Clickstream Research Articles

JobFormer: Skill-Aware Job Recommendation with Semantic-Enhanced Transformer

Job recommendation aims to provide potential talents with suitable job descriptions (JDs) consistent with their career trajectory, which plays an essential role in proactive talent recruitment. In real-world management scenarios, the available JD-user records always consist of JDs, user profiles, and click data, in which the user profiles are typically summarized as the user's skill distribution for privacy reasons. Although existing sophisticated recommendation methods can be directly employed, effective recommendation still has challenges considering the information deficit of JD itself and the natural heterogeneous gap between JD and user profile. To address these challenges, we proposed a novel skill-aware recommendation model based on the designed semantic-enhanced Transformer to parse JDs and complete personalized job recommendation. Specifically, we first model the relative items of each JD and then adopt an encoder with the local-global attention mechanism to better mine the intra-job and inter-job dependencies from JD tuples. Moreover, we adopt a two-stage learning strategy for skill-aware recommendation, in which we utilize the skill distribution to guide JD representation learning in the recall stage, and then combine the user profiles for final prediction in the ranking stage. Consequently, we can embed rich contextual semantic representations for learning JDs, while skill-aware recommendation provides effective JD-user joint representation for click-through rate (CTR) prediction. To validate the superior performance of our method for job recommendation, we present a thorough empirical analysis of large-scale real-world and public datasets to demonstrate its effectiveness and interpretability.

Does Online Fundraising Increase Charitable Giving? A Nationwide Field Experiment on Facebook

Does online fundraising increase charitable giving? Using the Facebook advertising tool, we implemented a natural field experiment across Germany, randomly assigning almost 8,000 postal codes to Save the Children fundraising videos or to a pure control. We studied changes in the donation revenue and frequency for Save the Children and other charities by postal code. Our geo-randomized design circumvented many difficulties inherent in studies based on click-through data, especially substitution and measurement issues. We found that (i) video fundraising increased donation revenue and frequency to Save the Children during the campaign and in the subsequent five weeks, that (ii) the campaign was profitable for the fundraiser, and that (iii) the effects were similar independent of video content and impression assignment strategy. However, we also found some crowding out of donations to other similar charities or projects. Finally, we demonstrated that click data may be an inappropriate proxy for donations and recommend that managers use careful experimental designs that can plausibly evaluate the effects of advertising on relevant outcomes. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Funding: M. Adena acknowledges financial support from the Deutsche Forschungsgemeinschaft [Grants 417014946 and CRC TRR 190 (Project 280092119)]. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2020.00596 .

A Novel Deep Learning Model for Student Performance Prediction Using Engagement Data

Technology-enhanced learning supported by virtual learning environments (VLEs) facilitates tutors and students. VLE platforms contain a wealth of information that can be used to mine insight regarding students’ learning behaviour and relationships between behaviour and academic performance, as well as to model data-driven decision-making. This study introduces a system that we termed ASIST: a novel Attention-aware convolutional Stacked BiLSTM network for student representation learning to predict their performance. ASIST exploits student academic registry, VLE click stream, and midterm continuous assessment information for their behaviour representation learning. ASIST jointly learns the student representation using five behaviour vectors. It processes the four sequential behaviour vectors using a separate stacked bidirectional long short term memory (LSTM) network. A deep convolutional neural network models the diurnal weekly interaction behaviour. It also employs the attention mechanism to assign weight to features based on their importance. Next, five encoded feature vectors are concatenated with the assessment information, and, finally, a softmax layer predicts the high-performer (H), moderate-performer (M), and at-risk (F) categories of students. We evaluate ASIST over three datasets from an Irish university, considering five evaluation metrics. ASIST achieves an area under the curve (AUC) score of 0.86 to 0.90 over the three datasets. It outperforms three baseline deep learning models and four traditional classification models. We also found that the attention mechanism has a slight impact on ASIST’s performance. The ablation analysis reveals that weekly event count has the greatest impact on ASIST, whereas diurnal weekly interaction has the least impact. The early prediction using the first seven weeks of data achieves an AUC of 0.83 up to 0.89 over the three datasets. In yearly analysis, ASIST performs best over the 2018/19 dataset and worst over the 2020/21 dataset.

