User Click Research Articles

Predicting Click Behavior Based on Machine Learning Models

As businesses spend a great amount of advertising dollars each year and hope to achieve marketing objectives, advertising practitioners often advise them to understand customer behavior and customize target audience. This study gains information about click behavior and uses those to help brands make targeting decisions. In the study, a predictive analysis on whether a viewer clicks on the ad was conducted to build a model that forecasts user clicks and forms a list of potentially interested viewers. The predictive model included variables that mainly contained viewer information, such as ‘Daily Time Spent on Site’, ‘Age’, ‘Area Income’, etc., and were chosen based on correlations with each other. After splitting the samples with a 7:3 ratio, machine Learning algorithms, including Logistic Regression, Random Forest, and Support Vector Machine, were used to build the model. Each model was implemented on both datasets, and with no sign of overfitting shown, Support Vector Machine model outperformed other models with a 97% prediction accuracy on training data and a 96% prediction accuracy on testing data. With this model being able to generalize well on new data points, businesses can start implementing it to filter out the audience who will click on an ad and show interests in the brand/product. This deeper understanding of the audience will help improve targeting for campaigns with different objectives and achieve higher advertising efficiency and effectiveness.

Personal or General? A Hybrid Strategy with Multi-factors for News Recommendation

News recommender systems have become an effective manner to help users make decisions by suggesting the potential news that users may click and read, which has shown the proliferation nowadays. Many representative algorithms made great efforts to discover users’ preferences from the histories for triggering news recommendations. However, there exist some limitations due to the following two main issues. First, they mainly rely on the sufficient user data, which cannot well capture users’ temporal interests with very limited records. Second, always perceiving users’ histories for recommendation may ignore some important news (e.g., breaking news). In this article, we propose a novel Multi-factors Fusion model for news recommendation by integrating both user-dependent preference effect and user-independent timeliness effect together. First, to track the preference of a certain user, we decompose her reading history into two user-related factors, including the long-term habit and the short-term interest. Specifically, we extract her persistent habit by exploring the category effect of news that she focuses on from her whole records. Then, we characterize her temporary interests by proposing a recurrent neural network of analyzing the homogeneous relations between her latest clicked news and the candidate ones. Second, to describe the user-independent news timeliness effect, we propose a novel survival analysis model to estimate the instantaneous click probability of a certain news as the occurring probability of an event, where much sensational news tends to be picked out. Last, we fuse all effects to determine the probability of a user clicking on a certain news under the independent event assumption. We conduct extensive experiments on two real-world datasets. Experimental results demonstrate that our model can generate better news recommendations on both general scenario and cold-start scenario.

Mining high-utility sequences with positive and negative values

Sequence pattern discovery is a fundamental topic in the domain of data mining. It has been widely used to solve various problems (e.g., behavior pattern discovery, gene pattern discovery in bioinformatics, user click pattern mining, etc.). High-utility sequence mining as a novel hot issue is more challenging and has generally attracted plenty of attention. Our paper focuses on mining high-utility sequences in a more complicated environment with high efficiency. Most of the previous methods for utility mining aim to find high-utility sequences suitable for items with positive values, but most real-world situations contain items with both positive and negative values. Several algorithms have been applied to the above sophisticated situation and can be used as our comparing algorithms. In this paper, we introduce the FHUSN (Fast mining High Utility Sequences with Negative item) algorithm to mine high-utility sequences in situations with or without negative utility values. FHUSN utilizes the new utility array to store data. Several new pruning strategies that apply to situations with or without negative values have been used to reduce search space. Experiments are carried out on several benchmark datasets, and experimental results illustrate that our method has better performance.

