User's Sequence Research Articles

Predicting Temporal Sets with Simplified Fully Connected Networks

Given a sequence of sets, where each set contains an arbitrary number of elements, temporal sets prediction aims to predict which elements will appear in the subsequent set. Existing methods for temporal sets prediction are developed on sophisticated components (e.g., recurrent neural networks, attention or gating mechanisms, and graph neural networks), which inevitably increase the model complexity due to more trainable parameters and higher computational costs. Moreover, the involved nonlinear activation may contribute little or even degrade the performance. In this paper, we present a succinct architecture that is solely built on the Simplified Fully Connected Networks (SFCNs) for temporal sets prediction to bring both effectiveness and efficiency together. In particular, given a user's sequence of sets, we employ SFCNs to derive representations of the user by learning inter-set temporal dependencies, intra-set element relationships, and intra-embedding channel correlations. Two families of general functions are introduced to preserve the permutation-invariant property of each set and the permutation-equivariant property of elements in each set. Moreover, we design a user representations adaptive fusing module to aggregate user representations according to each element for improving the prediction performance. Experiments on four benchmarks show the superiority of our approach over the state-of-the-art under both transductive and inductive settings. We also theoretically and empirically demonstrate that our model has lower space and time complexity than baselines. Codes and datasets are available at https://github.com/yule-BUAA/SFCNTSP.

E-Commerce Recommendation Algorithm based on Graph Neural Network

While recommender system is becoming an increasingly essential component in e-commerce websites, although previous models, which directly calculate the similarity of user/item history record, has obtained evidence of effectiveness, recommendations based solely on users' current sequence of actions, when user identity and history preference are not present, has been a popular area due to the growing privacy concerns. This paper demonstrates a model using a graph neural network, which takes the user's sequence of purchasing events as input and constructs a graph derived from it, to make the prediction of the most likely subsequent product that the customer may purchase and make personalized recommendations by the combination of session preference and user’s current interest. Experiments on a real-world e-commerce purchasing event dataset and analysis are carried out to test the model’s performance, as well as how the length of sequence may affect the model preference. The result shows that the model performance has attained a local peak on the dataset used.

Feature-Level Attentive Neural Model for Session-Based Recommendation

The main goal of session-based recommendation is to predict a user's next click based on historical anonymous session data. One important aspect is capturing the interest drift that occurs in a user's click sequences. Recent studies have mainly exploited the attention mechanism to extract the main intentions of users and reduce the influence of unintentional actions on the performance of recommendation systems. However, the previous works attempt to extract a user's interests at the item level; they ignore the inherent correlations between a user's interest drift. Therefore, we propose a novel Feature-level Attentive Neural Model (FANM) as a solution capable of capturing a user's current interests by considering interest drift at the feature level. Our model exploits a gated recurrent unit (GRU) and a multihead attention mechanism to extract a user's current interests at the feature level from click sequences; then, it integrates the user's long-term and short-term interests with their last actions (e.g., clicks) to predict the next action. Our proposed model effectively captures a user's interest drift by performing sufficient modeling of a user's sequence data, resulting in increased recommendation accuracy. The experiments on the two real-world datasets show that FANM performs significantly better than baseline methods.

Authenticity and credibility aware detection of adverse drug events from social media

ObjectivesAdverse drug events (ADEs) are among the top causes of hospitalization and death. Social media is a promising open data source for the timely detection of potential ADEs. In this paper, we study the problem of detecting signals of ADEs from social media. MethodsDetecting ADEs whose drug and AE may be reported in different posts of a user leads to major concerns regarding the content authenticity and user credibility, which have not been addressed in previous studies. Content authenticity concerns whether a post mentions drugs or adverse events that are actually consumed or experienced by the writer. User credibility indicates the degree to which chronological evidence from a user's sequence of posts should be trusted in the ADE detection. We propose AC-SPASM, a Bayesian model for the authenticity and credibility aware detection of ADEs from social media. The model exploits the interaction between content authenticity, user credibility and ADE signal quality. In particular, we argue that the credibility of a user correlates with the user's consistency in reporting authentic content. ResultsWe conduct experiments on a real-world Twitter dataset containing 1.2 million posts from 13,178 users. Our benchmark set contains 22 drugs and 8089 AEs. AC-SPASM recognizes authentic posts with F1 – the harmonic mean of precision and recall of 80%, and estimates user credibility with precision@10 = 90% and NDCG@10 – a measure for top-10 ranking quality of 96%. Upon validation against known ADEs, AC-SPASM achieves F1 = 91%, outperforming state-of-the-art baseline models by 32% (p < 0.05). Also, AC-SPASM obtains precision@456 = 73% and NDCG@456 = 94% in detecting and prioritizing unknown potential ADE signals for further investigation. Furthermore, the results show that AC-SPASM is scalable to large datasets. ConclusionsOur study demonstrates that taking into account the content authenticity and user credibility improves the detection of ADEs from social media. Our work generates hypotheses to reduce experts’ guesswork in identifying unknown potential ADEs.

