Whiting, Stewart William (2015) Temporal dynamics in information retrieval. PhD thesis. http://theses.gla.ac.uk/6850/ Copyright and moral rights for this thesis are retained by the author A copy can be downloaded for personal non-commercial research or study This thesis cannot be reproduced or quoted extensively from without first obtaining permission in writing from the Author The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the Author When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given Glasgow Theses Service http://theses.gla.ac.uk/ [email protected] Temporal Dynamics in Information Retrieval Stewart William Whiting School of Computing Science College of Science and Engineering University of Glasgow, Scotland, UK. A thesis submitted for the degree of Doctor of Philosophy (Ph.D) October, 2015 I hereby declare that except where specific reference is made to the work of others, the con- tents of this dissertation are original and have not been submitted in whole or in part for consideration for any other degree or qualification in this, or any other University. This dissertation is the result of my own work, under the supervision of Professor Joemon M. Jose and Dr Gethin Norman, and includes nothing which is the outcome of work done in collaboration, except where specifically indicated in the text. Permission to copy without fee all or part of this thesis is granted provided that the copies are not made or distributed for commercial purposes, and that the name of the author, the title of the thesis and date of submission are clearly visible on the copy. Stewart William Whiting October, 2015 Abstract The passage of time is unrelenting. Time is an omnipresent feature of our existence, serving as a context to frame change driven by events and phenomena in our personal lives and social constructs. Accordingly, various elements of time are woven throughout information itself, and information behaviours such as creation, seeking and utilisation. Time plays a central role in many aspects of information retrieval (IR). It can not only dis- tinguish the interpretation of information, but also profoundly influence the intentions and expectations of users’ information seeking activity. Many time-based patterns and trends – namely temporal dynamics – are evident in streams of information behaviour by individuals and crowds. A temporal dynamic refers to a periodic regularity, or, a one-off or irregular past, present or future of a particular element (e.g., word, topic or query popularity) – driven by predictable and unpredictable time-based events and phenomena. Several challenges and opportunities related to temporal dynamics are apparent throughout IR. This thesis explores temporal dynamics from the perspective of query popularity and meaning, and word use and relationships over time. More specifically, the thesis posits that temporal dynamics provide tacit meaning and structure of information and information seek- ing. As such, temporal dynamics are a ‘two-way street’ since they must be supported, but also conversely, can be exploited to improve time-aware IR effectiveness. Real-time temporal dynamics in information seeking must be supported for consistent user satisfaction over time. Uncertainty about what the user expects is a perennial problem for IR systems, further confounded by changes over time. To alleviate this issue, IR systems can: (i) assist the user to submit an effective query (e.g., error-free and descriptive), and (ii) better anticipate what the user is most likely to want in relevance ranking. I first explore methods to help users formulate queries through time-aware query auto-completion, which can suggest both recent and always popular queries. I propose and evaluate novel approaches for time-sensitive query auto-completion, and demonstrate state-of-the-art performance of up to 9.2% improvement above the hard baseline. Notably, I find results are reflected across di- verse search scenarios in different languages, confirming the pervasive and language agnostic nature of temporal dynamics. Furthermore, I explore the impact of temporal dynamics on the motives behind users’ information seeking, and thus how relevance itself is subject to tempo- ral dynamics. I find that temporal dynamics have a dramatic impact on what users expect over time for a considerable proportion of queries. In particular, I find the most likely meaning of ambiguous queries is affected over short and long-term periods (e.g., hours to months) by sev- eral periodic and one-off event temporal dynamics. Additionally, I find that for event-driven multi-faceted queries, relevance can often be inferred by modelling the temporal dynamics of changes in related information. In addition to real-time temporal dynamics, previously observed temporal dynamics offer a complementary opportunity as a