Bumblebee friendly planting recommendations with citizen science data

Several citizen science projects engage with the public around pollinator species, typically requesting data (e.g. in the form of photo-records of different species tagged by place and date). While such projects help scientists collect data, these data are rarely fed back to the public in any meaningful manner. In this paper, we address this through a recommender system based on Matrix Factorization over a matrix of observed bumblebee-plant interactions derived from data submitted to a citizen science project BeeWatch. The system recommends pollinator-friendly plants for domestic gardens and takes into account both the fact that different bumblebee species exhibit differing preferences for flowers, and that plants flower at different times of the year. The goal is to attract a range of bumblebee species to a garden and to ensure that these species have sufficient food sources through the season.

[1]  Pasquale Lops,et al.  Content-based Recommender Systems: State of the Art and Trends , 2011, Recommender Systems Handbook.

[2]  C. Mellish,et al.  The role of automated feedback in training and retaining biological recorders for citizen science , 2016, Conservation biology : the journal of the Society for Conservation Biology.

[3]  George Karypis,et al.  A Comprehensive Survey of Neighborhood-based Recommendation Methods , 2011, Recommender Systems Handbook.

[4]  H. Sebastian Seung,et al.  Algorithms for Non-negative Matrix Factorization , 2000, NIPS.

[5]  Jon Rosewell,et al.  Crowdsourcing the identification of organisms: A case-study of iSpot , 2015, ZooKeys.

[6]  J. Bobadilla,et al.  Recommender systems survey , 2013, Knowl. Based Syst..

[7]  Shuang-Hong Yang,et al.  Functional matrix factorizations for cold-start recommendation , 2011, SIGIR.

[8]  Candie C. Wilderman,et al.  Public Participation in Scientific Research: a Framework for Deliberate Design , 2012 .

[9]  Chris Mellish,et al.  Mapping species distributions: A comparison of skilled naturalist and lay citizen science recording , 2015, Ambio.

[10]  H. Sebastian Seung,et al.  Learning the parts of objects by non-negative matrix factorization , 1999, Nature.

[11]  John Riedl,et al.  Collaborative Filtering Recommender Systems , 2011, Found. Trends Hum. Comput. Interact..

[12]  Yehuda Koren,et al.  The BellKor Solution to the Netflix Grand Prize , 2009 .

[13]  Mehrbakhsh Nilashi,et al.  Collaborative filtering recommender systems , 2013 .

[14]  Michael W. Berry,et al.  Algorithms and applications for approximate nonnegative matrix factorization , 2007, Comput. Stat. Data Anal..

[15]  Roy A. Sanderson,et al.  Quantifying and comparing bumblebee nest densities in gardens and countryside habitats , 2007 .

[16]  Yehuda Koren,et al.  Matrix Factorization Techniques for Recommender Systems , 2009, Computer.

[17]  Justin Dillon.  50 Years of JBE: From Science and Environmental Education to Civic Science , 2016 .

[18]  Greg Linden,et al.  Amazon . com Recommendations Item-to-Item Collaborative Filtering , 2001 .

[19]  Chris Mellish,et al.  Natural Language Generation for Nature Conservation: Automating Feedback to Help Volunteers Identify Bumblebee Species , 2012, COLING.

[20]  Taghi M. Khoshgoftaar,et al.  A Survey of Collaborative Filtering Techniques , 2009, Adv. Artif. Intell..

[21]  Federica Cena,et al.  The impact of rating scales on user's rating behavior , 2011, UMAP'11.

[22]  R. Plemmons,et al.  Optimality, computation, and interpretation of nonnegative matrix factorizations , 2004 .

[23]  J. Dillon.  On the Convergence Between Science and Environmental Education , 2018 .