Authenticity and credibility aware detection of adverse drug events from social media

ObjectivesAdverse drug events (ADEs) are among the top causes of hospitalization and death. Social media is a promising open data source for the timely detection of potential ADEs. In this paper, we study the problem of detecting signals of ADEs from social media. MethodsDetecting ADEs whose drug and AE may be reported in different posts of a user leads to major concerns regarding the content authenticity and user credibility, which have not been addressed in previous studies. Content authenticity concerns whether a post mentions drugs or adverse events that are actually consumed or experienced by the writer. User credibility indicates the degree to which chronological evidence from a user's sequence of posts should be trusted in the ADE detection. We propose AC-SPASM, a Bayesian model for the authenticity and credibility aware detection of ADEs from social media. The model exploits the interaction between content authenticity, user credibility and ADE signal quality. In particular, we argue that the credibility of a user correlates with the user's consistency in reporting authentic content. ResultsWe conduct experiments on a real-world Twitter dataset containing 1.2 million posts from 13,178 users. Our benchmark set contains 22 drugs and 8089 AEs. AC-SPASM recognizes authentic posts with F1 – the harmonic mean of precision and recall of 80%, and estimates user credibility with precision@10 = 90% and NDCG@10 – a measure for top-10 ranking quality of 96%. Upon validation against known ADEs, AC-SPASM achieves F1 = 91%, outperforming state-of-the-art baseline models by 32% (p < 0.05). Also, AC-SPASM obtains precision@456 = 73% and NDCG@456 = 94% in detecting and prioritizing unknown potential ADE signals for further investigation. Furthermore, the results show that AC-SPASM is scalable to large datasets. ConclusionsOur study demonstrates that taking into account the content authenticity and user credibility improves the detection of ADEs from social media. Our work generates hypotheses to reduce experts’ guesswork in identifying unknown potential ADEs.

Synthetic Gene Design Using Codon Optimization On-Line (COOL).

Codon optimization has been widely used for designing native or synthetic genes to enhance their expression in heterologous host organisms. We recently developed Codon Optimization On-Line (COOL) which is a web-based tool to provide multi-objective codon optimization functionality for synthetic gene design. COOL provides a simple and flexible interface for customizing codon optimization based on several design parameters such as individual codon usage, codon pairing, and codon adaptation index. User-defined sequences can also be compared against the COOL optimized ones to show the extent by which the user's sequences can be evaluated and further improved. The utility of COOL is demonstrated via a case study where the codon optimized sequence of an invertase enzyme is generated for the enhanced expression in E. coli.

MultiSyn: A Webtool for Multiple Synteny Detection and Visualization of User's Sequence of Interest Compared to Public Plant Species.

Information on multiple synteny between plants and/or within a plant is key information to understand genome evolution. In addition, visualization of multiple synteny is helpful in interpreting evolution. So far, some web applications have been developed to determine and visualize multiple homology regions at once. However, the applications are not fully convenient for biologists because some of them do not include the function of synteny determination but visualize the multiple synteny plots by allowing users to upload their synteny data by determining the synteny based only on BLAST similarity information, with some algorithms not designed for synteny determination. Here, we introduce a web application that determines and visualizes multiple synteny from two types of files, simplified browser extensible data and protein sequence file by MCScanX algorithm, which have been used in many synteny studies.

PTree: Mining Sequential Patterns Efficiently in Multiple Data Streams Environment

Although issues of data streams have been widely studied and utilized, it is nevertheless challenging to deal with sequential mining of data streams. In this paper, we assume that the transaction of a user is partially coming and that there is no auxiliary for buffering and integrating. We adopt the Path Tree for mining frequent sequential patterns over data streams and integrate the user's sequences efficiently. Algorithms with regards to accuracy (PAlgorithm) and space (PSAlgorithm) are proposed to meet the different aspects of users, although GAlgorithm for mining frequent sequential patterns with a gap limitation is proposed. Many pruning properties are used to further reduce the space usage and improve the accuracy of our algorithms. We also prove that PAlgorithm mine frequent sequential patterns with the approximate support of error guarantee. Through thoughtful experiments, synthetic and real datasets are utilized to verify the feasibility of our algorithms.

