Browse
Submit Data
Databases
API
Help

Dataset Information

0 Views

0 Connections

0 Citations

0 Reanalyses

0 Downloads

Omics score: 0

Retrofitting Embeddings for Unsupervised User Identity Linkage

ABSTRACT: User Identity Linkage (UIL) is the problem of matching user identities across multiple online social networks (OSNs) which belong to the same person. The solutions to UIL problem facilitate cross-platform research on OSN users and enable many useful applications such as user profiling and recommendation. As the UIL labeled data are often lacking and costly to obtain, learning user embeddings for matching user identities using an unsupervised approach is therefore highly desired. In this paper, we propose a novel unsupervised UIL framework for enhancing existing user embedding-based UIL methods. Our proposed framework incorporates two key ideas, user-discriminative features and retrofitting embedding. The user-discriminative features enable us to differentiate a specific user identity from other users in its OSN. From the user-discriminative features, we derive pairs of similar user identities across OSNs for retrofitting the base user embeddings of existing UIL methods. Through extensive experiments on three real-world OSN datasets, we show that our framework can leverage user-discriminative features to improve the accuracy of different user embedding-based UIL methods significantly. The quantum of improvement can also be surprisingly good even for existing UIL methods with very poor matching accuracy. Electronic supplementary material The online version of this chapter (10.1007/978-3-030-47426-3_30) contains supplementary material, which is available to authorized users.

SUBMITTER: Lauw H

PROVIDER: S-EPMC7206306 | biostudies-literature | 2020 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Json Xml

Similar Datasets

Unsupervised protein embeddings outperform hand-crafted sequence and structure features at predicting molecular function.

Project description:MotivationProtein function prediction is a difficult bioinformatics problem. Many recent methods use deep neural networks to learn complex sequence representations and predict function from these. Deep supervised models require a lot of labeled training data which are not available for this task. However, a very large amount of protein sequences without functional labels is available.ResultsWe applied an existing deep sequence model that had been pretrained in an unsupervised setting on the supervised task of protein molecular function prediction. We found that this complex feature representation is effective for this task, outperforming hand-crafted features such as one-hot encoding of amino acids, k-mer counts, secondary structure and backbone angles. Also, it partly negates the need for complex prediction models, as a two-layer perceptron was enough to achieve competitive performance in the third Critical Assessment of Functional Annotation benchmark. We also show that combining this sequence representation with protein 3D structure information does not lead to performance improvement, hinting that 3D structure is also potentially learned during the unsupervised pretraining.Availability and implementationImplementations of all used models can be found at https://github.com/stamakro/GCN-for-Structure-and-Function.Supplementary informationSupplementary data are available at Bioinformatics online.

| S-EPMC8055213 | biostudies-literature

Characterizing the macrostructure of electronic health record work using raw audit logs: an unsupervised action embeddings approach.

Project description:Raw audit logs provide a comprehensive record of clinicians' activities on an electronic health record (EHR) and have considerable potential for studying clinician behaviors. However, research using raw audit logs is limited because they lack context for clinical tasks, leading to difficulties in interpretation. We describe a novel unsupervised approach using the comparison and visualization of EHR action embeddings to learn context and structure from raw audit log activities. Using a dataset of 15 767 634 raw audit log actions performed by 88 intern physicians over 6 months of EHR use across inpatient and outpatient settings, we demonstrated that embeddings can be used to learn the situated context for EHR-based work activities, identify discrete clinical workflows, and discern activities typically performed across diverse contexts. Our approach represents an important methodological advance in raw audit log research, facilitating the future development of metrics and predictive models to measure clinician behaviors at the macroscale.

| S-EPMC9933072 | biostudies-literature

Retrofitting metal-organic frameworks.

Project description:The post-synthetic installation of linker molecules between open-metal sites (OMSs) and undercoordinated metal-nodes in a metal-organic framework (MOF) - retrofitting - has recently been discovered as a powerful tool to manipulate macroscopic properties such as the mechanical robustness and the thermal expansion behavior. So far, the choice of cross linkers (CLs) that are used in retrofitting experiments is based on qualitative considerations. Here, we present a low-cost computational framework that provides experimentalists with a tool for evaluating various CLs for retrofitting a given MOF system with OMSs. After applying our approach to the prototypical system CL@Cu3BTC2 (BTC?=?1,3,5-benzentricarboxylate) the methodology was expanded to NOTT-100 and NOTT-101 MOFs, identifying several promising CLs for future CL@NOTT-100 and CL@NOTT-101 retrofitting experiments. The developed model is easily adaptable to other MOFs with OMSs and is set-up to be used by experimentalists, providing a guideline for the synthesis of new retrofitted MOFs with modified physicochemical properties.

