Session Logs Research Articles

Implementation of training and coaching to improve goal planning and family engagement in early childhood home visiting

Early childhood home visiting services aim to support child and family health and development, yet challenges engaging and retaining families persist. One promising strategy to improve family engagement involves enhancing the capacity of the home visiting workforce to partner with families to set and implement service goals that align with families’ needs and preferences. This study examined implementation of ‘Goal Plan Strategies’ (GPS), a training and coaching innovation designed to improve family-centered goal planning. Aims were to evaluate (1) the extent to which training and eight sessions of audio-based coaching were implemented with fidelity, (2) whether coaching fidelity was associated with home visitors’ perceptions of coaching and confidence using goal planning strategies, and (3) whether GPS was perceived as feasible, acceptable, and useful by coaches and home visitors. Home visitors (N = 22) and coaches (N = 10) took part in GPS and completed surveys, session logs, and interviews. Training and coaching sessions were observed and coded for fidelity (e.g., dose, adherence, quality). GPS was implemented with high levels of fidelity. Findings support feasibility, acceptability, and usefulness of GPS, overall, and suggest the need to examine coaching dosage and overlap with other implementation supports more closely. Findings highlight the benefits of audio-based coaching, mentoring of coaches, and observational measures of implementation fidelity.

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.

Learning from the Wisdom of Crowds: Exploiting Similar Sessions for Session Search

Search engines are essential internet services, enabling users to efficiently find the information they need. Session search employs users’ session logs of queries to solve complex retrieval tasks, in which users search multiple times until interested documents are found. Most existing session search models focus on the contextual information within the current search, ignoring the evidence from historical search sessions. Considering the fact that many ongoing retrieval tasks should have already been carried out by other users with a similar intent, we argue that historical sessions with similar intents can help improve the accuracy of the current search task. We propose a novel Similar Session-enhanced Ranking (SSR) model to improve the session search performance using historical sessions with similar intents. Specifically, the candidate historical sessions are matched by query-level and session-level semantic similarity, and then query-level neighbor behaviors are aggregated by a Query-guided GNN (QGNN) while session-level neighbor behaviors are aggregated using the attention mechanism. Finally, we integrate the refined and aggregated historical neighbor information into the current search session. Experimental results on AOL and Tiangong-ST datasets show that our SSR model significantly outperforms the state-of-the-art models.

LWS: A framework for log-based workload simulation in session-based SUT

Artificial intelligence for IT Operations (AIOps) plays a critical role in operating and managing cloud-native systems and microservice-based applications but is limited by the lack of high-quality datasets with diverse scenarios. Realistic workloads are the premise and basis of generating such AIOps datasets, with the session-based workload being one of the most typical examples. Due to privacy concerns, complexity, variety, and requirements for reasonable intervention, it is difficult to copy or generate such workloads directly, showing the importance of effective and intervenable workload simulation. In this paper, we formulate the task of workload simulation and propose a framework for Log-based Workload Simulation (LWS) in session-based systems. LWS extracts the workload specification including the user behavior abstraction based on agglomerative clustering as well as relational models and the intervenable workload intensity from session logs. Then LWS combines the user behavior abstraction with the workload intensity to generate simulated workloads. The experimental evaluation is performed on an open-source cloud-native application with both well-designed and public real-world workloads, showing that the simulated workload generated by LWS is effective and intervenable, which provides the foundation of generating high-quality AIOps datasets.

Process-outcome relations in music therapy versus music listening for people with schizophrenia viewed through a mediational model: the role of the therapeutic alliance.