General Purpose Recommendation System

A recommendation system is a class of information retrieval systems that provides personalized suggestions to users. Our approach is to generalize the process of creating a recommendation system so that it can be used effectively in multiple domains. This approach can be achieved by collaborative filtering by using the website’s clickstream analytics. Data related to user clicks can easily be extracted and this data's use is limited to analyse the user behaviour. We can use this extracted data to generate recommendations for the users and increase engagement on the website. The identifiers of users and items can be passed along with the user click rate to a supervised machine learning model of matrix factorization. We are using alternating least square (ALS) to predict the recommended items. The suggested system interfaces with the client system using a web API which will collect data and generate recommendations

An Enhanced Graphical Authentication Scheme Using Multiple-Image Steganography

Most remote systems require user authentication to access resources. Text-based passwords are still widely used as a standard method of user authentication. Although conventional text-based passwords are rather hard to remember, users often write their passwords down in order to compromise security. One of the most complex challenges users may face is posting sensitive data on external data centers that are accessible to others and do not be controlled directly by users. Graphical user authentication methods have recently been proposed to verify the user identity. However, the fundamental limitation of a graphical password is that it must have a colorful and rich image to provide an adequate password space to maintain security, and when the user clicks and inputs a password between two possible grids, the fault tolerance is adjusted to avoid this situation. This paper proposes an enhanced graphical authentication scheme, which comprises benefits over both recognition and recall-based graphical techniques besides image steganography. The combination of graphical authentication and steganography technologies reduces the amount of sensitive data shared between users and service providers and improves the security of user accounts. To evaluate the effectiveness of the proposed scheme, peak signal-to-noise ratio and mean squared error parameters have been used.

An Adaptive Refinement Scheme for Depth Estimation Networks

Deep learning has proved to be a breakthrough in depth generation. However, the generalization ability of deep networks is still limited, and they cannot maintain a satisfactory performance on some inputs. By addressing a similar problem in the segmentation field, a feature backpropagating refinement scheme (f-BRS) has been proposed to refine predictions in the inference time. f-BRS adapts an intermediate activation function to each input by using user clicks as sparse labels. Given the similarity between user clicks and sparse depth maps, this paper aims to extend the application of f-BRS to depth prediction. Our experiments show that f-BRS, fused with a depth estimation baseline, is trapped in local optima, and fails to improve the network predictions. To resolve that, we propose a double-stage adaptive refinement scheme (DARS). In the first stage, a Delaunay-based correction module significantly improves the depth generated by a baseline network. In the second stage, a particle swarm optimizer (PSO) delineates the estimation through fine-tuning f-BRS parameters-that is, scales and biases. DARS is evaluated on an outdoor benchmark, KITTI, and an indoor benchmark, NYUv2, while for both, the network is pre-trained on KITTI. The proposed scheme was effective on both datasets.

Retrieval-Based Factorization Machines for Human Click Behavior Prediction.

Human click behavior prediction is crucial for recommendation scenarios such as online commodity or advertisement recommendation, as it is helpful to improve the quality and user satisfaction of services. In recommender systems, the concept of click-through rate (CTR) is used to estimate the probability that a user will click on a recommended candidate. Many methods have been proposed to predict CTR and achieved good results. However, they usually optimize the parameters through a global objective function such as minimizing logloss or root mean square error (RMSE) for all training samples. Obviously, they intend to capture global knowledge of user click behavior but ignore local information. In this work, we propose a novel approach of retrieval-based factorization machines (RFM) for CTR prediction, which can effectively predict CTR by combining global knowledge which is learned from the FM method with the neighbor-based local information. We also leverage the clustering technique to partition the large training set into multiple small regions for efficient retrieval of neighbors. We evaluate our RFM model on three public datasets. The experimental results show that RFM performs better than other models in metrics of RMSE, area under ROC (AUC), and accuracy. Moreover, it is efficient because of the small number of model parameters.

