Joint Hierarchical Semantic Clipping and Sentence Extraction for Document Summarization

Wanying Yan,Junjun Guo 한국정보처리학회 2020 Journal of information processing systems Vol.16 No.4

Extractive document summarization aims to select a few sentences while preserving its main information on a given document, but the current extractive methods do not consider the sentenceinformation repeat problem especially for news document summarization. In view of the importance and redundancy of news text information,in this paper, we propose a neural extractive summarization approach with joint sentence semantic clipping and selection, which can effectively solve the problem of news text summary sentence repetition. Specifically, a hierarchical selective encoding network is constructed for both sentencelevel and documentlevel document representations, and data containing important information is extracted on news text; a sentence extractor strategy is then adopted for joint scoring and redundant information clipping. This way, our model strikes a balance between important information extraction and redundant information filtering. Experimentalresults on both CNN/Daily Mail dataset and Court Public Opinion News dataset we built are presented to show the effectiveness of our proposed approach in terms of ROUGE metrics, especially for redundant information filtering.

Exploring Resident’s Daily Activity-Travel Behavior: Activity Pattern, Duration and Competition

Wanying Li,Hongzhi Guan,Yan Han,Haiyan Zhu,Pengfei Zhao 대한토목학회 2021 KSCE Journal of Civil Engineering Vol.25 No.8

This paper aims to deeply analyze the regularity of residents' daily activity-travel behaviors to help traffic management departments predict travel demand and evaluate transportation policies. Taking time use data in American as empirical analysis, the influencing factors such as residents’ activity pattern, duration and competition were discussed and analyzed based on the competing risk model. Results can be concluded that: a) people feel happy and meaningful during most of the activities, while they feel sad or pain in relatively rare occasions; b) there is no significant difference between young people and the elderly as for the total number of activities during one day; c) estimation results show that personal, household and activity characteristics have significant influence on activity duration and pattern, and the pattern of last activity (N) has a great influence on the occurrence of next activity (N+1); d) the competition between residents' daily activities is confirmed based on the competing risk model, and travel is closely related to various activities and can be seen as the derived demand of various activities. In all combinations of activity-travel chain schedule, personal - household (30.0%), household - recreation (25.9%), work - travel (45.2%), purchase - travel (33.7%), recreation - travel (22.5%), volunteer - travel (55.3%) and travel - purchase (28.4%) have the highest proportion respectively.

Short-Term Holiday Travel Demand Prediction for Urban Tour Transportation: A Combined Model Based on STC-LSTM Deep Learning Approach

Wanying Li,Hongzhi Guan,Yan Han,Haiyan Zhu,Ange Wang 대한토목학회 2022 KSCE Journal of Civil Engineering Vol.26 No.9

Short-term prediction of holiday travel demand is a complex but key issue to the planning and management of tour transportation system in big cities. This paper develops an improved spatial and temporal correlation long short-term memory model (STC-LSTM) to forecast short-term holiday travel demand based on deep learning approach. Analysis results show six kinds of tourist flow correlations appears in different sets of tourist attractions, and 27.94 percent of the tourist attractions have mid- or high- positive tourist flow correlation with others, meaning that a positive synchronization mechanism exists between suburban tourist attractions in Beijing. The proposed model predicts the holiday travel demand on the basis of the historical data of the spatial and temporal related tourist flows, and the auxiliary data including meteorological data, temporal data, and Internet search index. Based on actual case study with tourist flow data of the suburban tourist attraction in Beijing, the proposed STC-LSTM is carried out to compared with other conventional prediction approaches. Results show that the proposed approach can improve the prediction accuracy and well capture the different spatial and temporal correlations of tourist flows.

Experimental and numerical study on shear studs connecting steel girder and precast concrete deck

Shengfang Qiao,Huqing Liang,Jia-Bao Yan,Wanying Wang 국제구조공학회 2019 Structural Engineering and Mechanics, An Int'l Jou Vol.71 No.4

Shear studs are often used to connect steel girders and concrete deck to form a composite bridge system. The application of precast concrete deck to steel-concrete composite bridges can improve the strength of decks and reduce the shrinkage and creep effect on the long-term behavior of structures. How to ensure the connection between steel girders and concrete deck directly influences the composite behavior between steel girder and precast concrete deck as well as the behavior of the structure system. Compared with traditional multi-I girder systems, a twin-I girder composite bridge system is more simplified but may lead to additional requirements on the shear studs connecting steel girders and decks due to the larger girder spacing. Up to date, only very limited quantity of researches has been conducted regarding the behavior of shear studs on twin-I girder bridge systems. One convenient way for steel composite bridge system is to cast concrete deck in place with shear studs uniformly-distributed along the span direction. For steel composite bridge system using precast concrete deck, voids are included in the precast concrete deck segments, and they are casted with cast-in-place concrete after the concrete segments are erected. In this paper, several sets of push-out tests are conducted, which are used to investigate the heavier of shear studs within the voids in the precast concrete deck. The test data are analyzed and compared with those from finite element models. A simplified shear stud model is proposed using a beam element instead of solid elements. It is used in the finite element model analyses of the twin-I girder composite bridge system to relieve the computational efforts of the shear studs. Additionally, a parametric study is developed to find the effects of void size, void spacing, and shear stud diameter and spacing. Finally, the recommendations are given for the design of precast deck using void for twin I-girder bridge systems.

FNC inhibits non-small cell lung cancer by activating the mitochondrial apoptosis pathway

Jing Xiang,Niu Shuai,Liang Yi,Chen Huiping,Wang Ning,Peng Youmei,Ma Fang,Yue Wanying,Wang Qingduan,Chang Junbiao,Zhang Yi,Zhang Yan 한국유전학회 2022 Genes & Genomics Vol.44 No.1

Background: Previously, we published that 4'-azid-2'-deoxy-2'-fluorarabinoside (FNC), a novel cytosine nucleoside analog, has good anti-viral and anti-tumor activity. Objective: This study aimed to further explore the role and molecular mechanism of FNC in non-small cell lung cancer (NSCLC). Methods: FNC was tested in the NSCLC H460 cell line, the Lewis mouse model, and the H460 cell xenograft model. The effects of FNC were assessed by cell viability, transwell migration, and wound scratch analyses of cell migration and invasion. Apoptosis was assessed by flow cytometry. Proteins expression was assessed by western blot and immunohistochemistry staining (IHC). Results: FNC inhibits the proliferation and metastasis of H460 cells in a time- and dose-dependent manner. FNC treatment showed efficacy and low toxicity in the Lewis mouse lung cancer model as well as in the H460 cell xenograft model. Further, FNC induced H460 cell apoptosis through the activation of the mitochondrial pathway. Notably, FNC inhibited invasion by increasing E-cadherin protein and reducing the protein expression of VEGF, MMP-2, MMP-9, and CD31. Conclusion: FNC inhibits NSCLC by activating the mitochondrial apoptosis pathway and regulating the expressions of multiple proteins related to cell adhesion and invasion, highlighting its potential as an NSCLC therapeutic.