Examine whether change in clinical outcomes for patients with schizophrenia and negative symptoms randomized to either Music Therapy (MT) or Music Listening (ML) is associated to moderators and mediators, with focus on alliance, attendance and dropout. An exploratory post-hoc analysis of data from an original randomized controlled trial (RCT) investigating the effect of MT vs. ML for people with schizophrenia and negative symptoms. Inclusion to the study was implemented through screening of referred patients for symptoms of schizophrenia and negative symptoms. A total of 57 patients were randomly assigned, 28 to MT and 29 to ML. Session logs and notes were included in this study. Statistical analysis investigated moderator and mediator relation to outcome variables: Negative symptoms, functioning, quality of life, and retention to treatment. On average, participants in MT attended 18.86 sessions (SD = 7.17), whereas those in ML attended 12.26 (SD = 9.52), a statistically significant difference (p = 0.0078). Dropout at 25 weeks was predicted by intervention, with dropout being 2.65 (SE = 1.01) times more likely in ML than in music therapy (p = 0.009). Helping alliance score at weeks was explained by intervention, with mean score being 0.68 (SE = 0.32) points lower in ML than in MT (p = 0.042). The number of sessions attended was also explained by intervention, with participants in ML attending on average 6.17 (SE = 2.24) fewer sessions than those randomized to MT (p = 0.008). Though both groups improved significantly, improvements in negative symptoms, depression, and functioning tended to be higher in ML, whereas improvements in alliance and quality of life tended to be higher in MT. The analysis could not detect a direct link between helping alliance score and outcome variables. However, the analysis documented a stronger alliance developed in the MT group, a lower dropout rate, as well as higher attendance in treatment.Clinical Trial Registration: www.ClinicalTrials.gov, identifier: NCT02942459.

DATA AND WEB MINING

Data and web mining involves the use of data mining techniques to extract information from the web in order to gain insights into online behavior, customer preferences, and trends. The process requires data preprocessing to clean and prepare the data for analysis, and pattern recognition to identify relationships and patterns within the data. The benefits of data and web mining include improved decision-making and business growth, while challenges include the large volume of data and ensuring data quality. Web mining can be divided into three main categories: web content mining, web structure mining, and web usage mining. Web content mining involves extracting information from the content of web pages, such as text, images, and videos. Web structure mining involves analyzing the links between web pages to identify patterns and relationships. Web usage mining involves analyzing user behavior on the web, such as clickstream data and session logs, to extract useful information. KEYWORDS: data mining, web mining, information extraction, data preprocessing, pattern recognition, online behavior, customer preferences,

A gradient boosting classifier for purchase intention prediction of online shoppers

Early purchase prediction plays a vital role for an e-commerce website. It enables e-shoppers to enlist consumers for product suggestions, offer discount and for many other interventions. Several work has already been done using session log for analyzing customer behavior whether he performs a purchase on the product or not. In most cases, it is difficult to find out and make a list of customers and offer them discount when their session ends. In this paper, we propose a customer's purchase intention prediction model where e-shoppers can detect customer's purpose earlier. First, we apply feature selection technique to select best features. Then the extracted features are fed to train supervised learning models. Several classifiers like support vector machine (SVM), random forest (RF), multilayer perceptron (MLP), decision tree (DT), and XGBoost classifiers have been applied along with oversampling method for balancing the dataset. The experiments were performed on a standard benchmark dataset. Experimental results show that XGBoost classifier with feature selection techniques and oversampling method has the significantly higher area under ROC curve (auROC) score and are under precision-recall curve (auPR) score which are 0.937 and 0.754 respectively. On the other hand accuracy achieved by XGBoost and Decision tree are significantly improved and they are 90.65% and 90.54% respectively. Overall performance of the gradient boosting method is significantly improved compared to other classifiers and state-of-the-art methods. In addition to this, a method for explainable analysis on the problem was outlined.

