Options
Rafter, Rachael
Preferred name
Rafter, Rachael
Official Name
Rafter, Rachael
Research Output
Now showing 1 - 10 of 20
- PublicationTowards Conversational Collaborative Recommendation(2004-09)
; Traditionally, collaborative recommender systems have been based on a single-shot model of recommendation where a single set of recommendations are generated based on a user's (past) stored preferences. However, content-based recommender system research has begun to look towards more conversational models of recommendation, where the user is actively engaged in directing search at recommendation time. Such interactions can range from deep dialogues with the user that may involve natural language dialogues, to more simple interactions where the user is, for example, asked to indicate a preference for one of k suggested items. Importantly, the feedback attained from these interactions can help to di erentiate between the user's long-term stored preferences, and her current (short-term) requirements, which may be quite di erent. We argue that such interactions can also be bene cial to collaborative recommendation and provide preliminary experimental evidence in support of this.209 - PublicationA Game with a Purpose for Recommender SystemsRecommender systems learn about our preferences to make targeted suggestions. In this paper we outline a novel game with-a-purpose designed to infer preferences at scale as aside-effect of gameplay. We evaluate the utility of this data in a recommendation context as part of a small live-user trial.
150 - PublicationPassive Profiling from Server Logs in an Online Recruitment Environment(2001-08)
; The success of recommender systems ultimately depends on the availability of comprehensive user profiles that accurately capture the interests of endusers. However, the automatic compilation of such profiles represents a complex learning task. In this paper, we focus on how accurate user profiles can be generated directly from analysing the behaviours of Web users in the CASPER project. In CASPER user profiles are constructed by passively monitoring the click-stream and read-time behaviour of users. We will argue that building accurate profiles from such data is far from straightforward. In particular, we will describe the techniques that are used in CASPER to generate highquality graded user profiles from server logs. Finally, we will describe a comparative evaluation of these different techniques on real user data.764 - PublicationPersonalised Retrieval for Online Recruitment Services(2000-04-05)
; ; Internet search technology is largely based on exact-match retrieval. Such systems rely on the user to provide an adequate description of their requirements. However, most users submit poorly specified queries, leading to imprecise search results. Furthermore, there is no facility for personalising searches to reflect implicit likes and dislikes of users. We describe two complementary solutions to these problems as implemented in CASPER, an intelligent online recruitment service. We describe an approach to personalised similarity-based retrieval, and a queryless collaborative filtering recommendation technique.1365 - PublicationGreat Explanations: Opinionated Explanations for RecommendationExplaining recommendations helps users to make better, more satisfying decisions. We describe a novel approach to explanation for recommender systems, one that drives the recommendation process, while at the same time providing the user with useful insights into the reason why items have been chosen and the trade-os they may need to consider when making their choice. We describe this approach in the context ofa case-based recommender system that harnesses opinions mined from user-generated reviews, and evaluate it on TripAdvisor Hotel data.
597Scopus© Citations 36 - PublicationConversational collaborative recommendation - an experimental analysisTraditionally, collaborative recommender systems have been based on a single-shot model of recommendation where a single set of recommendations is generated based on a user's (past) stored preferences. However, content-based recommender system research has begun to look towards more conversational models of recommendation, where the user is actively engaged in directing search at recommendation time. Such interactions can range from high-level dialogues with the user, possibly in natural language, to more simple interactions where the user is, for example, asked to indicate a preference for one of k suggested items. Importantly, the feedback attained from these interactions can help to differentiate between the user's long-term stored preferences, and her current (short-term) requirements, which may be quite different. We argue that such interactions can also be beneficial to collaborative recommendation and provide experimental evidence to support this claim.
681Scopus© Citations 22 - PublicationThe Recommendation GameThis paper describes a casual Facebook game to capture recommendation data as a side-effect of gameplay. We showhow this data can be used to make successful recommendations as part of a live-user trial.
191Scopus© Citations 8 - PublicationSticking with a Winning Team: Better Neighbour Selection for Conversational Collaborative Recommendation(2007-08)
; ; ; Conversational recommender systems have recently emerged as useful alternative strategies to their single-shot counterpart, especially given their ability to expose a user’s current preferences. These systems use conversational feedback to hone in on the most suitable item for recommendation by improving the mechanism that finds useful collaborators. We propose a novel architecture for performing recommendation that incorporates information about the individual performance of neighbours during a recommendation session, into the neighbour retrieval mechanism. We present our architecture and a set of preliminary evaluation results that suggest there is some merit to our approach.We examine these results and discuss what they mean for future research.142 - PublicationWhat have the neighbours ever done for us? A collaborative filtering perspectiveCollaborative filtering (CF) techniques have proved to be a powerful and popular component of modern recommender systems. Common approaches such as user-based and item-based methods generate predictions from the past ratings of users by combining two separate ratings components: a base estimate, generally based on the average rating of the target user or item, and a neighbourhood estimate, generally based on the ratings of similar users or items. The common assumption is that the neighbourhood estimate gives CF techniques a considerable edge over simpler average-rating techniques. In this paper we examine this assumption more carefully and demonstrate that the influence of neighbours can be surprisingly minor in CF algorithms, and we show how this has been disguised by traditional approaches to evaluation, which, we argue, have limited progress in the field.
1320Scopus© Citations 10