Attribute-Driven Hidden Markov Model Trees for Intention Prediction

In this paper, we introduce a novel approach to generate an intention prediction model of user interactions with systems. As part of this new approach, we include personal aspects, such as user characteristics, that can increase prediction accuracy. The model is automatically trained according to the user's fixed attributes (e.g., demographic data such as age and gender) and the user's sequences of actions in the system. The generated model has a tree structure. The building blocks of each node can be any probabilistic sequence model [such as hidden Markov models (HMMs) and conditional random fields (CRFs)] and each node is split according to user attributes. Thus, we refer to this algorithm as an attribute-driven model tree. The new model was first tested on simulated data in which users with different attributes (such as age and gender) behave differently when trying to accomplish various tasks. We then validated the ability of the algorithm to discover the relevant attributes. We tested our algorithm on two real datasets: from a web application and a mobile application dataset. The results were encouraging and indicate the capability of the proposed method to discover the correct user intention model and increasing intention prediction accuracy compared with single HMM or CRF models.

CGAS: comparative genomic analysis server

Summary: Comparative approach is one of the most essential methods for extracting functional and evolutionary information from genomic sequences. So far, a number of sequence comparison tools have been developed, and most are either for on-site use, requiring program installation but providing a wide variety of analyses, or for the online search of user's sequences against given databases on a server. We newly devised an Asynchronous JavaScript and XML (Ajax)-based system for comparative genomic analyses, CGAS, with highly interactive interface within a browser, requiring no software installation. The current version, CGAS version 1, provides functionality for viewing similarity relationships between user's sequences, including a multiple dot plot between sequences with their annotation information. The scrollbar-less ‘draggable’ interface of CGAS is implemented with Google Maps API version 2. The annotation information associated with the genomic sequences compared is synchronously displayed with the comparison view. The multiple-comparison viewer is one of the unique functionalities of this system to allow the users to compare the differences between different pairs of sequences. In this viewer, the system tells orthologous correspondences between the sequences compared interactively. This web-based tool is platform-independent and will provide biologists having no computational skills with opportunities to analyze their own data without software installation and customization of the computer system.Availability and Implementation: CGAS is available at http://cgas.ist.hokudai.ac.jp/.Contact: watanabe@ist.hokudai.ac.jp

VisualRepbase: an interface for the study of occurrences of transposable element families.

BackgroundRepbase is a reference database of eukaryotic repetitive DNA, which includes prototypic sequences of repeats and basic information described in annotations. Repbase already has software for entering new sequence families and for comparing the user's sequence with the database of consensus sequences.ResultsWe describe the software named VisualRepbase and the associated database, which allow for displaying and analyzing all occurrences of transposable element families present in an annotated genome. VisualRepbase is a Java-based interface which can download selected occurrences of transposable elements, show the distribution of given families on the chromosome, and present the localization of these occurrences with regard to gene annotations and other families of transposable elements in Repbase. In addition, it has several features for saving the graphical representation of occurrences, saving all sequences in FASTA format, and searching and saving all annotated genes that are surrounded by these occurrences.ConclusionVisualRepbase is available as a downloadable version. It can be found at .

Transterm: a database of mRNAs and translational control elements.

Transterm is a database that facilitates studies of translation and the translational control of protein synthesis. It contains a curated collection of elements in mRNAs that control translation, and biologically relevant mRNA regions extracted from GenBank. It is organised largely on a taxonomic basis with files and summaries for each species. Global patterns that may affect translation in particular species, for example bias in the context of initiation codons (Kozak's consensus or Shine-Dalgarno sequences) or termination codons, can be detected in the consensus and information content bias summaries. Several types of access are provided via a web browser interface. Transterm defined elements may be matched in a user's sequence or in the database. Alternatively, elements can be entered by the user to search specific sections of the database (for example, coding regions or 3' flanking regions or the 3'-UTRs) or the user's sequence. Each Transterm defined element has an associated biological description with references. The database is accessible at http://uther.otago.ac.nz/Transterm.html.