ARTP: Anomaly based real time prevention of Distributed Denial of Service attacks on the web using machine learning approach

Distributed Denial of Service (DDoS) attack is one of the most destructive internet network attacks, denying legitimate users access to resources and networks by maliciously blocking available computing resources. Intruders send a large number of packets to the network in order to create a crowding effect. Unlike a Denial of Service (DoS) attack, where a single compromised source generates all of the traffic, a Distributed Denial of Service (DDoS) attack generates traffic from multiple compromised nodes spread across multiple geographies. To address the challenges posed by the Distributed Denial of Service (DDoS) attack, several researchers proposed a variety of solutions for early detection and prevention of the attack. Effective solutions for the prevention and early detection of Distributed Denial of Service (DDoS) attacks, on the other hand, have yet to be developed, and the problem remains a prominent research focus area. This paper tries to present a novel and optimal solution for detecting Distributed Denial of Service (DDoS) attacks on internet networks more quickly and accurately. The proposed model is an anomaly-based real-time prevention model for web networks. The model is based on machine learning principles and can effectively counter new types of Distributed Denial of Service (DDoS) attacks. To demonstrate the efficiency, accuracy, model robustness, and relative of the proposed model, a simulation study was run on an LLDOS session log, and the results indicated that the model performed better than benchmark models found in the literature.

Adapting in-person diabetes group visits to a virtual setting across federally qualified health centers.

Diabetes group visits (GVs) have been shown to improve glycemic control, enrich patient self-care, and decrease healthcare utilization among patients with type 2 diabetes mellitus (T2DM). While telehealth has become routine, virtual GVs remain understudied, especially in federally qualified health centers (FQHCs). We conducted a 5-year cluster randomized trial with a waitlist control group to test the impact of diabetes GVs on patients' outcomes in Midwestern FQHCs. Due to COVID-19, the 6 waitlisted FQHCs adapted to virtual GVs. FQHC staff were provided training and support to implement virtual GVs. The GV intervention included 6 monthly 1-1.5-h long education sessions and appointments with a primary care provider. We measured staff perspectives and satisfaction via GV session logs, monthly webinars, and staff surveys and interviews. Adaptations for implementation of virtual GV included: additional staff training, video conferencing platform use, decreased session length and group size, and adjusting study materials, activities, and provider appointments. Sites enrolled a total of 48 adults with T2DM for virtual GVs. Most FQHCs were urban and all FQHCs predominantly had patients on public insurance. Patients attended 2.1 ± 2.2 GVs across sites on average. Thirty-four patients (71%) attended one or more virtual GVs. The average GV lasted 79.4 min. Barriers to virtual GVs included patient technology issues and access, patient recruitment and enrollment, and limited staff availability. Virtual GV facilitators included providing tablets, internet access from the clinic, and technical support. Staff reported spending on average 4.9 h/week planning and implementing GVs (SD = 5.9). On average, 6 staff from each FQHC participated in GV training and 1.2 staff reported past GV experience. All staff had worked at least 1 year at their FQHC and most reported multiple years of experience caring for patients with T2DM. Staff-perceived virtual GV benefits included: empowered patients to manage their diabetes, provided patients with social support and frequent contact with providers, improved relationships with patients, increased team collaboration, and better patient engagement and care-coordination. Future studies and health centers can incorporate these findings to implement virtual diabetes GVs and promote accessible diabetes care.

Profile of adults users of smartphone applications for monitoring the level of physical activity and associated factors: A cross-sectional study.

