  methods to analyze and interpret these data must be efficient and robust. Hierarchical structures is one of  the most common kind of complex structures, and a proper approach is required. The multilevel modeling used to study hierarchical structures is a powerful tool which allows the collected data to be  analyzes in several levels. This study has as objective to make a literature review on multilevel linear models and to illustrate a three level model through a matrix procedure, without the use of specific software to estimate the parameters. With this model, we analyzed the vertical gingival retraction when using the substances: Naphazoline Chloridrate, Aluminium Chloride and without any substance. The intraclass correlation coefficient on dental level within patients showed that the hierarchical structure was important to accommodate the dependence within clusters.


2019 ◽  
Vol 13 (4) ◽  
pp. 916-934 ◽  
Author(s):  
Ajay Kumar Pandey ◽  
Manjushree Ghodke

Purpose The purpose of this paper is to develop an interpretive structural modeling (ISM) of barriers related to viability of Power Distribution Companies (discoms) in India. Design/methodology/approach Feedback from the Experts of Indian power sector has been taken as the basis to develop the model for barriers to viability of discoms, where major barriers have been identified through extent review of literature and through discussions with experts in the power sector keeping the viability of discoms in focus, and the hierarchical structure of barriers has been developed using ISM. Findings An interpretive structural model has been developed for discom-related factors (barriers) affecting its viability. The hierarchical structure portrays the impeding factors of viability and showcases that lack of regulatory effectiveness, inadequate tariffs and lack of government’s expenditures on power sector are the key barriers. Research limitations/implications This paper has implications for both practitioners and academics. For practitioners, it provides an indicative list of major barriers affecting the viability of Indian discoms. For academics, the methodology used provides a mechanism to conduct an exploratory study by identifying the key variables of interest and emphasizing their interactions through hierarchical structures. Originality/value The proposed model for barriers to viability of discoms developed through qualitative modeling technique is a pioneering effort altogether in the context of power distribution companies in India. Understanding contextual relationships among key barriers to viability of discom’s is neglected in existing literature, and this paper makes a contribution in this regard.


2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Binbin Jiang ◽  
Yong Yu ◽  
Hongyi Chen ◽  
Juan Cui ◽  
Xixi Liu ◽  
...  

AbstractWe demonstrate that the thermoelectric properties of p-type chalcogenides can be effectively improved by band convergence and hierarchical structure based on a high-entropy-stabilized matrix. The band convergence is due to the decreased light and heavy band energy offsets by alloying Cd for an enhanced Seebeck coefficient and electric transport property. Moreover, the hierarchical structure manipulated by entropy engineering introduces all-scale scattering sources for heat-carrying phonons resulting in a very low lattice thermal conductivity. Consequently, a peak zT of 2.0 at 900 K for p-type chalcogenides and a high experimental conversion efficiency of 12% at ΔT = 506 K for the fabricated segmented modules are achieved. This work provides an entropy strategy to form all-scale hierarchical structures employing high-entropy-stabilized matrix. This work will promote real applications of low-cost thermoelectric materials.


Sign in / Sign up

Export Citation Format

Share Document