Research on intelligent information system of user intelligent behavior data based on computer big data

This paper collects and analyzes users' online behaviors through computer big data technology, uses computer big data intelligent analysis system to analyze users' online shopping behavior, and uses information-based data analysis system to detect consumers' online shopping needs under the e-commerce platform. The main technique used in this paper is the computer browser log mining method. In the user's click stream data, the function keys of Tmall and Taobao webpages are used as data information for classification and collection. This paper uses the Bisecting K-means clustering algorithm to mine the state of interest. Finally, the feature maps of interests and behaviors are summarized. By processing four typical types of e-commerce user demand status, including background management type, continuous search type, product browsing type and information search data, and clustering based on page type, an effective method for dynamically changing demand judgment is obtained. The state of online shoppers is analyzed through data processing, which also proves the effectiveness of the computer intelligent information system.

SeaRank: relevance prediction based on click models in a reinforcement learning framework

PurposeUser feedback inferred from the user's search-time behavior could improve the learning to rank (L2R) algorithms. Click models (CMs) present probabilistic frameworks for describing and predicting the user's clicks during search sessions. Most of these CMs are based on common assumptions such as Attractiveness, Examination and User Satisfaction. CMs usually consider the Attractiveness and Examination as pre- and post-estimators of the actual relevance. They also assume that User Satisfaction is a function of the actual relevance. This paper extends the authors' previous work by building a reinforcement learning (RL) model to predict the relevance. The Attractiveness, Examination and User Satisfaction are estimated using a limited number of the features of the utilized benchmark data set and then they are incorporated in the construction of an RL agent. The proposed RL model learns to predict the relevance label of documents with respect to a given query more effectively than the baseline RL models for those data sets.Design/methodology/approachIn this paper, User Satisfaction is used as an indication of the relevance level of a query to a document. User Satisfaction itself is estimated through Attractiveness and Examination, and in turn, Attractiveness and Examination are calculated by the random forest algorithm. In this process, only a small subset of top information retrieval (IR) features are used, which are selected based on their mean average precision and normalized discounted cumulative gain values. Based on the authors' observations, the multiplication of the Attractiveness and Examination values of a given query–document pair closely approximates the User Satisfaction and hence the relevance level. Besides, an RL model is designed in such a way that the current state of the RL agent is determined by discretization of the estimated Attractiveness and Examination values. In this way, each query–document pair would be mapped into a specific state based on its Attractiveness and Examination values. Then, based on the reward function, the RL agent would try to choose an action (relevance label) which maximizes the received reward in its current state. Using temporal difference (TD) learning algorithms, such as Q-learning and SARSA, the learning agent gradually learns to identify an appropriate relevance label in each state. The reward that is used in the RL agent is proportional to the difference between the User Satisfaction and the selected action.FindingsExperimental results on MSLR-WEB10K and WCL2R benchmark data sets demonstrate that the proposed algorithm, named as SeaRank, outperforms baseline algorithms. Improvement is more noticeable in top-ranked results, which usually receive more attention from users.Originality/valueThis research provides a mapping from IR features to the CM features and thereafter utilizes these newly generated features to build an RL model. This RL model is proposed with the definition of the states, actions and reward function. By applying TD learning algorithms, such as the Q-learning and SARSA, within several learning episodes, the RL agent would be able to learn how to choose the most appropriate relevance label for a given pair of query–document.

Click-Through Rate Prediction Algorithm Based on Modeling of Implicit High-Order Feature Importance

<p>Click-through rate (CTR) prediction plays a central role in online advertising and recommendation systems. In recent years, with the successful application of deep neural networks (DNNs) in many fields, researchers have integrated deep learning into CTR prediction algorithms to model implicit high-order features. However, most of these existing methods unify the weights of implicit higher-order features to predict user behaviors. The importance of such features of different dimensions for predicting user click behaviors are different. Base on this, we propose a prediction method that dynamically learns the importance of implicit high-order features. Specifically, we integrate the output features of deep and shallow components, and adaptively learn the weights of implicit high-order features from among all features through the designed attention network, which effectively capturing the deep interests of users. In addition, this framework has strong versatility and can be combined with shallow models such as Logistic Regression (LR) and Factorization Machines (FMs) to form different models and achieve optimal performance. The extended experiment is conducted on two large-scale datasets, AVAZU and SafeDrive, and the experimental results show that the performance of the proposed model is superior to that of existing baseline models.</p> <p> </p>