BackgroundThere are currently more than 200 million smartphones in Brazil. The potential of mobile technologies for favorable changes in health behavior such as physical activity has been previously described in the literature. Results of surveys in developed countries indicate that applications (APPs) are developed for people who are better educated, younger, and with higher incomes compared to non-users. However, the profile of users in developing countries like Brazil is not well-known. Understanding the profile of APP users might ease the development turned to physically inactive people and those at higher cardiovascular risk. Furthermore, the physiological and functional factors associated with the use of such APP are unknown.ObjectivesTo characterize the profile of APP users to monitor the physical activity level (PAL) and assess the demographic, socioeconomic, clinical, physiological, and functional characteristics associated with the use of smartphone APPs to monitor physical activity in Brazilian adults.MethodsWe assessed 176 asymptomatic men and 178 women (43 ± 12 years; 27 ± 5 kg/m2). We initially asked participants about their current use of a smartphone APP containing PAL monitoring functionality, such as exercise session logs and/or step counts. In a cross-sectional design, we investigated schooling, socioeconomic status (Critério Brasil), and classic self-reported cardiovascular risk factors. We evaluated several physiological and functional variables such as maximum O2 consumption on a treadmill (VO2 max), blood pressure, body composition (bioelectrical impedance), handgrip strength, and isokinetic muscle strength of the dominant lower limb. Participants used a triaxial accelerometer for 7 days to quantify daily physical activity. We also assessed health-related quality of life (WHOQOL BREF), perceived stress (PSS14 Scale), and the built environment (NEWS Scale). We compared continuous variables using the Student's t-test and categorical variables using the χ2 test, between APP users and non-users. After univariate analysis, we included the main variables associated with the use of APP in a multiple logistic regression model.ResultsOne hundred and two participants (28.3%), unrelated to gender, reported using a smartphone APP for physical activity at the time of assessment. Except for perceived stress and the built environment that were not associated with the use of APP, users of APP were younger and had higher education, lower cardiovascular risk, better socioeconomic status, a better quality of life, better cardiorespiratory function, better body composition, greater physical fitness and more moderate to vigorous physical activity in daily life. The results of the multiple logistic regression showed that age, hypertension, VO2 max, socioeconomic status (Critério Brasil), and quality of life (WHOQOL BREF total score) were the variables most significantly associated with the use of the APP.ConclusionsOur results indicate that smartphone APPs to monitor physical activity are developed for younger adults with better socioeconomic status, lower cardiovascular risk, higher quality of life, and greater cardiorespiratory fitness. Greater efforts are needed to develop a science-based APP for people who most need this technology, enabling greater potential to prevent undesirable health outcomes in asymptomatic adults.

Blockchain-Enabled and Data-Driven Smart Healthcare Solution for Secure and Privacy-Preserving Data Access

The major advances in body-mounted sensors and wireless technologies have been revolutionizing the healthcare industry, where patient’s conditions can be remotely monitored by medical staff. Such a model is gaining broad support due to its economic and social advantages. However, the wealth of sensor measurements pose major technical challenges on where to store the collected data, how to ensure its integrity, who control access permissions, and how to enable secure interaction between patients and medical facilities and professionals. This article aspires to provide a holistic solution based on blockchain technology. Our solution puts the patient in charge for granting and revoking access permissions and makes it easy for healthcare organizations and providers to meet privacy regulations. The sensor data are to reside on cloud storage, while access control and session logs are maintained on blockchain. In addition, a novel data-driven authentication and secure communication protocol is proposed to mitigate the risk of fraud and identity theft. In order to enforce such a protocol, all interactions between the cloud and patients and healthcare providers are regulated through smart contracts. The security properties of our solution are analyzed using AVISPA; it is also shown to be computationally efficient.

The DIAMONDS intervention to support self-management of type 2 diabetes in people with severe mental illness: Study protocol for a single-group feasibility study