Predict student learning styles and suitable assessment methods using click stream

Adaptive learning, which aims to give each learner engaging, effective learning experiences, is one method of offering modified education. Adaptive learning seeks to consider the student's unique characteristics by personalizing the learning course materials and evaluation procedures. To determine the student's preferred learning strategies, we first ascertain their attributes utilizing VAK learning styles. In this study, we developed an integrated model to classify learners based on their learning activity clicks by combining machine learning algorithms like K-Nearest Neighbor (KNN), random forest (RF), and support vector machine (SVM) and Logistic regression (LR) with semantic association, which is used to help us map learning activity with VAK learning style. This enables us to classify learners, determine their preferred methods of learning, and offer the most suitable; as a result, we were able to group pupils according to their learning styles and provide the best evaluation technique or strategies. To assess the effectiveness of the suggested model, several tests were executed on the actual dataset (Open University Learning Analytics Dataset, or OULAD). According to studies, using a Random Forest algorithm, the suggested model can predict which evaluation strategy or strategies will be most effective for each student and can classify individuals with the highest degree of accuracy—98%.

Validating Synthetic Usage Data in Living Lab Environments

Evaluating retrieval performance without editorial relevance judgments is challenging, but instead, user interactions can be used as relevance signals. Living labs offer a way for small-scale platforms to validate information retrieval systems with real users. If enough user interaction data are available, then click models can be parameterized from historical sessions to evaluate systems before exposing users to experimental rankings. However, interaction data are sparse in living labs, and little is studied about how click models can be validated for reliable user simulations when click data are available in moderate amounts. This work introduces an evaluation approach for validating synthetic usage data generated by click models in data-sparse human-in-the-loop environments like living labs. We ground our methodology on the click model’s estimates about a system ranking compared to a reference ranking for which the relative performance is known. Our experiments compare different click models and their reliability and robustness as more session log data become available. In our setup, simple click models can reliably determine the relative system performance with already 20 logged sessions for 50 queries. In contrast, more complex click models require more session data for reliable estimates, but they are a better choice in simulated interleaving experiments when enough session data are available. While it is easier for click models to distinguish between more diverse systems, it is harder to reproduce the system ranking based on the same retrieval algorithm with different interpolation weights. Our setup is entirely open, and we share the code to reproduce the experiments.

Using multimodal learning analytics to understand effects of block‐based and text‐based modalities on computer programming

AbstractBackgroundWith the development of computational literacy, there has been a surge in both research and practice application of text‐based and block‐based modalities within the field of computer programming education. Despite this trend, little work has actually examined how learners engaging in programming process when utilizing these two major programming modalities, especially in the context of secondary education settings.ObjectivesTo further compare programming effects between and within text‐based and block‐based modalities, this research conducted a quasi‐experimental research in China's secondary school.MethodsAn online programming platform, Code4all, was developed to allow learners to program in text‐based and block‐based modalities. This research collected multimodal data sources, including programming platform data, process data, and performance data. This research further utilized multiple learning analytics approaches (i.e., clustering analysis, click stream analysis, lag‐sequential analysis and statistics) to compare learners' programming features, behavioural patterns and knowledge gains under two modalities.Results and ConclusionsThe results indicated that learners in text‐based modality tended to write longer lines of code, encountered more syntactical errors, and took longer to attempt debugging. In contrast, learners in block‐based modality spent more time operating blocks and attempt debugging, achieving better programming knowledge performances compared to their counterparts. Further analysis of five clusters from the two modalities revealed discrepancies in programming behavioural patterns.ImplicationsThree major pedagogical implications were proposed based on empirical research results. Furthermore, this research contributed to the learning analytics literature by integrating process‐oriented and summative analysis to reveal learners' programming learning quality.

