HCI Bibliography Home | HCI Conferences | IIiX Archive | Detailed Records | RefWorks | EndNote | Hide Abstracts
IIiX Tables of Contents: 0608101214

Proceedings of the 2014 Symposium on Information Interaction in Context

Fullname:Proceedings of the 5th Symposium on Information Interaction in Context
Editors:David Elsweiler; Bernd Ludwig; Leif Azzopardi; Max L. Wilson
Location:Regensburg, Germany
Dates:2014-Aug-26 to 2014-Aug-29
Publisher:ACM
Standard No:ISBN: 978-1-4503-2976-7; ACM DL: Table of Contents; hcibib: IIiX14
Papers:64
Pages:356
Links:Conference Website
  1. Keynote presentations
  2. Workshops
  3. Full papers
  4. Short papers
  5. Demos
  6. Doctoral consortium

Keynote presentations

Information retrieval evaluation with humans in the loop BIBAFull-Text 1
  Gabriella Kazai
The evaluation and tuning of information retrieval (IR) systems based on the Cranfield paradigm requires purpose built test collections, which include sets of human contributed relevance labels, indicating the relevance of search results to a set of user queries. Traditional methods of collecting relevance labels rely on a fixed group of hired expert judges, who are trained to interpret user queries as accurately as possible and label documents accordingly. Human judges and the obtained relevance labels thus provide a critical link within the Cranfield style IR evaluation framework, where disagreement among judges and the impact of variable judgment sets on the final outcome of an evaluation is a well studied issue. There is also reported evidence that experiment outcomes can be affected by changes to the judging guidelines or changes in the judge population.
   Recently, the growing volume and diversity of the topics and documents to be judged is driving the increased adoption of crowdsourcing methods in IR evaluation, offering a viable alternative that scales with modest costs. In this model, relevance judgments are distributed online over a large population of humans, a crowd, facilitated, for example, by a crowdsourcing platform, such as Amazon's Mechanical Turk or Clickworker. Such platforms allow millions of anonymous crowd workers to be hired temporarily for micro-payments to complete so-called human intelligence tasks (HITs), such as labeling images or documents. Studies have shown that workers come from diverse backgrounds, work in a variety of different environments, and have different motivations. For example, users may turn to crowdsourcing as a way to make a living, to serve an altruistic or social purpose or simply to fill their time. They may become loyal crowd workers on one or more platforms, or they may leave after their first couple of encounters. Clearly, such a model is in stark contrast to the highly controlled methods that characterize the work of trained judges. For example, in a micro-task based crowdsourcing setup, worker training is usually minimal or non-existent. Furthermore, it is widely reported that labels provided by crowd workers can vary in quality, leading to noisy labels. Crowdsourcing can also suffer from undesirable worker behaviour and practices, e.g., dishonest behaviour or lack of expertise, that result in low quality contributions. While a range of quality assurance and control techniques have now been developed to reduce noise during or after task completion, little is known about the workers themselves and possible relationships between workers' characteristics, behaviour and the quality of their work.
   In this talk, I will review the findings of recent research that examines and compares trained judges and crowd workers hired to complete relevance assessment tasks of varying difficulty. The investigations include a range of aspects from how HIT design, judging instructions, worker demographics and characteristics may impact work quality. The main focus of the talk will be on experiments aimed to uncover characteristics of the crowd by monitoring their behaviour during different relevance assessment tasks, and compare them to professional judges' behaviour on the same tasks. Throughout the talk I will highlight challenges of quality assurance and control in crowdsourcing and propose a possible direction for solving the issue without relying on gold standard data sets, which are expensive to create and have limited application.
Seeking answers, making sense, changing lifestyles: cognitive models of human-information interaction BIBAFull-Text 2-3
  Peter Pirolli
This presentation will discuss complex human-information interaction problems involving information foraging, sensemaking, and lifestyle change (behavior change), predictive models of human cognition in these contexts, as well as novel interaction techniques inspired by these models.
Engaging users with situational recommendations: challenges and results BIBAFull-Text 4-5
  Francesco Ricci
Recommender Systems are popular tools that automatically compute personalised suggestions for items that are predicted to be interesting and useful to a user [24, 17]. For instance, in the music domain recommender systems support information search and discovery tasks by helping the user to find music tracks or artists that the user may not even know, but he will like [7, 15, 14].
   Recommender systems accomplish their functionality by explicitly requesting users to enter their preferences and by tracking users' actions and behaviours, which implicitly signal users' preferences. Then, they aggregate these observation data and build predictive models of the users' future interests. Several techniques have been proposed to model user preferences and generate recommendations for them. But, ultimately, most of the implemented systems use content-, collaborative- or social-based approaches, or even more often, hybrid combinations of these three basic approaches [6].
   In addition to long-term interests, which are normally acquired and modelled in RSs, other session specific factors do influence the user's response to the suggested items and therefore should be taken into consideration. These factors include: the ephemeral needs of the users [21, 19], their decision biases [8, 25], the context of the search [10, 18] and the context of items' usage [1]. However, appropriately modeling the user's preferences and behaviour in the possible various and diverse situational contexts and reasoning upon them in order to identify useful, convincing, diverse and relevant recommendations is still challenging. Major technical and practical difficulties must yet to be solved.
   First of all, one should parsimoniously narrow down the various types and the number of contextual dimensions that the system should model to those that actually influence the user decision making processes [2, 23]. Then, it is important to understand the dynamics of the impact of such contextual dimensions on the user preferences and the decision-making process [8]. This impact is strongly coupled with the full interaction design of the system [5, 16]. Moreover, it is important to implement technical solutions that enable the system to continuously acquire context-dependent user evaluations (e.g., ratings) for the suggested items, during the full life cycle of the system [20, 11, 12, 22]. Finally, one must embed the contextual dimensions and leverage the acquired data in a recommendation computational model [3, 9], while dealing with the typically very limited knowledge of the system for the users, the items and the contextual situations [4, 13].
   These topics will be illustrated in the talk, making examples taken from the recommender systems that we have developed in the tourism and music domains.