Introduction.The DIAMONDS programme aims to evaluate a novel supported diabetes self-management intervention for people with severe mental illness (the “DIAMONDS intervention”). The purpose of this study is to test the feasibility of intervention delivery and data collection procedures to inform a definitive randomised controlled trial (RCT).Methods.Adults aged 18 years or over with a diagnosis of type 2 diabetes and severe mental illness (schizophrenia, schizoaffective disorder, or bipolar disorder) will be eligible for inclusion. Individuals with other types of diabetes or non-psychotic mental illness and those lacking capacity to consent will not be eligible. Participants will be recruited from NHS mental health trusts and general practices across the North of England. All participants will receive the DIAMONDS intervention: weekly one-to-one sessions with a trained facilitator (“DIAMONDS Coach”) to support goal setting, action planning, and diabetes education; ongoing self-management supported by a paper-based workbook and optional digital application (app); and monthly peer-support group sessions with other participants. The primary outcomes are: 1. Recruitment rate, measured as proportion of the recruitment target (N ​= ​30) achieved at 5 months from start of recruitment, 2. Attrition measured as the proportion of missing outcomes data at the end of the recruitment period (5 months from start of recruitment) for physiological and self-reported data items, 3. Intervention delivery rate recorded as the proportion of planned sessions delivered (measured by the number of completed intervention session logs per participant within 15 weeks of the first intervention session). Secondary outcomes include completeness of data collection at baseline and of process evaluation data at follow-up as well as the feasibility and acceptability of the intervention and of wearing a blinded continuous glucose monitoring device. An intervention fidelity framework will also be developed. Recruitment started in July 2021. The study was prospectively registered: ISRCTN15328700 (12th March 2021).Discussion.The results of this feasibility study will inform the refinement of the content and delivery of the DIAMONDS intervention, as well as research procedures, including recruitment and data collection, in preparation for the main DIAMONDS RCT.

Investigating the Experience of Boredom During Reading Sessions in the Foreign Language Classroom

Despite the fact that boredom appears to be one of the most commonly experienced emotions in school settings, this negative emotion remains vastly underappreciated in the field of SLA. This is the gap this article seeks to rectify by reporting the findings of a classroom-based study whose purpose was to investigate changes in the experience of boredom in an English language classroom during reading sessions. The sample consisted of 18 second-year students studying English at a Polish high school. The data were collected by means of session logs, observations and reading session plans. The gathered data were analysed quantitatively and qualitatively. The findings showed that the participants reported different levels of boredom over the course of single reading sessions and from one session to the next. Factors responsible for the detected variation in the levels of boredom were related, among other things, to inactivity, performance of too easy/difficult tasks, teacher’s decisions regarding choice and use of language materials, the design of the reading sessions or individual characteristics of the learner.

An SSH predictive model using machine learning with web proxy session logs

An SSH predictive model using machine learning with web proxy session logs

Application of the behaviour change technique taxonomy (BCTTv1) to an inclusive physical literacy-based sport program for children and youth

Coaches play a critical role in shaping the sporting experiences of children and youth. In introductory-level grassroots sport programs, the coaching strategies used can greatly influence children’s and youths’ physical literacy development. This study identified behaviour change techniques (BCTs) used by coaches in an inclusive physical literacy-based grassroots sport program for children and youth. Specifically, the study aimed to assess the type and frequency of BCTs reported by coaches in their weekly one-on-one coaching session logs. This study focused on a community-based program (Igniting Fitness Possibilities) that supports children and youth with varied abilities in becoming engaged in local physical activity opportunities. 162 weekly coaching logs from 13 coaches (n = 8 females) were coded using Michie and colleagues’ (2013) BCT Taxonomy (BCTTv1) consisting of 93 different BCTs. A codebook was developed in accordance with the BCTTv1 to assist in identifying BCTs used within the Igniting Fitness Possibilities context. Thirty-six BCTs were identified across the coaching logs, with an average of 3.62 (SD = 2.43) different BCTs coded per log. The most commonly coded BCTs were ‘demonstration of the behaviour’ (146 codings), ‘instruction on how to perform the behaviour’ (145 codings), and ‘behavioural practice/rehearsal’ (137 codings). The findings support the applicability of the BCTTv1 to a paediatric inclusive grassroots context and provide insight into BCTs that might be most relevant to target for physical literacy development in this setting. BCT identification can also promote reflective coaching practices, help identify gaps in training, and provide a foundation for evaluating the effectiveness of coaching strategies.

Microplastic Spectral Classification Needs an Open Source Community: Open Specy to the Rescue!