| S-EPMC6820732 | biostudies-literature

Peer-to-Peer User Identity Verification Time Optimization in IoT Blockchain Network

Project description:Blockchain introduces challenges related to the reliability of user identity and identity management systems; this includes detecting unfalsified identities linked to IoT applications. This study focuses on optimizing user identity verification time by employing an efficient encryption algorithm for the user signature in a peer-to-peer decentralized IoT blockchain network. To achieve this, a user signature-based identity management framework is examined by using various encryption techniques and contrasting various hash functions built on top of the Modified Merkle Hash Tree (MMHT) data structure algorithm. The paper presents the execution of varying dataset sizes based on transactions between nodes to test the scalability of the proposed design for secure blockchain communication. The results show that the MMHT data structure algorithm using SHA3 and AES-128 encryption algorithm gives the lowest execution time, offering a minimum of 36% gain in time optimization compared to other algorithms. This work shows that using the AES-128 encryption algorithm with the MMHT algorithm and SHA3 hash function not only identifies malicious codes but also improves user integrity check performance in a blockchain network, while ensuring network scalability. Therefore, this study presents the performance evaluation of a blockchain network considering its distinct types, properties, components, and algorithms’ taxonomy.

| S-EPMC9961683 | biostudies-literature

Polypropylene as a Retrofitting Material for Shear Walls.

Project description:In recent years, on account of their excellent mechanical properties, composite materials (made of epoxy-bonded carbon, glass, or aramid fibers) have been used to reinforce masonry walls against in-plane actions. These materials have proven to be an effective solution for the strengthening of unreinforced masonry (URM) walls. Lately, research has shifted to the study of different types of fibers to avoid the use of epoxy adhesives, whose long-term behavior and compatibility with masonry are poor. This paper describes an experimental program that investigated the behavior of URM shear walls strengthened with two types of commercially available polypropylene products: short fibers (fiber length = 12 mm) and polypropylene nets. This investigation aimed to evaluate the influence of polypropylene reinforcement, embedded into an inorganic matrix, in terms of the improvement of the lateral load-carrying capacity, failure mechanism, ductility, and energy dissipation capacity of URM wall panels, where nine walls were subjected to in-plane loads using a racking test setup. The study showed that using two layers of polypropylene fibers embedded into a cementitious matrix greatly increased the in-plane load capacity of the brickwork masonry. On the other hand, the test results indicated that polypropylene nets, used as a repair method for cracked shear walls, cannot improve the structural performance of the walls.

| S-EPMC7321489 | biostudies-literature

Unsupervised low-dimensional vector representations for words, phrases and text that are transparent, scalable, and produce similarity metrics that are not redundant with neural embeddings.

Project description:Neural embeddings are a popular set of methods for representing words, phrases or text as a low dimensional vector (typically 50-500 dimensions). However, it is difficult to interpret these dimensions in a meaningful manner, and creating neural embeddings requires extensive training and tuning of multiple parameters and hyperparameters. We present here a simple unsupervised method for representing words, phrases or text as a low dimensional vector, in which the meaning and relative importance of dimensions is transparent to inspection. We have created a near-comprehensive vector representation of words, and selected bigrams, trigrams and abbreviations, using the set of titles and abstracts in PubMed as a corpus. This vector is used to create several novel implicit word-word and text-text similarity metrics. The implicit word-word similarity metrics correlate well with human judgement of word pair similarity and relatedness, and outperform or equal all other reported methods on a variety of biomedical benchmarks, including several implementations of neural embeddings trained on PubMed corpora. Our implicit word-word metrics capture different aspects of word-word relatedness than word2vec-based metrics and are only partially correlated (rho?=?0.5-0.8 depending on task and corpus). The vector representations of words, bigrams, trigrams, abbreviations, and PubMed title?+?abstracts are all publicly available from http://arrowsmith.psych.uic.edu/arrowsmith_uic/word_similarity_metrics.html for release under CC-BY-NC license. Several public web query interfaces are also available at the same site, including one which allows the user to specify a given word and view its most closely related terms according to direct co-occurrence as well as different implicit similarity metrics.

| S-EPMC6557457 | biostudies-literature

Seismic retrofitting of Maori wharenui in Aotearoa New Zealand.