Workshops

Workshop on searching for fun 2014 BIBAFull-Text 6
  Morgan Harvey; Max Wilson; Karen Church
The Searching for Fun workshop brings together researchers who are interested in furthering our knowledge about searching and browsing in a casual leisure scenario. These are scenarios where the searcher does not have a specific information need to fulfil, but rather simply wishes to be entertained with no specific goal in mind. This can include: online window shopping with nothing to buy, reading online (including the news), watching funny videos, finding funny pictures, browsing Wikipedia or social networking sites. Following on from the successful workshop at ECIR in 2012 and later discussions at Dagstuhl meetings, this event involves researchers from several IR sub-communities (e.g. recommender systems, result diversity, multimedia retrieval) and related disciplines, discussing new and early research and creating a vision for future work in this area.
Searching as learning (SAL) workshop 2014 BIBAFull-Text 7
  Luanne Freund; Jiyin He; Jacek Gwizdka; Noriko Kando; Preben Hansen; Soo Young Rieh
In this paper we describe the Searching as Learning Workshop (SAL 2014) taking place at IIiX 2014 in Regensburg, Germany.

Full papers

Finding information about mental health in microblogging platforms: a case study of depression BIBAFull-Text 8-17
  Max L. Wilson; Susan Ali; Michel F. Valstar
Searching for online health information has been well studied in web search, but social media, such as public microblogging services, are well known for different types of tacit information: personal experience and shared information. Finding useful information in public microblogging platforms is an on-going hard problem and so to begin to develop a better model of what health information can be found, Twitter posts using the word "depression" were examined as a case study of a search for a prevalent mental health issue. 13,279 public tweets were analysed using a mixed methods approach and compared to a general sample of tweets. First, a linguistic analysis suggested that tweets mentioning depression were typically anxious but not angry, and were less likely to be in the first person, indicating that most were not from individuals discussing their own depression. Second, to understand what types of tweets can be found, an inductive thematic analysis revealed three major themes: 1) disseminating information or link of information, 2) self-disclosing, and 3) the sharing of overall opinion; each had significantly different linguistic patterns. We conclude with a discussion of how different types of posts about mental health may be retrieved from public social media like Twitter.
Usability and perception of young users and adults on targeted web search engines BIBAFull-Text 18-27
  Tatiana Gossen; Juliane Höbel; Andreas Nürnberger
The usability of web search engines is an important factor that influences user experience and correlates with users' success in finding the relevant information. Currently, there are different search engines online available whose main target audience are children. In this paper, we investigate the differences between children and adults in terms of usability and perception of targeted search engines, i.e. search engines designed specifically for that audience. To this end, an eye-tracking study was conducted to compare children's and adults' search behavior and perception of search interface elements on search engine results pages (SERPs) during an informational and a navigational search with a standard search engine and a search engine for children. We identified differences in the information-seeking behavior and perception of search engines SERPs between children and adults. Based on these findings we propose criteria on how to design search user interfaces that are more appropriate for children.
Books' interest grading and dwell time in metadata in selecting fiction BIBAFull-Text 28-37
  Pertti Vakkari; Arto Luoma; Janna Pöntinen
It is studied by eye-tracking how searchers explore metadata in book pages when selecting novels of varying interest levels. 30 participants searched interesting novels for four search tasks in two public library catalogs. The results showed that the associations of dwell time in book pages and in many metadata types, and novels' interest grading were non-linear. Most time was used for assessing a somewhat interesting novel compared to a non-interesting or very interesting one. Therefore, the binary classification of interest grading hides and over-emphasizes the contribution of "somewhat interesting" category in modeling interest by dwell time in book pages or metadata. Non-linear regression models showed that the explanatory power was greater in a three-level classification of interest grading compared to a binary classification.
Relating user interaction to experience during festivals BIBAFull-Text 38-47
  Richard Schaller; Morgan Harvey; David Elsweiler
Festivals held in a city (or number of cities) contain many geographically distributed events often occurring contemporaneously. Visitors to the festival need to make numerous cognitively-challenging decisions about which events to see, and in which order. Consequently, the visitors' information interactions with online and mobile guides are likely to influence their experience of the festival. In this paper we investigate how such interactions with a mobile app, designed to provide visitors with information about the festival and to help them plan their itinerary, relate to their experience and how they participated in the festival. The app was deployed in a large-scale naturalistic study (n=1159). Our analysis reveals that different information interaction styles corresponded to itineraries with different properties. The results of a follow-up survey (n=59), completed by a sub-sample of these users, suggests that this is no coincidence. Analysing what people reported in terms of their desires for their evening reveals trends indicating that user groups who made use of the same interface features (i.e. search, browse or recommendation) had similar priorities when planning their evening and ended up visiting events that reflect those priorities. These findings suggest that users are able to adapt their interaction style to use the features most appropriate to their needs. We conclude by discussing what our findings mean in terms of the information behaviour literature and evaluating interactive information retrieval systems embedded in a real context.