Does diversity facilitate consumer decisions: a comparative perspective based on single-category versus multi-category products

PurposeThis study aims to explore the opposite effects of single-category versus multi-category products information diversity on consumer decision making. Further, the authors investigate the moderating role of three categories of visitors – direct, hesitant and hedonic – in the relationship between product information diversity and consumer decision making.Design/methodology/approachThe research utilizes a sample of 1,101,062 product click streams from 4,200 consumers. Visitors are clustered using the k-means algorithm. The diversity of information recommendations for single and multi-category products is characterized using granularity and dispersion, respectively. Empirical analysis is conducted to examine their influence on the two-stage decision-making process of heterogeneous online visitors.FindingsThe study reveals that the impact of recommended information diversity on consumer decision making differs significantly between single-category and multiple-category products. Specifically, information diversity in single-category products enhances consumers' click and purchase intention, while information diversity in multiple-category products reduces consumers' click and purchase intention. Moreover, based on the analysis of online visiting heterogeneity, hesitant, direct and hedonic features enhance the positive impact of granularity on consumer decision making; while direct features exacerbate the negative impact of dispersion on consumer decision making.Originality/valueFirst, the article provides support for studies related to information cocoon. Second, the research contributes evidence to support the information overload theory. Third, the research enriches the field of precision marketing theory.

A Survey of Sequential Pattern Based E-Commerce Recommendation Systems

E-commerce recommendation systems usually deal with massive customer sequential databases, such as historical purchase or click stream sequences. Recommendation systems’ accuracy can be improved if complex sequential patterns of user purchase behavior are learned by integrating sequential patterns of customer clicks and/or purchases into the user–item rating matrix input of collaborative filtering. This review focuses on algorithms of existing E-commerce recommendation systems that are sequential pattern-based. It provides a comprehensive and comparative performance analysis of these systems, exposing their methodologies, achievements, limitations, and potential for solving more important problems in this domain. The review shows that integrating sequential pattern mining of historical purchase and/or click sequences into a user–item matrix for collaborative filtering can (i) improve recommendation accuracy, (ii) reduce user–item rating data sparsity, (iii) increase the novelty rate of recommendations, and (iv) improve the scalability of recommendation systems.

Ad Click Prediction: A Comparative Evaluation of Logistic Regression and Performance Metrics

Abstract: This research paper investigates the effectiveness of logistic regression as a predictive modelling technique for ad click prediction. The study aims to explore the performance of logistic regression and evaluate its predictive power using various evaluation metrics. The primary objective is to assess the accuracy, precision, recall, F1 score, and area under the receiver operating characteristic curve (AUC-ROC) of logistic regression models in predicting ad clicks. The paper begins with a comprehensive review of the literature on ad click prediction and logistic regression. It highlights the importance of accurate click-through rate (CTR) estimation for advertisers and the potential benefits of logistic regression in this context. The theoretical background of logistic regression is also discussed, providing an understanding of its underlying principles and assumptions. Next, the methodology section describes the dataset used for the study, which includes historical ad impressions and click data. The data pre-processing steps, including feature selection and transformation, are explained. Logistic regression models are then trained on the pre-processed data, and the model performance is evaluated using various evaluation metrics. The results section presents the findings of the study. It includes a detailed analysis of the accuracy, precision, recall, F1 score, and AUC-ROC obtained from the logistic regression models. The performance of the models is compared against benchmark models or alternative algorithms commonly used in ad click prediction. The results highlight the strengths and limitations of logistic regression in predicting ad clicks. Furthermore, the discussion section provides insights into the implications of the results. It discusses the interpretability of logistic regression models and their potential for providing actionable insights to advertisers. The limitations and potential challenges of logistic regression in ad click prediction are also addressed. Finally, the conclusion section summarizes the key findings of the research and provides recommendations for future studies. It emphasizes the significance of logistic regression as a reliable and interpretable method for ad click prediction, while also recognizing the need for further research to improve its performance.

Fuzzy based optimized itemset mining in high dimensional transactional database using adaptable FCM

One of the extremely deliberated data mining processes is HUIM (High Utility Itemset Mining). Its applications include text mining, e-learning bioinformatics, product recommendation, online click stream analysis, and market basket analysis. Likewise lot of potential applications availed in the HUIM. However, HUIM techniques could find erroneous patterns because they don’t look at the correlation of the retrieved patterns. Numerous approaches for mining related HUIs have been presented as an outcome. The computational expense of these methods continues to be problematic, both in terms of time and memory utilization. A technique for extracting weighted temporal designs is therefore suggested to rectify the identified issue in HUIM. Preprocessing of time series-based information into fuzzy item sets is the first step of the suggested technique. These feed the Graph Based Ant Colony Optimization (GACO) and Fuzzy C Means (FCM) clustering methodologies used in the Improvised Adaptable FCM (IAFCM) method. The suggested IAFCM technique achieves two objectives: optimal item placement in clusters using GACO; and ii) IAFCM clustering and information decrease in FCM cluster. The proposed technique yields high-quality clusters by GACO. Weighted sequential pattern mining, which considers facts of patterns with the highest weight and low frequency in a repository that is updated over a period, is used to locate the sequential patterns in these clusters. The outcomes of this methodology make evident that the IAFCM with GACO improves execution time when compared to other conventional approaches. Additionally, it enhances information representation by enhancing accuracy while using a smaller amount of memory.