Exploiting Post-Click Behaviors for Recommender System

In recent years, recommender systems become an effective technique to address the information overload problem on the Web, targeting the delivery of most relevant items and information to individual users based on their historical behaviours. To build an effective recommender system, implicit feedback like user click has been harvested. Since click data is usually very noisy, recent works also leverage dwell time as a proxy to optimize user engagement. However, dwell time is just one-dimensional post-click behaviour. The multi-dimensional post-click behaviours, e.g., reading comments, browsing images, and other sub-modules, have not been fully exploited. To address this issue, in this paper, we study leveraging post-click behaviours to improve user engagement in recommendation systems. We first take an E-commerce recommender system as an example to define post-click behaviours and demonstrate their effects with respect to user conversions. Based on the analysis, we then propose a tree-based labelling model, which provides a new perspective to understand user engagement beyond CTR or dwell time. The labelling model can be further incorporated into state-of-the-art recommendation methods. Extensive experiments on the dataset from JD.com, a real-world E-commerce website, demonstrate the competitive performance of the proposed method in both offline and online scenarios.

Learned iterative segmentation of highly variable anatomy from limited data: Applications to whole heart segmentation for congenital heart disease.

Training deep learning models that segment an image in one step typically requires a large collection of manually annotated images that captures the anatomical variability in a cohort. This poses challenges when anatomical variability is extreme but training data is limited, as when segmenting cardiac structures in patients with congenital heart disease (CHD). In this paper, we propose an iterative segmentation model and show that it can be accurately learned from a small dataset. Implemented as a recurrent neural network, the model evolves a segmentation over multiple steps, from a single user click until reaching an automatically determined stopping point. We develop a novel loss function that evaluates the entire sequence of output segmentations, and use it to learn model parameters. Segmentations evolve predictably according to growth dynamics encapsulated by training data, which consists of images, partially completed segmentations, and the recommended next step. The user can easily refine the final segmentation by examining those that are earlier or later in the output sequence. Using a dataset of 3D cardiac MR scans from patients with a wide range of CHD types, we show that our iterative model offers better generalization to patients with the most severe heart malformations.

Causal Inference in the Presence of Interference in Sponsored Search Advertising

In classical causal inference, inferring cause-effect relations from data relies on the assumption that units are independent and identically distributed. This assumption is violated in settings where units are related through a network of dependencies. An example of such a setting is ad placement in sponsored search advertising, where the likelihood of a user clicking on a particular ad is potentially influenced by where it is placed and where other ads are placed on the search result page. In such scenarios, confounding arises due to not only the individual ad-level covariates but also the placements and covariates of other ads in the system. In this paper, we leverage the language of causal inference in the presence of interference to model interactions among the ads. Quantification of such interactions allows us to better understand the click behavior of users, which in turn impacts the revenue of the host search engine and enhances user satisfaction. We illustrate the utility of our formalization through experiments carried out on the ad placement system of the Bing search engine.

User dynamic topology-information-based matrix factorization for e-government recommendation

Today, there is a growing tendency to integrate all public services on e-government websites, a trend that increases the risk of information overload. Matrix factorization (MF)-based recommendation models can effectively predict the user preference for an item, thus demonstrating their superiority in alleviating information overload problems. However, factors such as poor latent representations and the linear interaction function limit the further performance of these models To address these issues, we propose a novel method named user dynamic topology-information-based matrix factorization (User-DTMF). User-DTMF contains two designed modules: the dynamic topology feature learning module and the interactive preference learning module. The former first constructs the user dynamic topology sequence based on user historical behaviors. A convolutional neural network (CNN) is then used to extract features from the sequence. To obtain the final representations of users and items, the latter module uses MF and a fully connected (FC) neural network to model the dynamic topology features and static user-item interactive information. Next, another FC layer, which can be regarded as a nonlinear interaction function, is used to predict the probability of a user clicking on an item. Comprehensive experiments on a real dataset show that User-DTMF outperforms the best benchmark method by a margin of 1.68% and 3.05% in terms of Recall@10 and NDCG@10, respectively.