tacit dimension which can be exploited to inform more effec- tive IR systems. IR approaches are typically based on methods which characterise the nature of information through the statistical distributions of words and phrases. In this thesis I look to model and exploit the temporal dimension of the collection, characterised by temporal dy- namics, in these established IR approaches. I explore how the temporal dynamic similarity of word and phrase use in a collection can be exploited to infer temporal semantic relationships between the terms. I propose an approach to uncover a query topic’s “chronotype” terms – that is, its most distinctive and temporally interdependent terms, based on a mix of temporal and non-temporal evidence. I find exploiting chronotype terms in temporal query expansion leads to significantly improved retrieval performance in several time-based collections. Temporal dynamics provide both a challenge and an opportunity for IR systems. Overall, the findings presented in this thesis demonstrate that temporal dynamics can be used to derive tacit structure and meaning of information and information behaviour, which is then valu- able for improving IR. Hence, time-aware IR systems which take temporal dynamics into account can better satisfy users consistently by anticipating changing user expectations, and maximising retrieval effectiveness over time. Acknowledgements I dedicate this thesis to my parents, Richard and Pamela Whiting. Their love, support and encouragement has been unwavering. Without them I would never have been able to follow this opportunity in my life. I can never express quite how truly appreciative I am. Writing this thesis has been a long journey. I am deeply thankful to all my friends, family and colleagues who have aided me throughout. I would like to thank my siblings – Marc, Rebecca and Emma Whiting – for being there for me, especially during the difficult past two years. My aunty and uncle, Sandy and Brian Talbot, opened the doors that led me down this path of discovery. Their help, guidance and inspiration changed the way I look at the world, and for that I am ever grateful. I will be forever indebted to my first mentor, Wayne Kerridge, who provided the early inspiration and support that allowed me to develop a career in technology. My colleagues have been a source of much inspiration and comedy over my years as a PhD student. Guido Zuccon and Teerapong Leelanupab helped me to become established when I first started. Ke Zhou, Jesus Rodriguez Perez, Philip McParlane, James McMinn, Horatiu Bota, Rami Alkhawaldeh, Fajie Yuan and Stefan Raue – their company and collaboration has been a pleasure. I am extremely appreciative of my supervisor, Joemon Jose, for handing me the opportunity to do a PhD funded by the EPSRC DTA scheme. He granted me the freedom to follow many new research ideas, yet provided the counsel when needed. I would also like to take this opportunity to thank my viva examiners, Arjen de Vries and Milad Shokouhi, for their hard work in providing excellent comments and suggestions to improve the thesis. From the very start of my time as a PhD student, it has been an absolute privilege to have Yashar Moshfeghi as my mentor. He voluntarily took a central role in helping me shape my research ideas and this thesis, and for that I will be forever appreciative. Special thanks must go to Omar Alonso, who gave me the incredible opportunity to join Mi- crosoft Research in Silicon Valley as an intern in 2012, and again in 2014. Omar and my other supervisors at Microsoft – Aditi (Shubha) Nabar and Alex Dow – mentored me to develop a research and development approach that has laid the foundations of my career. Finally, I need to thank my partner, Jodie Clarke, for supporting me during this journey. We have gone through this together, and for that I can never quite thank her enough. The only constant is change: “All is flux, nothing is stationary; no man ever steps in the same river twice, for it is not the same river and he is not the same man.” – Plato, around 369 BC. Contents Glossary xvi I Introduction and Background2 1 Introduction3 1.1 Preface . .3 1.2 Thesis Statement . .5 1.3 Motivation . .6 1.4 Research Questions . .9 1.5 Outline . .9 1.6 Publications . 11 2 General IR Background 13 2.1 Introduction . 13 2.1.1 Chapter Outline . 13 2.2 Information Retrieval Tasks . 13 2.3 Retrieval Models . 14 2.3.1 Boolean Model . 15 2.3.2 Vector Space Model . 15 2.3.3 Probabilistic Models . 16 2.3.4 Language Models . 17 2.4 Experimental Methodologies . 18 2.4.1 System-oriented Evaluation . 19 2.4.2 User-oriented Evaluation . 20 2.5 Evaluating Retrieval Effectiveness . 20 2.5.1 Recall . 21 2.5.2 Precision . 21 2.5.3 Mean Average Precision (MAP) . 21 2.5.4 Chapter Summary . 22 vi CONTENTS 3 Time-aware IR Background and Motivation 23 3.1 Introduction .