Doubly Robust Estimation for Correcting Position Bias in Click Feedback for Unbiased Learning to Rank

Clicks on rankings suffer from position bias: generally items on lower ranks are less likely to be examined—and thus clicked—by users, in spite of their actual preferences between items. The prevalent approach to unbiased click-based learning-to-rank (LTR) is based on counterfactual inverse-propensity-scoring (IPS) estimation. In contrast with general reinforcement learning, counterfactual doubly robust (DR) estimation has not been applied to click-based LTR in previous literature. In this article, we introduce a novel DR estimator that is the first DR approach specifically designed for position bias. The difficulty with position bias is that the treatment—user examination—is not directly observable in click data. As a solution, our estimator uses the expected treatment per rank, instead of the actual treatment that existing DR estimators use. Our novel DR estimator has more robust unbiasedness conditions than the existing IPS approach, and in addition, provides enormous decreases in variance: our experimental results indicate it requires several orders of magnitude fewer datapoints to converge at optimal performance. For the unbiased LTR field, our DR estimator contributes both increases in state-of-the-art performance and the most robust theoretical guarantees of all known LTR estimators.

Ad Fraud Detector for Mobile Applications

Ad fraud happens in cost per click ad networks where publishers charge advertisers for each click. Ad fraud is posing the huge loss to the mobile advertising industry. The conventional technologies use ensemble machine learning methods, neglecting the cost of incorrect classification for a fraud publisher is higher than a normal publisher. An effective classification model for variable ad fraud is proposed in this paper. Cost-sensitive Back Propagation Neural Network is combined with the novel Artificial Bee Colony algorithm in this research (CSBPNN-ABC). Feature selection is synchronously optimized with BPNN connection weights by ABC to reduce the interaction between features and weights. Cost Parameters are added to BPNN by correcting the error function. Experiments on real world click data in mobile advertising show that its superior classification performance compared with the state-of-the-art technology.

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.

Towards a unified search: Improving PubMed retrieval with full text

ObjectiveA significant number of recent articles in PubMed have full text available in PubMed Central®, and the availability of full texts has been consistently growing. However, it is not currently possible for a user to simultaneously query the contents of both databases and receive a single integrated search result. In this study, we investigate how to score full text articles given a multitoken query and how to combine those full text article scores with scores originating from abstracts and achieve an overall improved retrieval performance. Materials and methodsFor scoring full text articles, we propose a method to combine information coming from different sections by converting the traditionally used BM25 scores into log odds ratio scores which can be treated uniformly. We further propose a method that successfully combines scores from two heterogenous retrieval sources – full text articles and abstract only articles – by balancing the contributions of their respective scores through a probabilistic transformation. We use PubMed click data that consists of queries sampled from PubMed user logs along with a subset of retrieved and clicked documents to train the probabilistic functions and to evaluate retrieval effectiveness. Results and conclusionsRandom ranking achieves 0.579 MAP score on our PubMed click data. BM25 ranking on PubMed abstracts improves the MAP by 10.6%. For full text documents, experiments confirm that BM25 section scores are of different value depending on the section type and are not directly comparable. Naïvely using the body text of articles along with abstract text degrades the overall quality of the search. The proposed log odds ratio scores normalize and combine the contributions of occurrences of query tokens in different sections. By including full text where available, we gain another 0.67%, or 7% relative improvement over abstract alone. We find an advantage in the more accurate estimate of the value of BM25 scores depending on the section from which they were produced. Taking the sum of top three section scores performs the best.

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.

