Human-Computer Interaction (HCI) research aims to make systems versatile, easy to use, and accessible for most people. While computer technologies have successfully remodeled and improved the learning process, students with vision impairment (SVI) are at a disadvantage due to the lack of accessibility to their learning tools. This research aims to present the tools and studies that promote self-learning in mathematics for students who are blind or have low vision. First, we present the initial work in MyAccessible Math, an open-source web platform that integrates user-centered design principles for SVIs to learn and practice mathematics. The prototype integrates speech recognition and text-to-speech libraries to provide personalized real-time feedback. We evaluated interaction modalities in the prototype using experimental research designs with visually impaired elementary, middle, and high school students from Alabama Institute for the Deaf and Blind (AIDB) and Alabama School for the Blind (ASB). We recognize the limitations of interaction features in the prototype and propose further improvements. Then, we transform MyAccessible Math into a fully functional cloud-based system. We introduce the math evaluator that generates step-by-step solutions to math expressions. We incorporate speech recognition libraries with Natural Language Toolkit (NLTK) to answer open-ended questions related to mathematics. Finally, we evaluate improved interaction and motivational techniques in MyAccessible Math with five individuals to assess the system's usability. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]