Inverse Foraging

Users' engagement with pervasive displays has been extensively studied, however, determining how their content is interesting remains an open problem. Tracking of body postures and gaze has been explored as an indication of attention; still, existing works have not been able to estimate the interest of passers-by from readily available data, such as the display viewing time. This article presents a simple yet accurate method of estimating users' interest in multiple content items shown at the same time on displays. The proposed approach builds on the information foraging theory, which assumes that users optimally decide on the content they consume. Through inverse foraging, the parameters of a foraging model are fitted to the values of viewing times observed in practice, to yield estimates of user interest. Different foraging models are evaluated by using synthetic data and with a controlled user study. The results demonstrate that inverse foraging accurately estimates interest, achieving an R2 above 70% in comparison to self-reported interest. As a consequence, the proposed solution allows to dynamically adapt the content shown on pervasive displays, based on viewing data that can be easily obtained in field deployments.

How do Web-Active End-User Programmers Forage?

Web-active end-user programmers spend substantial time and cognitive effort seeking information while debugging web mashups, which are platforms for creating web applications by combining data and functionality from two or more different sources. The debugging on these platforms is challenging as end user programmers need to forage within the mashup environment to find bugs and on the web to forage for the solution to those bugs. To understand the foraging behavior of end-user programmers when debugging, we used information forging theory. Information foraging theory helps understand how users forage for information and has been successfully used to understand and model user behavior when foraging through documents, the web, user interfaces, and programming environments. Through the lens of information foraging theory, we analyzed the data from a controlled lab study of eight web-active end-user programmers. The programmers completed two debugging tasks using the Yahoo! Pipes web mashup environment. On analyzing the data, we identified three types of cues: clear, fuzzy, and elusive. Clear cues helped participants to find and fix bugs with ease while fuzzy and elusive cues led to useless foraging. We also identified the strategies used by the participants when finding and fixing bugs. Our results give us a better understanding of the programming behavior of web-active end-users and can inform researchers and professionals how to create better support for the debugging process. Further, this study methodology can be adapted by researchers to understand other aspects of programming such as implementing, reusing, and maintaining code.

IoT-based group size prediction and recommendation system using machine learning and deep learning techniques

In an open source software development environment, it is hard to decide the number of group members required for resolving software issues. Developers generally reply to issues based totally on their domain knowledge and interest, and there are no predetermined groups. The developers openly collaborate on resolving the issues based on many factors, such as their interest, domain expertise, and availability. This study compares eight different algorithms employing machine learning and deep learning, namely—Convolutional Neural Network, Multilayer Perceptron, Classification and Regression Trees, Generalized Linear Model, Bayesian Additive Regression Trees, Gaussian Process, Random Forest and Conditional Inference Tree for predicting group size in five open source software projects developed and managed using an open source development framework GitHub. The social information foraging model has also been extended to predict group size in software issues, and its results compared to those obtained using machine learning and deep learning algorithms. The prediction results suggest that deep learning and machine learning models predict better than the extended social information foraging model, while the best-ranked model is a deep multilayer perceptron((R.M.S.E. sequelize—1.21, opencv—1.17, bitcoin—1.05, aseprite—1.01, electron—1.16). Also it was observed that issue labels helped improve the prediction performance of the machine learning and deep learning models. The prediction results of these models have been used to build an Issue Group Recommendation System as an Internet of Things application that recommends and alerts additional developers to help resolve an open issue.