Personalized Smart Clothing Design Based on Multimodal Visual Data Detection.

In the traditional clothing customization system, only the designer participates in the clothing design, and the style is single. In the face of numerous styles, the user just repeatedly arranges and combines the styles, but does not realize the user's innovative design. In this paper, we propose a novel multitask deep convolutional neural network training method for task-by-task transfer learning, and learn deep image features for image retrieval tasks on noisy user click data. Image retrieval model based on image-text multimodal correlation features: this paper uses image-text multimodal correlation features to calculate the correlation between query keywords and images, and calculates the correlation between images and images. In this paper, the method of automatic generation of clothing style is researched, and parameterized coding is designed for it. Taking the typical style of suits as the initial population, through the human-computer interaction interface, the scoring value is assigned to the fitness value to carry out the evolution process. The binary string of the suit style generated by the genetic algorithm is decoded through the decoding algorithm and rules, decoded into a visual style diagram of the style of the suit style, and the style diagram of the suit style is automatically drawn. From the perspective of clothing design, this paper summarizes the general design methods of outdoor sports smart clothing, follows the integration of people-clothing-environment, and proposes a wearer-centered design concept to deeply explore the rationality of outdoor sports smart clothing design methods. This paper further solves the key problems in the design process, and takes the fashion of clothing as the core principle to design the structure of clothing modeling. The fabric selection of suits is based on the principle of clothing comfort. The key is to realize the outdoor sports monitoring function of suits through sensing technology. The design uses Arduino as an electronic prototype platform, so as to detect the heart rate of the human body during exercise and the microclimate temperature under the clothes. This kind of suit with monitoring function is ultimately a combination of sensing device and clothing. It not only has monitoring function, but also has the aesthetic concept of clothing design and conforms to the performance of human body structure, which will provide reference and reference for the design of outdoor sports smart clothing.

Multi-scale and multi-channel neural network for click-through rate prediction

Click-through rate (CTR) prediction aims to predict the probability of a user clicking on items and ads. Feature engineering plays an important role in improving the accuracy of CTR predict ion because features are normally multi-field. Recently, existing neural network based approaches mostly focus on the feature interactions in both low- and high-order. However, this methodology ignores salient features in both local and global contexts and diverse semantic features in various aspects. In this paper, we propose a novel framework called Multi-Scale and Multi-Channel neural network (MSMC) to learn the feature importance and feature semantics for enhancing CTR prediction. MSMC consists of two parallel modules: a salient feature encoder (SFE) and a diverse semantic feature encoder (DFE). The SFE employs an attentive global–local contexts module to extract the salient features. The DFE uses an attentive semantic module to aggregate the diverse semantic features. Then, a fusion function is adopted to adaptively combine features from the SFE and the DFE so as to obtain high-order feature interactions. Experimental results on two public real datasets illustrate the effectiveness of MSMC compared to the state-of-the-art CTR prediction methods. Our extensive analysis of MSMC shows how the salient features, diverse semantic features, and fusion function positively impact the performance of CTR prediction.

Session interest model for CTR prediction based on self-attention mechanism

Click-through rate prediction, which aims to predict the probability of the user clicking on an item, is critical to online advertising. How to capture the user evolving interests from the user behavior sequence is an important issue in CTR prediction. However, most existing models ignore the factor that the sequence is composed of sessions, and user behavior can be divided into different sessions according to the occurring time. The user behaviors are highly correlated in each session and are not relevant across sessions. We propose an effective model for CTR prediction, named Session Interest Model via Self-Attention (SISA). First, we divide the user sequential behavior into session layer. A self-attention mechanism with bias coding is used to model each session. Since different session interest may be related to each other or follow a sequential pattern, next, we utilize gated recurrent unit (GRU) to capture the interaction and evolution of user different historical session interests in session interest extractor module. Then, we use the local activation and GRU to aggregate their target ad to form the final representation of the behavior sequence in session interest interacting module. Experimental results show that the SISA model performs better than other models.

