Recipient Organization
UNIV OF IDAHO
875 PERIMETER DRIVE
MOSCOW,ID 83844-9803
Performing Department
(N/A)
Non Technical Summary
Data science is an integrated part of 21st-century agriculture, yet the educational needs have not been adequately met in schools and universities. The SUPERMinDS project goals are to train the next generation of professionals with data science skills and fill key gaps in agricultural data and information.To achieve these goals, three diverse student groups are met with three sets of immersive research and extension activities. Out of the 5-year total of 72 SUPERMinDS student participants, 20 students will be recruited from College of Southern Idaho, Idaho's first Hispanic-Serving Institution; 24 from Randolph College, a former women's college; and 28 from the University of Idaho main campus.The three project objectives dictate that SUPERMinDS activities should encompass training in all aspects of data science with directly applicable results for Idaho agriculture: students involved in data-gathering activities will engage with local stakeholders and produce up-to-date enterprise budgets for major Idaho crops; students involved in data analytics activities will utilize big data analytics techniques to produce real-time forecasts, what-if calculators, and risk management strategies for Idaho farmers and ranchers; and students involved in data management activities will use a variety of coding software to manage the gathered data, visualize results from data analytics, and design user-friendly website.The diverse group of mentors will leverage their past success to gauge interest from target students and work closely with the project evaluator to measure short-term learning and long-term metrics, such as student placement and where the experiential learning occurs. The long-term sustainability of the program is maintained through industry partners, directed studies, and capstone projects.
Animal Health Component
0%
Research Effort Categories
Basic
20%
Applied
80%
Developmental
0%
Goals / Objectives
The goals of this Research and Extension Experiences for Undergraduates (REEU) project are to train the next generation of professionals with data science skills (including data gathering, data analytics, and data management), and fill key gaps in agricultural data and information. Furthermore, such training shall be sustained in long-term agricultural education.To achieve these goals, we developed three objectives for SUPERMinDS student participants:Objective 1. Faculty at the University of Idaho (UI) Twin Falls Research and Extension Center (TFREC) will instruct undergraduate students on data gathering and database-building activities for generating enterprise budgets for Idaho's main crops. Students will also learn how to disseminate the gathered information via Extension research reports, fact sheets, and associated presentations and videos.Objective 2. Faculty on the UI main campus in Moscow, Idaho, will instruct undergraduates on data analytics activities such as empirical estimation and forecasting and building web-based decision aid tools.Objective 3. Our collaborator at Randolph College will lead training on digital data management for undergraduates. Specifically, mathematics/computer science students will develop websites and database management systems that use algorithms to automatically gather and store publicly available agriculture-related data (e.g., prices, production). The final product that encompasses all these components will be a website we call RAPIDS (Real-time Analytics Programming Information Dissemination System), on which we will disseminate outputs, host the decision aid tools, and link to an REEU student recruitment and project outcome website. The novelty of this project lies not only in the immersive research and Extension experiences for undergraduate students but also in a new form of collaboration: we will both train agricultural students with new technological tools from STEM disciplines and STEM students with new applications in agriculture.
Project Methods
Efforts:Activity 1: Data Gathering (for Objective 1).A1.1 Enterprise budget basics: include introducing students to the basic concept of enterprise budgets and their main cost and return components in guest lectures in relevant CSI Agriculture courses. After these guest lectures, students will be invited to apply to officially join SUPERMinDS as an Agricultural Data Associate. After application review by Dr. Hatzenbuehler and Dr. Quesnell of CSI, students will be selected to form an Agricultural Data Associate team.A1.2 Data gathering strategies and methods: will include implementing more formal training in data-gathering strategies and methods based on the existing UI Extension curriculum on enterprise budget development that is currently delivered to agricultural producers throughout the state in Farm Management courses. This curriculum includes a detailed discussion of the budget cost and returns components, identification of data sources for each cost and return variable, strategies for sampling among multiple sources, and the discussion of the usefulness of indexes from secondary sources for obtaining variable estimates.A1.3 Data gathering implementation: involves actual data gathering by engaging with local stakeholders to obtain agricultural data via email, phone calls, and/or in-person meetings. To implement this aspect, each Agricultural Data Associate will develop a Data Gathering Plan, in which they identify whether primary or secondary data are needed, potential sources of data, and strategies for obtaining the data. Data Gathering Plans will be reviewed by Dr. Hatzenbuehler and feedback will be provided after review. Students will then implement the data-gathering activities as consistent with the approved plan. The gathered data will be encompassed in a database that is shared among all Agricultural Data Associates and project partners.A1.4 