A5114949071- Information Retrieval and Search Behavior
- Web Data Mining and Analysis
- Mobile Crowdsensing and Crowdsourcing
- Complex Network Analysis Techniques
- Advanced Text Analysis Techniques
- Pulmonary Hypertension Research and Treatments
- Fetal and Pediatric Neurological Disorders
- Prenatal Screening and Diagnostics
- Digital Marketing and Social Media
- Expert finding and Q&A systems
- Semantic Web and Ontologies
- Data Visualization and Analytics
- Recommender Systems and Techniques
- Cardiovascular Issues in Pregnancy
- Digital Imaging for Blood Diseases
- Musculoskeletal pain and rehabilitation
- Heme Oxygenase-1 and Carbon Monoxide
- Digital Communication and Language
- Experimental Behavioral Economics Studies
- Personal Information Management and User Behavior
- Spam and Phishing Detection
- Advanced Database Systems and Queries
- Advanced Machining and Optimization Techniques
- Big Data and Business Intelligence
- Advanced Optical Imaging Technologies
Jiaying University
2025
Xiamen University
2014-2024
Guangzhou Medical University
2013-2024
Sichuan University
2024
State Key Laboratory of Biotherapy
2024
State Key Laboratory of Respiratory Disease
2013-2022
First Affiliated Hospital of Guangzhou Medical University
2022
University of California, San Diego
2016-2018
LinkedIn (United States)
2013-2018
Google (United States)
2009
A self-supervised model based on CycleGAN is proposed to reconstruct the 3D particle fields from unpaired data sets. The ground truth can be obtained by converting information of field into a 2D grayscale image. hologram particles input network model. training designed. not only alleviates pressure demand but also simplifies design holographic device. Thus, proceed with incomplete data. With cycle consistency mode, characterization more accurate. In order assess viability scheme, structural...
In this paper we investigate the effectiveness of a document-independent technique for eliciting feedback from users about their information problems. We propose that such can be used to elicit terms use in query expansion and as follow-up when ambiguous queries are initially posed by users. design form obtain additional users, administer after initial querying, create series experimental runs based on obtained form. Results demonstrate was successful at more significantly improved retrieval...
Term relevance feedback has had a long history in information retrieval. However, research on interactive term yielded mixed results. In this paper, we investigate several aspects related to the elicitation of feedback: display document surrogates, technique for identifying or selecting terms, and sources expansion terms. We conduct between subjects experiment (n=61) three interfaces using 2005 TREC HARD collection, evaluate each interface with respect query length retrieval performance....
User engagement in search refers to the frequency for users (re-)using engine accomplish their tasks. Among factors that affected users' visit frequency, relevance of results is believed play a pivotal role. While multiple work past has demonstrated correlation between success and user based on longitudinal analysis, we examine this problem from different perspective work. Specifically, carefully designed large-scale controlled experiment large commercial Web engine, which were separated...
Abstract Background Patients with pulmonary hypertension (PH) and chronic obstructive disease (COPD) have an increased risk of exacerbation decreased survival. We aimed to develop validate a non‐invasive nomogram for predicting COPD associated severe PH prognostic patients concurrent (COPD–PH). Methods This study included 535 COPD–PH from six hospitals. A multivariate logistic regression analysis was used analyse the factors in Cox COPD–PH. Performance assessed using calibration, area under...
Copy Number Variations (CNVs) are an important genetic cause of a number neurodevelopmental disorders (NDs). However, the association between CNVs and development prognosis fetal isolated mild ventriculomegaly (IMV) is unclear.To investigate possible associations IMV.This retrospective study recruited 154 subjects with ultrasound-confirmed IMV 190 in control cohort who underwent high-risk prenatal serum screening program. The exclusion criteria included fetus G-banding chromosomal...
Abstract This exploratory study investigates the needs Web users have to make annotations for their personal use when they view pages. Three forms of observed on printed documents ‐ text selection and emphasis, association building, document re‐segmentation are examined in environment. The shows that building through notes or symbols remain dominant annotation Web, while structural (re‐segmentation) layout (change font, color, etc.) also prevalent. users' preferences tools designed...
Abstract Motivation Circular RNAs (circRNAs) play important roles in gene expression and their involvement tumorigenesis is emerging. circRNA-related database a powerful tool for researchers to investigate circRNAs. However, existing databases lack advanced platform integrating comprehensive information analysis tools of cancer-related Results We developed called CircRNA Cancer Database (C2CDB), encompassing 318 158 circRNAs expressed tumors adjacent tissues across 30 types cancers. C2CDB...
The study investigates the relationship between types of behavior that can be captured from Web searches and searchers' interests. search cases which involve underspecification information needs at beginning modification strategies during process will collected examined by human analysts. focuses on identifying rules used analysts to infer searcher These put into algorithms as basis for systems provide query suggestions or implicitly reformulate continues work.
"You make what you measure" is a familiar mantra at data-driven companies. Accordingly, companies must be careful to choose North Star metrics that create better product. Metrics fall into two general categories: direct count such as total revenue and monthly active users, nuanced quality regarding value or other aspects of the user experience. Count metrics, when used exclusively Star, might inform product decisions harm Therefore, play an important role in development. We present five-step...
Pregnant women with pulmonary hypertension (PH) have higher mortality rates and poor foetal/neonatal outcomes. Tools to assess these risk factors are not well established.Predictive prognostic nomograms were constructed using data from a "Development" cohort of 420 pregnant patients PH, recorded between January 2009 December 2018. Logistic regression analysis established models predict the probability adverse maternal events overall survival by Cox analysis. An independent "Validation"...
We describe a preliminary analysis of queries created by 81 users for 4 topics from the TREC Robust Track. Our goal was to explore potential benefits using multiple on retrieval performance difficult topics. first examine overlap in users' and results with respect different same topic. then combining various ways. provide some evidence that having access can improve individual searchers
"Well begun is half done". This proverb especially true for a web product when it comes to creating delightful and proactive user experience. article describes our work in the last few years optimizing new experience at LinkedIn, driven by application of data science advanced analytical methods. Through mining logs generated users past, we uncovered signals from their initial session that can predict retention. We established difference between models predictions inform strategy. found...
Abstract Numerous studies have demonstrated that annotation is an important part of human reading behavior in both printed and electronic environments. Annotation the environment requires special support due to limited media affordances. We witnessed continuous improvement functions some environments, such as text documents Microsoft Word or Adobe Acrobat images Flickr. However, comparatively little research has been conducted understand people's needs for making annotations when they watch...