Microplastic pollution research has suffered from inadequate data and tools for spectral (Raman and infrared) classification. Spectral matching tools often are not accurate for microplastics identification and are cost-prohibitive. Lack of accuracy stems from the diversity of microplastic pollutants, which are not represented in spectral libraries. Here, we propose a viable software solution: Open Specy. Open Specy is on the web (www.openspecy.org) and in an R package. Open Specy is free and allows users to view, process, identify, and share their spectra to a community library. Users can upload and process their spectra using smoothing (Savitzky-Golay filter) and polynomial baseline correction techniques (IModPolyFit). The processed spectrum can be downloaded to be used in other applications or identified using an onboard reference library and correlation-based matching criteria. Open Specy's data sharing and session log features ensure reproducible results. Open Specy houses a growing library of reference spectra, which increasingly represents the diversity of microplastics as a contaminant suite. We compared the functionality and accuracy of Open Specy for microplastic identification to commonly used spectral analysis software. We found that Open Specy was the only open source software and the only software with a community library, and Open Specy had comparable accuracy to popular software (OMNIC Picta and KnowItAll). Future developments will enhance spectral identification accuracy as the reference library and functionality grows through community-contributed spectra and community-developed code. Open Specy can also be used for applications beyond microplastic analysis. Open Specy's source code is open source (CC-BY-4.0, attribution only) (https://github.com/wincowgerDEV/OpenSpecy).

Using remote learning to teach clinicians manual wheelchair skills: a cohort study with pre- vs post-training comparisons

Purpose To test the hypothesis that remote learning to teach clinicians manual wheelchair skills is efficacious. Materials and methods A convenience sample of therapists (physical and occupational) and students were enrolled in pairs in a cohort study with pre- versus post-training comparisons. The intervention was a hybrid of self-study and hands-on practice paired with remote feedback for ten intermediate and advanced manual wheelchair skills. Participants practiced with self-selected frequency and duration, uploading a session log and video(s) to an online platform. A remote trainer provided asynchronous feedback prior to the next practice session. Capacity and confidence in completing the ten skills were evaluated using the Wheelchair Skills Test Questionnaire (WST-Q). Knowledge of wheelchair skills training and motor learning was assessed using a 62-item Knowledge Test. Secondary outcome measures included skill achievement, as confirmed by submitted video recordings, and participant feedback about the training. Results Across 41participants, scores were higher at follow-up compared to baseline for WST-Q capacity (73.9 ± 19.1 vs 16.8 ± 15.6, p < 0.001), WST-Q confidence (80.1 ± 12.2 vs 47.6 ± 18.2, p = 0.003) and knowledge (70.8 ± 7.5 vs 67.0 ± 5.4, p = 0.004). Conclusions Remote learning can increase wheelchair skills capacity and confidence as well as knowledge about such training and assessment. This model should be further investigated as a delivery method for training rehabilitation professionals. Clinical Trial Registration Number NCT01807728. Implications for rehabilitation Wheelchair skills training is one of the 8 steps of wheelchair provision as outlined by the World Health Organization. Wheelchair skills are not a core part of most clinical curriculums and many clinicians cite a lack of resources and uncertainty on how to implement wheelchair skills training into practice as major barriers to providing such training. Remote learning offers the benefits of structured wheelchair skills training with expert feedback on an individual’s own schedule that is not afforded by one-day “bootcamp”-type courses or on-the-job training, which are how many clinicians currently learn wheelchair skills. In a sample of physical and occupational therapists and students, remote learning was effective at increasing capacity and confidence to perform manual wheelchair skills as well as knowledge of wheelchair training.

Can intelligent agents improve data quality in online questiosnnaires? A pilot study.