Feature distillation and accumulated selection for automated fraudulent publisher classification from user click data of online advertising

PurposeThe problem of choosing the utmost useful features from hundreds of features from time-series user click data arises in online advertising toward fraudulent publisher's classification. Selecting feature subsets is a key issue in such classification tasks. Practically, the use of filter approaches is common; however, they neglect the correlations amid features. Conversely, wrapper approaches could not be applied due to their complexities. Moreover, in particular, existing feature selection methods could not handle such data, which is one of the major causes of instability of feature selection.Design/methodology/approachTo overcome such issues, a majority voting-based hybrid feature selection method, namely feature distillation and accumulated selection (FDAS), is proposed to investigate the optimal subset of relevant features for analyzing the publisher's fraudulent conduct. FDAS works in two phases: (1) feature distillation, where significant features from standard filter and wrapper feature selection methods are obtained using majority voting; (2) accumulated selection, where we enumerated an accumulated evaluation of relevant feature subset to search for an optimal feature subset using effective machine learning (ML) models.FindingsEmpirical results prove enhanced classification performance with proposed features in average precision, recall, f1-score and AUC in publisher identification and classification.Originality/valueThe FDAS is evaluated on FDMA2012 user-click data and nine other benchmark datasets to gauge its generalizing characteristics, first, considering original features, second, with relevant feature subsets selected by feature selection (FS) methods, third, with optimal feature subset obtained by the proposed approach. ANOVA significance test is conducted to demonstrate significant differences between independent features.

Phishing short URL detection based on link jumping on social networks

Nowadays, a large number of people frequently use social networks. Social networks have become important platforms for people to publish and obtain information. However, social networks have also become the main venue for hackers to initiate online fraud. Phishing is a common way used by hackers to launch online fraud on social networks. This paper proposes a method for detecting phishing short URL based on the link jumping. The method uses a hierarchical hidden Markov model with two-layer structure to describe the link jumping process after user clicking on short URL, so as to identify phishing short URL on social networks. The proposed method includes a training phase and an identification phase. In the identification phase, the average log-likelihood probability of the observation sequence is calculated. An experiment based on real datasets of Weibo is conducted to evaluate this method. The experiment results validate the effectiveness of this method.

Research on Improvement of the Click-Through Rate Prediction Model Based on Differential Privacy

Click-through rate prediction is crucial in network applications such as recommendation systems and online networks. Designing feature extraction schemes to obtain features and modeling users’ click behavior are used to estimate the probability of users clicking on recommended items. The AutoInt model is a recent and effective research finding. It constructs combined features by referencing the multi-head attention mechanism but does not fully mine meaningful high-order cross-features and ignores user privacy protection. To address this problem, this study proposes the differential privacy bidirectional long short-term memory network (DP-Bi-LSTM-AutoInt) model, which is an improved AutoInt model. A bidirectional long short-term memory network is added after the embedding layer to deeply mine the nonlinear relationship between user click behaviors and construct high-order features. Further, differential privacy technology is adopted for user privacy protection, and the Gaussian mechanism is used to randomly perturb the gradient descent algorithm of the model. Using the Criteo dataset to conduct experiments, the experimental results show that the accuracy of the Bi-LSTM-AutoInt model proposed herein is improved by 0.65 % compared to the original AutoInt model. When the privacy budget is greater than 3.0, the accuracies of the DP-Bi-LSTM-AutoInt and Bi-LSTM-AutoInt models are nearly equivalent. However, the DP-Bi-LSTM-AutoInt model algorithm is more secure and reliable than the AutoInt model.