Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2409.15441 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913513973219328 |
|---|---|
| author | Tang, Brian Shin, Kang G. |
| author_facet | Tang, Brian Shin, Kang G. |
| contents | Recently, large language models (LLMs) have demonstrated exceptional capabilities in serving as the foundation for AI assistants. One emerging application of LLMs, navigating through websites and interacting with UI elements across various web pages, remains somewhat underexplored. We introduce Steward, a novel LLM-powered web automation tool designed to serve as a cost-effective, scalable, end-to-end solution for automating web interactions. Traditional browser automation frameworks like Selenium, Puppeteer, and Playwright are not scalable for extensive web interaction tasks, such as studying recommendation algorithms on platforms like YouTube and Twitter. These frameworks require manual coding of interactions, limiting their utility in large-scale or dynamic contexts. Steward addresses these limitations by integrating LLM capabilities with browser automation, allowing for natural language-driven interaction with websites. Steward operates by receiving natural language instructions and reactively planning and executing a sequence of actions on websites, looping until completion, making it a practical tool for developers and researchers to use. It achieves high efficiency, completing actions in 8.52 to 10.14 seconds at a cost of $0.028 per action or an average of $0.18 per task, which is further reduced to 4.8 seconds and $0.022 through a caching mechanism. It runs tasks on real websites with a 40% completion success rate. We discuss various design and implementation challenges, including state representation, action sequence selection, system responsiveness, detecting task completion, and caching implementation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_15441 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Steward: Natural Language Web Automation Tang, Brian Shin, Kang G. Artificial Intelligence Recently, large language models (LLMs) have demonstrated exceptional capabilities in serving as the foundation for AI assistants. One emerging application of LLMs, navigating through websites and interacting with UI elements across various web pages, remains somewhat underexplored. We introduce Steward, a novel LLM-powered web automation tool designed to serve as a cost-effective, scalable, end-to-end solution for automating web interactions. Traditional browser automation frameworks like Selenium, Puppeteer, and Playwright are not scalable for extensive web interaction tasks, such as studying recommendation algorithms on platforms like YouTube and Twitter. These frameworks require manual coding of interactions, limiting their utility in large-scale or dynamic contexts. Steward addresses these limitations by integrating LLM capabilities with browser automation, allowing for natural language-driven interaction with websites. Steward operates by receiving natural language instructions and reactively planning and executing a sequence of actions on websites, looping until completion, making it a practical tool for developers and researchers to use. It achieves high efficiency, completing actions in 8.52 to 10.14 seconds at a cost of $0.028 per action or an average of $0.18 per task, which is further reduced to 4.8 seconds and $0.022 through a caching mechanism. It runs tasks on real websites with a 40% completion success rate. We discuss various design and implementation challenges, including state representation, action sequence selection, system responsiveness, detecting task completion, and caching implementation. |
| title | Steward: Natural Language Web Automation |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2409.15441 |