Tracing and Predicting Collaboration for Junior Scholars

23rd International Conference on Intelligent User Interfaces

Brusilovsky

2018

Self Cite

The beyond-relevance objectives of recommender systems have been drawing more and more attention. For example, a diversity-enhanced interface has been shown to associate positively with overall levels of user satisfaction. However, little is known about how users adopt diversity-enhanced interfaces to accomplish various real-world tasks. In this paper, we present two attempts at creating a visual diversity-enhanced interface that presents recommendations beyond a simple ranked list. Our goal was to design a recommender system interface to help users explore the different relevance prospects of recommended items in parallel and to stress their diversity. Two within-subject user studies in the context of social recommendation at academic conferences were conducted to compare our visual interfaces. Results from our user study show that the visual interfaces significantly reduced the exploration efforts required for given tasks and helped users to perceive the recommendation diversity. We show that the users examined a diverse set of recommended items while experiencing an improvement in overall user satisfaction. Also, the users' subjective evaluations show significant improvement in many user-centric metrics. Experiences are discussed that shed light on avenues for future interface designs.

“…(1) The Academic feature is determined by the degree of publication similarity between two attendees using cosine similarity [43]. The function is defined as:…”

Section: First Attempt: Scatter Vizmentioning

confidence: 99%

“…We retrieve longitude and latitude data based on attendees' affiliation information. We used the Haversine formula to compute the geographic distance between any pair of attendees [43].…”

Section: First Attempt: Scatter Vizmentioning

confidence: 99%

Beyond the Ranked List

23rd International Conference on Intelligent User Interfaces

Brusilovsky

2018

Self Cite

“…e RelExplorer uses three separate recommender engines that suggest co-a endees to meet on the basis of: 1) e similarity of past publications (Academic feature); 2) Social network distance (Social feature); and 3) Similarity of interests (Interest feature). Academic Feature: e academic feature is determined by publication similarity between two a endees using cosine similarity [13,22]. e function is de ned as:…”

Section: Recommendation Componentsmentioning

confidence: 99%

Providing Control and Transparency in a Social Recommender System for Academic Conferences

Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization

Brusilovsky

2017

Self Cite

A social recommender system aims to provide useful suggestion to the user and prevent social overload problem. Most of the research e orts are spent on push high relevant item on top of the ranked list, using a weight ensemble approach. However, we argue the "learned" static fusion is not enough to speci c contexts. In this paper, we develop a series visual recommendation components and control panel for the user to interact with the recommendation result of an academic conference. e system o ers a be er recommendation transparency and user-driven fusion through recommended sources. e experiment result shows the user did fuse the di erent recommended sources and exploration pa erns among tasks. e post-study survey is positively associated with the system and explanation function e ectiveness. is nding shed light on the future research of design a recommender system with human intervention and the interface beyond the static ranked list.

“…4) Geographic Distance (GD) [14]: The GD is used to measure the actual geographic distance between two businesses. We used the Haversine formula to compute the geographic distance between two points on earth, based on longitude and latitude data.…”

Section: Review Networkmentioning

confidence: 99%

“…The goal of this study is to predict the likelihood that two businesses will attract the same user's review in the future. The key to fulfilling such a prediction task is the predictor selection [14]. This study further considers various prediction features, including geographical distance, reviewing network, the fuzzy-businesses vector, and content similarity.…”

Section: Introductionmentioning

confidence: 99%

A Fuzzy-Based Personalized Recommender System for Local Businesses

Proceedings of the 27th ACM Conference on Hypertext and Social Media

2016

Self Cite

On-line reviewing systems have become prevalent in our society. User-provided reviews of local businesses have provided rich information in terms of users' preferences regarding businesses and their interactions in reviewing systems; however, little is known about how the reviewing behaviors of users can benefit businesses in terms of suggesting potential collaboration opportunities. In the current study, we aim to build a recommendation system for businesses to provide suggestions for business collaboration. Based on historical data from Yelp that shows two businesses being reviewed by the same users within a same season, we were able to identify businesses that might attract the same customers in the future, and hence provide them with a collaboration suggestion. Our results suggest that the evidence-two businesses sharing reviews from same users-can provide recommendations for businesses to pursue future collaborative marketing opportunities.