Thread by @kashtodi

How can adaptive interfaces and #HCI benefit from #AI and reinforcement learning?
🧵 A thread on our #CHI2021 paper w. @gilles_bailly @luileito @oulasvirta
🌐 Project page: userinterfaces.aalto.fi/adaptive
🎥 Watch the video:
👇👇 @AaltoResearch @sig_chi @sigchi

The secret sauce is to make decisions via #planning 🧠 Remember how AlphaGo by @DeepMind could plan moves and consistently win at Go? Turns out you can use similar #RL methods for HCI applications too📱💻

One such promising case is where UIs #adapt automatically to users 🔄 But what does a "win" even mean here? How could the system tell whether the decisions it is making are actually good?🤔

We suggest that #predictive models can be used to evaluate costs and benefits of any changes – or in other words, the #reward. HCI literature is rich with #models for GUIs, pointing, etc. We extend #menu search models to predict how changes influence selection time⏱

We can now use Monte-Carlo Tree Search (#MCTS) for planning! Self-adapting menus can efficiently simulate several sequences of adaptations and plan changes 🤖 We also develop a value network for speeding up computations to address larger problem sizes and online use 🚀

In contrast to myopic/greedy changes, this approach encourages long-term improvements to usability through carefully planned adaptations 💚 Our first technical + empirical evaluations show improvements over static and adaptive baselines, and offer promising evidence ✅

We hope our work sparks further exploration in this exciting area that combines #HCI and #AI 💪🦾

To encourage it, and in the #openscience spirit:
📝 Paper: userinterfaces.aalto.fi/adaptive/resou…
📂 Code: github.com/aalto-ui/chi21…

Stop by during the #CHI2021 sessions: programs.sigchi.org/chi/2021/progr…