Machine Learning and AI Application Behaviour Prediction for User Experience Modelling and Optimization

The purpose of this research is to offer a technique for assessing user experience in mobile applications utilizing AIAM technology. Due to ineffective and time-consuming nature of conventional data gathering techniques (such as user interviews and user inference), AIAM concentrates on using Artificial Intelligence (AI) to assess and enhance user experience. Logs from a mobile application may be used to gather information about user activity. Only a few parameters of data are utilized in the process of surfing and running mobile applications to ensure the privacy of users. The method's objective is to create the deep neural network prototype as close as feasible to a user's experience when using a mobile app. For particular objectives, we create and employ application interfaces to train computational models. The click data from all users participating in a certain task is shown on these projected pages. User activity may therefore be mapped in connected and hidden layers of the system. Finally, the social communications application is used to test the efficacy of the suggested method by implementing the improved design.

Determinants of paying attention to the logos of online travel agencies

A logo is an important part of brand image and a key visual element that has an impact on cognitive, emotional and behavioural reactions. This article examines how often tourists consciously pay attention to the logos of online travel agencies and brand name types in the context of their demographic characteristics and tourism behaviour. There were a total of 170 participants in the study. Empirical data was collected using two research techniques: a survey questionnaire and click tracking. Data analysis was carried out in two stages. In the first stage, it was assessed whether there is a relationship between the demographic characteristics and tourism behaviour of the respondents and their perception of the logos of online travel agencies. In the second stage, correspondence analysis (CA) was performed to obtain information on the relationship between the frequency with which the respondents pay attention to logos and their preferred type of logo. The study showed that the frequency with which the respondents pay attention to the logos of online travel agencies varies with their education, tourism expenditure, and the number of tourism trips they make in a year. The perception of logos as regards the type of brand name they include varies depending on the frequency of tourism trips and the preferred type of travel arrangements. This study provides important guidance on the design of logos and visual communication strategies, thus helping to facilitate the brand value creation process.

A Novel Approach for Discovering the Patterns by using PDBD Model in Big Data

Big data is a term that refers to information collected from a wide range of sources, including transaction processing systems, sensors, digital photographs, internet click stream logs, movies, and social media. The goal of text mining is to reveal new patterns and relationships that could lead to the discovery of previously unknown sources. It provides ample scope for interpretation of genuine user intent or contextual meanings as the case be. This research focuses on the creation of a novel approach for discovering patterns in big data using the Pattern Discovery in Big Data (PDBD) model in text mining. From textual sources, text-mining techniques extract a range of patterns and important facts. Sequential pattern mining helps to extract informative elements from a set of sequences based on the frequency of occurrences. The Pattern Discovery in Big Data (PDBD) model study has four primary contributions to uncovering patterns uniquely and effectively: Semantic information extraction, pattern improvements, LDA Topic Modeling, and cluster assignments. The proposed work compared the Pattern Taxonomy Model (Inner Pattern Evolving) pattern inner evolving model with other existing similar benchmark models to evaluate the model performance. The empirical results prove that the proposed model outperformed the compared models with significant improvement in accuracy Improvement in this Pattern Discovery in Big Data (PDBD) model yields better results in discovering the patterns.

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.

Predicting the At-Risk Online Students Based on the Click Data Distribution Characteristics

High fail and dropout rates are the major problems in distance education. Due to a large number of online learners and limited teacher resources, it is essential to accurately identify these potential at-risk students in advance and provide timely aids, which will help to improve the educational outcome. In the online learning environment, students’ online learning behaviors can be recorded easily, with the click data being the most common one. Students’ learning behavior can reflect their learning situation and may differ among different students and periods. This paper proposed a model that uses the short-period activity characteristic and long-term changing pattern to predict the potential at-risk students. The model contains two stages: information extraction and information utilization. The first stage extracts data from the log files and organizes it in a form suitable for the model. In the second stage, according to the different characteristics of students’ short-term and long-term learning behavior, a convolution residual recurrent neural network (CRRNN) model is proposed. The convolutional neural network is used to obtain the representation of the student’s learning behavior in a certain period. Then, the residual recurrent neural network is used to get the behavior changing pattern over the periods. The experimental results indicate that the proposed model has higher performance than the three widely used baseline methods on the OULA dataset and has good practical application value for teaching and management.