Abstract
Search engines and virtual assistants receive a large number of queries, and also report a large number of issues such as user queries that didn’t work, triggered the wrong features, or were not understood due to speech recognition errors, etc. This disclosure presents techniques that cluster similar issues, identify issues with the largest impact, and identify the root cause of an issue. The techniques scale easily, detect large patterns of similar issues, and prevent one-off fixes that need repeated application across similar issues. The techniques help improve the search engine or virtual assistant to provide responses that are more reliable, accurate, and satisfactory to the user.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Pilar, Petr; Zuniga, Mike; Yakovlev, Vladimir; Balle, Daniel; Kalb, Norbert; Wenger, Stephan; and Baeuml, Martin, "Automatic Issue Identification and Clustering", Technical Disclosure Commons, (May 01, 2020)
https://www.tdcommons.org/dpubs_series/3201