We explored the utility of chatbots for improving data quality arising from collection via sonline surveys. Three-hundred Australian adults sampled via Prolific Academic were randomized across chatbot-supported or unassisted online questionnaire conditions. The questionnaire comprised validated measures, along with challenge items formulated to be confusing yet aligned with the validated targets. The chatbot condition provided optional assistance with item clarity via a virtual support agent. Chatbot use and user satisfaction were measured through session logs and user feedback. Data quality was operationalized as between-group differences in relationships among validated and challenge measures. Findings broadly supported chatbot utility for online surveys, showing that most participants with chatbot access utilized it, found it helpful, and demonstrated modestly improved data quality (vs. controls). Absence of confusion for one challenge item is believed to have contributed to an underestimated effect. Findings show that assistive chatbots can enhance data quality, will be utilized by many participants if available, and are perceived as beneficial by most users. Scope constraints for this pilot study are believed to have led to underestimated effects. Future testing with longer-form questionnaires incorporating expanded item difficulty may further understanding of chatbot utility for survey completion and data quality.

Feasibility of a multicomponent cognitive behavioral intervention for fear of falling after hip fracture: process evaluation of the FIT-HIP intervention

BackgroundThis study describes the process evaluation of an intervention developed to reduce fear of falling (FoF) after hip fracture, within an inpatient geriatric rehabilitation setting. This ‘FIT-HIP intervention’ is a multicomponent cognitive behavioral intervention, conducted by physiotherapists and embedded in usual care in geriatric rehabilitation in the Netherlands. A previous study (cluster randomized controlled trial) showed no beneficial effects of this intervention when compared to usual care. The aim of this study was to gain insight into factors related to the intervention process that may have influenced the effectiveness of the intervention.MethodsThis process evaluation was conducted using an observational prospective study design. Based on quantitative and qualitative data derived from session logs, evaluation questionnaires and interviews, we addressed: 1] recruitment and reach; 2] performance according to protocol; 3] patients’ adherence; and 4] opinions of patients and facilitators on the intervention. Participants in this study were: a) patients from 6 geriatric rehabilitation units, who were invited to participate in the intervention (39 adults aged ≥65 years with hip fracture and FoF) and; b) intervention facilitators (14 physiotherapists and 8 psychologists who provide coaching to the physiotherapists).ResultsThirty-six patients completed the intervention during inpatient geriatric rehabilitation. Apart from cognitive restructuring and telephonic booster (which was not provided to all patients), the intervention was performed to a fair degree in accordance with protocol. Patients’ adherence to the intervention was very good, and patients rated the intervention positively (average 8.1 on a scale 0–10). Although most facilitators considered the intervention feasible, a limited level of FoF (possibly related to timing of intervention), and physiotherapists’ limited experience with cognitive restructuring were identified as important barriers to performing the intervention according to protocol.ConclusionsThe FIT-HIP intervention was only partly feasible, which may explain the lack of effectiveness in reducing FoF. To improve the intervention’s feasibility, we recommend selecting patients with maladaptive FoF (i.e. leading to activity restriction), being more flexible in the timing of the intervention, and providing more support to the physiotherapists in conducting cognitive restructuring.Trial registrationNetherlands Trial Register: NTR5695 (7 March 2016).

Towards early purchase intention prediction in online session based retailing systems

Purchase prediction has an important role for decision-makers in e-commerce to improve consumer experience, provide personalised recommendations and increase revenue. Many works investigated purchase prediction for session logs by analysing users’ behaviour to predict purchase intention after a session has ended. In most cases, e-shoppers prefer to be anonymous while browsing the websites and after a session has ended, identifying users and offering discounts can be challenging. Therefore, after a session ends, predicting purchase intention may not be useful for the e-commerce strategists. In this work, we propose and develop an early purchase prediction framework using advanced machine learning models to investigate how early purchase intention in an ongoing session can be predicted. Since users could be anonymous, this could help to give real-time offers and discounts before the session ends. We use dynamically created session features after each interaction in a session, and propose a utility scoring method to evaluate how early machine learning models can predict the probability of purchase intention. The proposed framework is validated with a real-world dataset. Computational experiments show machine learning models can identify purchase intention early with good performance in terms of Area Under Curve (AUC) score which shows success rate of machine learning models on early purchase prediction.