Search

Online reviews significantly impact consumers' decision-making process and firms' economic outcomes and are widely seen as crucial to the success of online markets. Firms, therefore, have a strong incentive to manipulate ratings using fake reviews. This presents a problem that academic researchers have tried to solve for over two decades and on which platforms expend a large amount of resources. Nevertheless, the prevalence of fake reviews is arguably higher than ever. To combat this, we collect a dataset of reviews for thousands of Amazon products and develop a general and highly accurate method for detecting fake reviews. A unique difference between previous datasets and ours is that we directly observe which sellers buy fake reviews. Thus, while prior research has trained models using laboratory-generated reviews or proxies for fake reviews, we are able to train a model using actual fake reviews. We show that products that buy fake reviews are highly clustered in the product reviewer network. Therefore, features constructed from this network are highly predictive of which products buy fake reviews. We show that our network-based approach is also successful at detecting fake review buyers even without ground truth data, as unsupervised clustering methods can accurately identify fake review buyers by identifying clusters of products that are closely connected in the network. While text or metadata can be manipulated to evade detection, network-based features are more costly to manipulate because these features result directly from the inherent limitations of buying reviews from online review marketplaces, making our detection approach more robust to manipulation.

Background: Use of numerical and spatial language, also known as math talk, is critical to the development of foundational number and spatial skills in early childhood. However, caregivers and children of low socioeconomic status (SES) tend to use less math talk than their higher-SES peers. Objective: The current efficacy study tested the hypothesis that quantity of math talk among low-SES caregivers and children is increased via a caregiver education curriculum aimed at improving caregivers' language input to children. Methods: Caregiver-child dyads (n = 37; children aged 17 to 36 months) participated in either the language input or a control intervention. Math talk (operationalized as number and spatial word tokens) was coded from video recordings of each dyad engaging in free play at three time points: baseline, post-intervention, and follow-up. Results: The language input curriculum significantly increased caregivers' amount of spatial talk and children's amount of number and spatial talk for up to 4 months after the intervention. Conclusions A caregiver education intervention increased caregivers' use of math talk, which resulted in higher math talk usage by their children. Further verification is needed through an adequately powered longitudinal randomized controlled trial.

The Influenza Research Database (IRD) is a U.S. National Institute of Allergy and Infectious Diseases (NIAID)-sponsored Bioinformatics Resource Center dedicated to providing bioinformatics support for influenza virus research. IRD facilitates the research and development of vaccines, diagnostics and therapeutics against influenza virus by providing a comprehensive collection of influenza-related data integrated from various sources, a growing suite of analysis and visualization tools for data mining and hypothesis generation, personal workbench spaces for data storage and sharing, and active user community support. Here, we describe the recent improvements in IRD including the use of cloud and high performance computing resources, analysis and visualization of user-provided sequence data with associated metadata, predictions of novel variant proteins, annotations of phenotype-associated sequence markers and their predicted phenotypic effects, hemagglutinin (HA) clade classifications, an automated tool for HA subtype numbering conversion, linkouts to disease event data and the addition of host factor and antiviral drug components. All data and tools are freely available without restriction from the IRD website at https://www.fludb.org.

Motivation Sequence repositories have few well-annotated virus mature peptide sequences. Therefore post-translational proteolytic processing of polyproteins into mature peptides (MPs) has been performed in silico, with a new computational method, for over 200 species in 5 pathogenic virus families (Caliciviridae, Coronaviridae, Flaviviridae, Picornaviridae and Togaviridae). Results Using pairwise alignment with reference sequences, MPs have been annotated and their sequences made available for search, analysis and download. At publication the method had produced 156 216 sequences, a large portion of the protein sequences now available in https://www.viprbrc.org. It represents a new and comprehensive mature peptide collection. Availability and implementation The data are available at the Virus Pathogen Resource https://www.viprbrc.org , and the software at https://github.com/VirusBRC/vipr_mat_peptide. [ABSTRACT FROM AUTHOR]