Nowadays, recommender system has become one of the main tools to search for users' interested papers. Since one paper often contains only a part of keywords that a user is interested in, recommender system returns a set of papers that satisfy the user's need of keywords. Besides, to satisfy the users' requirements of further research on a certain domain, the recommended papers must be correlated. However, each paper of an existing paper citation network hardly has cited relationships with others, so the correlated links among papers are very sparse. In addition, while a mass of research approaches have been put forward in terms of link prediction to address the network sparsity problems, these approaches have no relationship with the effect of self-citations and the potential correlations among papers (i.e., these correlated relationships are not included in the paper citation network as their published time is close). Therefore, we propose a link prediction approach that combines time, keywords, and authors' information and optimizes the existing paper citation network. Finally, a number of experiments are performed on the real-world Hep-Th datasets. The experimental results demonstrate the feasibility of our proposal and achieve good performance.

Lei, MD have contributed equally to this work.Word count: 3166 All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. AbstractBackground: The clinical outcomes of COVID-19 patients in Hubei and other areas are different. We aim to investigate the epidemiological and clinical characteristics of patient with COVID-19 in Hunan which is adjacent to Hubei. Methods:In this double-center, observational study, we recruited all consecutive patients with laboratory confirmed COVID-19 from January 23 to February 14, 2020 in two designated hospitals in Hunan province, China. Epidemiological and clinical data from patients' electronic medical records were collected and compared between mild, moderate and severe/critical group in detail. Clinical outcomes were followed up to February 20, 2020. Findings: 291 patients with COVID-19 were categorized into mild group (10.0%), moderate group (72.8%) and severe/critical group (17.2%). The median age of all patients was 46 years (49.8% were male). 86.6% patients had an indirect exposure history. The proportion of patients that had been to Wuhan in severe/critical group (48.0% vs 17.2%, p=0.006) and moderate group (43.4% vs 17.2%, p=0.007) were higher than mild group. Fever (68.7%), cough (60.5%), and fatigue (31.6%) were common symptoms especially for severe and critical patients. Typical lung imaging finding were bilateral and unilateral ground glass opacity or consolidation. Leukopenia, lymphopenia and eosinopenia occurred in 36.1%, 22.7% and 50.2% patients respectively. Increased fibrinogen was detected in 45 of 58 (77.6%) patients with available results. 29 of 44 (65.9%) or 22 of 40 (55.0%) patients were positive in Mycoplasma pneumonia or Chlamydia pneumonia antibody test respectively.Compared with mild or moderate group, severe/critical group had a relative higher level of neutrophil, All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.