Agha Ali Raza

Publications/Speech-Based Interfaces & ICT for Development (ICT4D)

Journal · 2021

Personalized Weather Information for Low-Literate Farmers using Multimodal Dialog Systems

Muhammad Qasim, Haris Bin Zia, Awais Athar, Tania Habib, Agha Ali Raza

International Journal of Speech Technology, vol. 24(2), 2021

Summary

Speech interfaces over basic phones can bring information to people who cannot read, but accurate speech recognition alone does not make a service usable. A weather information line for farmers in Pakistan had only a 37.8% task success rate despite a very accurate recognizer. Through iterative design, we added multimodal input, adaptation to the user, and context-specific help. Deployed over 10 months with 23,997 calls from 6,893 users, the redesigned service reached a 96.3% success rate and went on to become Pakistan's national weather hotline.

Cite this paper

Muhammad Qasim, Haris Bin Zia, Awais Athar, Tania Habib, Agha Ali Raza. Personalized Weather Information for Low-Literate Farmers using Multimodal Dialog Systems. International Journal of Speech Technology, vol. 24(2), 2021.

@article{qasim2021personalized,
  author = {Muhammad Qasim and Haris Bin Zia and Awais Athar and Tania Habib and Agha Ali Raza},
  title = {{Personalized Weather Information for Low-Literate Farmers using Multimodal Dialog Systems}},
  journal = {International Journal of Speech Technology},
  volume = {24},
  number = {2},
  year = {2021},
  url = {https://aghaaliraza.com/publications/2021--IJST--Personalized_weather_information_for_low-literate_farmers_using_multimodal_dialog_systems.pdf}
}

More on this theme

All 14 works on this theme →