The effect of pets on happiness: A data-driven approach via large-scale social media

Abstract

Abstract-Psychologists have demonstrated that pets have a positive impact on owners' happiness. For example, lonely people are often advised to have a dog or cat to quell their social isolation. Conventional psychological research methods of analyzing this phenomenon are mostly based on surveys or self-reported questionnaires, which are time-consuming and lack of scalability. Utilizing social media as an alternative and complimentary resource could potentially address both issues and provide different perspectives on this psychological investigation. In this paper, we propose a novel and effective approach that exploits social media to study the effect of pets on owners' happiness. The proposed framework includes three major components: 1) collecting user-level data from Instagram consisting of about 300,000 images from 2905 users; 2) constructing a convolutional neural network (CNN) for pets classification, and combined with timeline information, further identifying pet owners and the control group; 3) measuring the confidence score of happiness by detecting and analyzing selfie images. Furthermore, various factors of demographics are employed to analyze the fine-grained effects of pets on happiness. Our experimental results demonstrate the effectiveness of the proposed approach and we believe that this approach can be applied to other related domains as a large-scale, high-confidence methodology of user activity analysis through social media.

The Grass is Greener on the Other Side

“…For example, instead of administering surveys to understand how a specific crisis or natural disaster affects people's emotional well-being, we can perform sentiment analysis on a large amount of tweets that are posted in close proximity to the natural disaster or by users residing near the natural disaster. In future, we intend to extend our study to utilize image recognition techniques alongside sentiment analysis on photo-sharing sites, similar to the studies on pet ownership and alcohol consumption using Instagram [30,51].…”

Sleep-deprived fatigue pattern analysis using large-scale selfies from social media

“…Some of our findings are beyond those reported in the literature, e.g., those related to the interplays of age, gender and race, pointing to the potential to discover factors that the empirical studies have overlooked. Our work is in the same vein as [3][17] [18], and such social mediadriven methods are expected to find more successes in computational sociology and psychology.…”

The Effect of Pets on Happiness: A Large-scale Multi-Factor Analysis using Social Multimedia

“…In contrast, the work in [23] analyzed 300,000 posts from around 2900 users, while our study analyzes around 2-million posts from roughly 20,000 Instagram users. [23] has constructed a reasonable convolutional neural network (CNN) as its pet classifier while we build a high-performance classifier by retraining the final layer of the Inception v3 model using remarkably fewer training samples. In addition, [23] assumed that the largest face in a selfie post is the face of the user, and analyzed the average happiness of this user according to all of those "largest" faces throughout this user's timeline posts.…”

“…Our study is inspired by the preliminary research in [23] but we extend the analyses in four distinguishable ways: 1) using nearly ten-times more data, 2) retraining the final layer of the Inception model v3 as our pet classifier, which achieves a superior accuracy, 3) improving the method of identifying user faces in user timeline posts, and most importantly, 4) considering multiple factors including user's relationship status and if having children. In contrast, the work in [23] analyzed 300,000 posts from around 2900 users, while our study analyzes around 2-million posts from roughly 20,000 Instagram users. [23] has constructed a reasonable convolutional neural network (CNN) as its pet classifier while we build a high-performance classifier by retraining the final layer of the Inception v3 model using remarkably fewer training samples.…”