A Survey of Accuracy Evaluation Metrics of Recommendation Tasks

By Asela Gunawardana, Guy Shani. 2009

715 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems

Citations in the corpus

Explainable Recommendation: A Survey and New Perspectives (2018, methodology): “The quality of the free-text explanations can also be evaluated in terms of readability based on frequently used readability measures, such as Gunning Fog Index (Gunning [1952]), Flesch Reading Ease (Flesch [1948]), Flesch Kincaid Grade Level (Kincaid et al.”
Collaborative Filtering Recommender Systems (2011, methodology): “Time-averaged RMSE [85] requires that predictions be performed in temporal sequence using only ratings before the current time t, as discussed by Gunawardana and Shani [53], and computes the RMSE over all predictions generated up to time t.”
Recommender systems survey (2013, background): “The most commonly used quality measures are the following [90,95]: (1) prediction evaluations, (2) evaluations for recommendation as sets, and (3) evaluations for recommendations as ranked lists.”
Research-paper recommender systems: a literature survey (2015, background): “Gunawardana and Shani published a survey about accuracy metrics [224].”
A comparative analysis of offline and online evaluations and discussion of research paper recommender system evaluation (2013, background): “In addition, some arguments have been voiced that offline-evaluations are not adequate to evaluate recommender systems [6,9,17].”
List-wise learning to rank with matrix factorization for collaborative filtering (2010, background): “The ultimate purpose of a recommender system is to provide users with a ranking or recommendation list and this objective has been argued to be more crucial than the accuracy of rating prediction [1][5][7][13].”
Evaluating recommender systems (2012, background): “An example is usage predictions [1]which mentioned above.”
Efficiency Improvement of Neutrality-Enhanced Recommendation (2013, methodology): “We formalize the task of information-neutral recommendation and present an algorithm for performing this task.”
Sequential User-based Recurrent Neural Network Recommendations (2017, background): “To adequately reflect user interests, functions that rank items according to some notion of utility are today considered most suitable for generating personalized recommendations [14].”