DEVELOPING CROPS THAT MEET THE GROWING NEEDS OF A RISING POPULATION IN A CHANGING CLIMATE REQUIRES INCREASING OUR UNDERSTANDING OF HOW PLANTS FUNCTION. RESEARCHERS CAN LEARN ABOUT PLANTS BY GROWING THEM AND MAKING OBSERVATIONS, BUT THERE IS A DISCONNECT BETWEEN RESEARCHERS DESCRIBING WHAT THEY SEE, AND HOW COMPUTER SYSTEMS INTERPRET THAT INFORMATION IN A WAY THAT IS USEFUL AND LEADS TO NEW DISCOVERIES. THIS PROJECT ADDRESSES THIS DISCONNECT BY DEVELOPING METHODS AND TOOLS THAT WORK DIRECTLY ON RESEARCHERS' OWN DESCRIPTIONS, AND USE THEM TO PREDICT NEW BIOLOGICAL DISCOVERIES.TO DO THIS, WE WILL EVALUATE A NUMBER OF EXISTING APPROACHES FOR HOW TO COMPUTATIONALLY INTERPRET LANGUAGE, AS WELL AS DEVELOPING OUR OWN, TO UNDERSTAND HOW USEFUL THEY ARE FOR INTERPRETING DESCRIPTIONS OF PLANTS PROVIDED BY RESEARCHERS. THE MOST USEFUL METHODS WILL GO INTO A WEB TOOL THAT RESEARCHERS CAN EASILY USE TO PROVIDE DESCRIPTIONS OF THEIR OWN PLANTS AND MAKE ASSOCIATIONS BETWEEN WHAT THEY ARE OBSERVING AND THE UNDERLYING BIOLOGY. GIVING RESEARCHERS ACCESS TO SUCH A TOOL WILL HELP FACILITATE NEW BIOLOGICAL DISCOVERIES THAT ULTIMATELY RESULT IN THE IMPROVEMENT OF CROPS TO HELP SUPPLY THE GENERAL PUBLIC WITH FOOD, FUEL, AND MATERIALS.