Past Meetup

User Satisfaction and String Matching

This Meetup is past

67 people went

Details

This Friday, we'll have two talks followed by drinks. Julia Kiseleva (http://juliakiseleva.com/) from ILPS-UVA (http://ilps.science.uva.nl) will give the academic talk. The industry talk will be given by Petro Protsyk from ZyLAB (https://www.zylab.com).

This edition of SEA will be held in SPUI25.

Program:

16:00 - 16:30 Julia Kiseleva

16:30 - 17:00 Petro Protsyk

17:00 - 18:00 Drinks & Snacks

Details of the talks:

----------------------------------------------------------------------------------------

Julia Kiseleva -- Predicting User Satisfaction with Intelligent Assistants

There is a rapid growth in the use of voice-controlled intelligent personal assistants on mobile devices, such as Microsoft’s Cortana, Google Now, and Apple’s Siri. They significantly change the way users interact with search systems, not only because of the voice control use and touch gestures, but also due to the dialogue-style nature of the interactions and their ability to preserve context across different queries. Predicting success and failure of such search dialogues is a new problem, and an important one for evaluating and further improving intelligent assistants. While clicks in web search have been extensively used to infer user satisfaction, their significance in search dialogues is lower due to the partial replacement of clicks with voice control, direct and voice answers, and touch gestures.
In this research, we propose an automatic method to predict user satisfaction with intelligent assistants that exploits all the interaction signals, including voice commands and physical touch gestures on the device. First, we conduct an extensive user study to measure user satisfaction with intelligent assistants, and simultaneously record all user interactions. Second, we show that the dialogue style of interaction makes it necessary to evaluate the user experience at the overall task level as opposed to the query level. Third, we train a model to predict user satisfaction, and find that interaction signals that capture the user reading patterns have a high impact: when including all available interaction signals, we are able to improve the prediction accuracy of user satisfaction from 71% to 81% over a baseline that utilizes only click and query features.

Julia Kiseleva is co-founder of usersat.com (http://usersat.com/), a spin-off company offering search and recommender systems optimization specifically for mobile devices, and researcher at University of Amsterdam. She has extensive industrial experience at leading IT companies including Microsoft Research, Microsoft Bing, Yandex.ru, Hewlett Packard Research, E-bay, and Booking.com. In 2016, Julia obtained a PhD from Eindhoven University of Technology on her research to improve the user’s search and browse experience. See http://juliakiseleva.com/ and http://usersat.com/ .

----------------------------------------------------------------------------------------

Petro Protsyk --Algorithms for approximate string matching

Approximate string matching or fuzzy search is the technique of finding strings in text or dictionary that match pattern approximately. The most common applications of fuzzy search are spell checking, syntax error correction, spam filtering and correction of OCR errors.

At ZyLAB we are developing an enterprise level full-text search engine that supports Boolean and proximity operators, fast wildcard, regular expression and fuzzy search. Over the years we have gained a lot of experience in this area and implemented various algorithms and data structures in ZyLAB Search Engine.In this talk we are going to present an overview of approximate string matching algorithms that use Levenshtein metric and their practical application in our software.

Petro Protsyk has a MSc degree in Computer Science at Taras Shevchenko National University of Kyiv, Cybernetics department in 2006. Since than Petro works as a professional software developer. He joined ZyLAB in 2010 to work on ZyLAB’s proprietary Search Engine and eDiscovery solution. During the last 6 years Petro focused on building distributed solutions for processing large volumes of data and implementing a new version of the full-text search engine for ZyLAB. For more information, see https://www.zylab.com