Information dissemination:this is disseminating the data and information through publishing reports, filming associated videos--in which the students describe how they developed the enterprise budgets and how they are different from older budgets--and delivering presentations at relevant UI Extension programs. Dr. Hatzenbuehler will facilitate the travel for students to UI Extension programs.Activity2: Data Analytics (for Objective 2).A2.1 Forecasting and big data analytics:The training will provide web-based analytical results using state-of-art tools such as machine learning-based forecasting for SUPERMinDS project participants. The algorithm will automatically update results as new data become available.The analysis will use big data in all forms: students will utilize publicly available data about the weather, demographics, economics, geography, production, and public opinion from multiple sources including data gathered from data-gathering activities.A2.2 Risk and risk management analysis:Utilizing enterprise budget information gathered in Objective 1, students are able to conduct research on comparing costs of production in Idaho relative to nationally representative benchmark data and help farmers be more confident in making marketing and production decisions about when to cut losses or expand a certain product. This study is especially useful in labor-intensive operations in Idaho where, due to fluctuation in labor availability, uncertain financing costs, and variation in other input costs, there is a growing trend of automation, which introduces new uncertainties in cost structure for farms of all sizes. From these baseline results, Du will alsocollaborate with SUPERMinDS students to do research on developing: (i) optimal risk management strategies for farms of varied sizes and (ii) indicators of the financial performance of various enterprises across farms of varied sizes. Examples of the latter include analysis by crop acreage, total farm acreage, the magnitude of sales, and livestock operations, by herd size. Simulation analyses will use these indicators as a base level to evaluate the likelihood of triggering policy responses from crop insurance or government support programs.A2.3 Information dissemination:this is co-building theRAPIDSwebsite in conjunction with students' activities in Data Management. Leveraging the UI AgBiz website, the project team will build the dashboard-style webpageRAPIDShosted under the UI AgBiz website.RAPIDSincludes a web-based interactive information dashboard where data visualization will be provided and can be customized according to users' needs. TheRAPIDSwebpage is expected to be deployed in January 2024. Example tools that will be hosted withinRAPIDSinclude data visualization for important production and market information for main Idaho crops, real-time forecasts of market conditions, and a "what-if" scenario simulator for issues like yield losses and insurance coverage.Activity3: Data Management (for Objective 3).Dr. Wan will lead data management activities.Students involved in this activity will focus on:A3.1 Conceptual Training and Coding Camp with USDA database:We will use it as the training for python script for data scraping, USDA database introduction, and time-series data basics.A3.2 USDA API time-series data training:a) work with Python packages such as Pandas, NumPy, and Plotly and b) data filtering and virtualization with Python and SQLite.A3.3 Automation with Cloud Services: a) researching cloud service solutions; b) getting data plotted automatically; c) redesigning and implementing data visualization with tools such as D3 and RShiny; d) researching the efficiency of data plotting for the above methods.A3.4 Frontend Redesign and Implementation: a) learning web page design basics: HTML, CSS, js, and D3; and b) redesigning frontend features such as smooth pop-out windows and intuitive navigation.A3.5 Big Data Algorithm Designand Implementation: a) learning introductory coding examples and linear interpolation methodsto manage missing data in R; b) being introduced to coding examples, such as the Holt-Winter exponential smoothing model in R for time-series data forecasting; c) being introduced to machine learning algorithms such as ARIMA for time series data smoothing; d) developing new features and/or adjustments as the other SUPERMinDS team members develop their study based on interactions with farmers and ranchers at Extension programs.Project EvaluationNav Ghimire will serve as the project evaluator. He will conduct a formative evaluation annually to assess the program outcomes. The data from the formative assessment will serve two purposes: to report project progress and to improve the training curriculum and programs throughout SUPERMinDS project implementation. Ghimire will conduct a summativeevaluation after the completion of the project to develop a final report required to submit to the funder. The objectives of the project and the following evaluative questions will serve as aframework to inform the design and implementation of the evaluation plan: a) To what extent does the program improves students' knowledge and skills about data science and itsapplications in the agricultural field, 2) To what extent does the program fill gaps in the agricultural data. 3) To what extent does the program attract STEM students pursuing careers inagriculture. We will use qualitative, quantitative, and mixed methods to collect evaluation data on the project's outcomes, success, and challenges. The evaluation audience includes students, mentors, stakeholders, and others interested in such student activities.SUPERMinDS student participants will be tracked using the following strategy: 1) before the exit, students will fill out a form that includes their mailing address, personal email, and phone number, their employment industry and/or graduate school, and 2) they will join a social media group named "SUPERMinDS Alumni."