Project description:The design of M?ori wharenui (meeting houses), refined over the last 180 years, has given rise to a distinctive typology that demonstrates an understanding of tectonic-narrative expression. It is a visual codex designed to be read. Set within the context of pre-disaster readiness, this current study investigated the resilience of M?ori buildings and communities through a dual interpretation of seismic, regulatory and historical frameworks. The literature on M?ori building technologies scarcely addresses matters of resilience and there is currently no formalized methodology to guide wharenui seismic retrofit practice. Through a macro level analysis of recently completed retrofit projects, this current study identified issues present in current practice. The resulting research presented a pathway for the future development of an approach (to inform methodologies) and strategy (to build capability) founded on the recognition of wharenui as a new category of building within building regulations. It is also recommended that research continues to develop alternative engineered solutions, expert knowledge bases and best practice guidelines. This article is part of the theme issue 'Environmental loading of heritage structures'.

| S-EPMC6710465 | biostudies-literature

Digital Identity: The effect of trust and reputation information on user judgement in the Sharing Economy.

Project description:The Sharing Economy (SE) is a growing ecosystem focusing on peer-to-peer enterprise. In the SE the information available to assist individuals (users) in making decisions focuses predominantly on community-generated trust and reputation information. However, how such information impacts user judgement is still being understood. To explore such effects, we constructed an artificial SE accommodation platform where we varied the elements related to hosts' digital identity, measuring users' perceptions and decisions to interact. Across three studies, we find that trust and reputation information increases not only the users' perceived trustworthiness, credibility, and sociability of hosts, but also the propensity to rent a private room in their home. This effect is seen when providing users both with complete profiles and profiles with partial user-selected information. Closer investigations reveal that three elements relating to the host's digital identity are sufficient to produce such positive perceptions and increased rental decisions, regardless of which three elements are presented. Our findings have relevant implications for human judgment and privacy in the SE, and question its current culture of ever increasing information-sharing.

| S-EPMC6292641 | biostudies-literature

Molecular Retrofitting Adapts a Metal-Organic Framework to Extreme Pressure.

Project description:Despite numerous studies on chemical and thermal stability of metal-organic frameworks (MOFs), mechanical stability remains largely undeveloped. To date, no strategy exists to control the mechanical deformation of MOFs under ultrahigh pressure. Here, we show that the mechanically unstable MOF-520 can be retrofitted by precise placement of a rigid 4,4'-biphenyldicarboxylate (BPDC) linker as a "girder" to afford a mechanically robust framework: MOF-520-BPDC. This retrofitting alters how the structure deforms under ultrahigh pressure and thus leads to a drastic enhancement of its mechanical robustness. While in the parent MOF-520 the pressure transmitting medium molecules diffuse into the pore and expand the structure from the inside upon compression, the girder in the new retrofitted MOF-520-BPDC prevents the framework from expansion by linking two adjacent secondary building units together. As a result, the modified MOF is stable under hydrostatic compression in a diamond-anvil cell up to 5.5 gigapascal. The increased mechanical stability of MOF-520-BPDC prohibits the typical amorphization observed for MOFs in this pressure range. Direct correlation between the orientation of these girders within the framework and its linear strain was estimated, providing new insights for the design of MOFs with optimized mechanical properties.

| S-EPMC5492252 | biostudies-literature

Multipoint approximations of identity-by-descent probabilities for accurate linkage analysis of distantly related individuals.

Project description:We propose an analytical approximation method for the estimation of multipoint identity by descent (IBD) probabilities in pedigrees containing a moderate number of distantly related individuals. We show that in large pedigrees where cases are related through untyped ancestors only, it is possible to formulate the hidden Markov model of the Lander-Green algorithm in terms of the IBD configurations of the cases. We use a first-order Markov approximation to model the changes in this IBD-configuration variable along the chromosome. In simulated and real data sets, we demonstrate that estimates of parametric and nonparametric linkage statistics based on the first-order Markov approximation are accurate. The computation time is exponential in the number of cases instead of in the number of meioses separating the cases. We have implemented our approach in the computer program ALADIN (accurate linkage analysis of distantly related individuals). ALADIN can be applied to general pedigrees and marker types and has the ability to model marker-marker linkage disequilibrium with a clustered-markers approach. Using ALADIN is straightforward: It requires no parameters to be specified and accepts standard input files.

| S-EPMC2427226 | biostudies-literature

OmicsDI is part of the ELIXIR infrastructure

OmicsDI is an Elixir interoperability service. Learn more ›

Tweets

OmicsDI Databases

PRIDE
PeptideAtlas
MassIVE
JPOST Repository
Physiome Model Repository

EGA
EVA
ENA
LINCS
PAXDB
Cell Collective

MetaboLights
Metabolomics Workbench
MetabolomeExpress
GNPS
BioModels
FAIRDOMHub

ArrayExpress
dbGaP
ExpressionAtlas
GEO
NODE

Information

Databases
Help
API
Contact us
Code on GitHub
Terms of use
Submit Data