Optimising Attribute Selection in Conversational Search
نویسندگان
چکیده
It has been shown that user modelling has the potential to improve the performance of conversational search systems, particularly in what concerns the problem of attribute selection, i.e., determining which attribute to ask the user at each step of the dialogue. In this paper we present a novel framework for attribute selection which allows the fine-tuning of the relative importance of profile-based and entropy-based heuristics. Based on this framework, we describe a number of experiments which allow us to quantify the bounds to such improvements.
منابع مشابه
Reduction of Combinatorial Space of Adjustable Kinetic Parameters of Biochemical Network Models in Optimisation Task
The search for minimal set of adjustable parameters through optimising a kinetic model of biochemical networks is needed in industrial biotechnology to increase the productivity of industrial organism strains while keeping low the chance of causing unwanted side effects of implemented changes. As the search for minimal set of adjustable parameters is of combinatorial nature, the search space be...
متن کاملThe Adaptive Place Advisor: A Conversational Recommendation System
In this paper, we describe the Adaptive Place Advisor, a conversational recommendation system designed to help users decide on a destination. We view the selection of destinations as an interactive, conversational process, with the advisory system inquiring about desired item characteristics and the human responding. The user model, which contains preferences regarding items, attributes, values...
متن کاملDesigning a model of intuitionistic fuzzy VIKOR in multi-attribute group decision-making problems
Multiple attributes group decision making (MAGDM) is regarded as the process of determining the best feasible solution by a group of experts or decision makers according to the attributes that represent different effects. In assessing the performance of each alternative with respect to each attribute and the relative importance of the selected attributes, quantitative/qualitative evaluations ar...
متن کاملEfficient Selection of Design Parameters in Multi-Objective Economic-Statistical Model of Attribute C Control Chart
Control chart is the most well-known chart to monitor the number of nonconformities per inspection unit where each sample consists of constant size. Generally, the design of a control chart requires determination of sample size, sampling interval, and control limits width. Optimally selecting these parameters depends on several process parameters, which have been considered from statistical and...
متن کاملA Personalized System for Conversational Recommendations
Increased computing power and the Web have made information widely accessible. In turn, this has encouraged the development of recommendation systems that help users find items of interest, such as books or restaurants. Such systems are more useful when they personalize themselves to each user’s preferences, thus making the recommendation process more efficient and effective. In this paper, we ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2003