A user defined taxonomy of factors that divide online information retrieval sessions BIBAFull-Text 48-57
  Chaoyu Ye; Max L. Wilson
Although research is increasingly interested in session-based retrieval, comparably little work has focused on how best to divide web histories into sessions. Most automated attempts to divide web histories into sessions have focused on dividing web logs using simplistic rules, including user identifiers and specific time gaps. This research, however, is focused on understanding the full range of factors that affect the division of sessions, so that we can begin to go beyond current naive techniques like fixed time periods of inactivity. To investigate these factors, 10,000 log items were manually analysed by their owners into 847 naturally occurring web sessions. During interviews, participants reviewed their own web histories to identify these sessions, and described the causes of divisions between sessions. This paper contributes a taxonomy of six factors that can be used to better model the divisions between sessions, along with initial insights into how the divided sessions manifested in web logs. The factors in our taxonomy provide focus for future work, including our own, for finding practical ways to more intelligently divide and identify sessions for improved session-based retrieval.
Characterizing relevance with eye-tracking measures BIBAFull-Text 58-67
  Jacek Gwizdka
Relevance, a fundamental concept in information search and retrieval, is 80-years old [4]. The recent decades have been ripe with work that brought a much better understanding of this rich concept. Yet, we still don't know which cognitive and affective processes are involved in relevance judgments. Empirical work that tackles these questions is scarce. This paper aims to contribute toward better understanding of cognitive processing of text documents at different degrees of relevance. Our approach takes advantage of a direct relationship between eye movement patterns, pupil size and cognitive processes, such as mental effort and attention. We examine gaze-based metrics in relation to individual word processing and reading text documents in the context of a constricted information search tasks. The findings indicate that text document processing depends on document relevance and on the user-perceived relevance. Statistical analyses show that relevant documents tended to be continuously read, while irrelevant documents tended to be scanned. Most eye-tracking-based measures indicate cognitive effort to be highest for partially relevant documents and lowest for irrelevant documents. However, pupil dilation indicates cognitive effort to be higher for relevant than partially relevant documents. Classification of selected eye-tracking measures show that an accuracy of 70-75% can be achieved for predicting binary relevance. These results show a promise for implicit relevance feedback.
How concept maps change if a user does search or not? BIBAFull-Text 68-75
  Yuka Egusa; Masao Takaku; Hitomi Saito
Previous studies have shown that a concept map can capture changes in the user knowledge structure during a search. However, these studies could not exclude the possibility of the influence of instructions or time-dependent changes. In this study, we have compared differences between concept maps created before and after a search condition and a non-search condition to reveal whether these changes are due to searching.
   In the experiment, participants were required to gather information on the Web in preparation for a group discussion. The participants were divided into two groups representing two tasks, convergent and divergent tasks. The convergent task required gathering web pages for a specific and detailed discussion, and the divergent task required gathering web pages for a wide-ranging discussion. Participants performed each task under search and filler conditions. In the search condition, they searched the Web. In the filler condition, they played a typing game on a PC.
   We compared pre- and post-task concept maps. Analysis of the number of nodes in the concept maps indicated that changes in the search condition are significant, whereas changes in the filler condition are insignificant. The analysis of the number of nodes at each distance from the center nodes in the concept maps showed that tasks had a greater effect in the search condition than in the filler condition. Finally, we consider whether the experimental results support our hypotheses.
Are episodic context features helpful for refinding tasks?: lessons learnt from a case study with lifelogs BIBAFull-Text 76-85
  Yi Chen; Gareth J. F. Jones
Both psychological theories and findings in information science suggest that people may remember the episodic context of previously encountered information. This implies that a user's episodic memory might be utilized to improve the efficiency or effectiveness of refinding tasks. In this paper, we report a case study which aims to explore the feasibility of integrating episodic context into the design of information refinding systems. The subjects in this study collected 20 months of rich contextual data along including the full text of all documents, emails, web pages and so on, which they accessed during the collection period. We developed a "memory-friendly" system based on psychological theories to test the hypothesis through user studies requiring the subjects to find their personal data using this system. From examination of the user activity log and a post-task questionnaire, we found that although our designed features, which support or utilize episodic context or autobiographical memory, were not used as frequently as we expected, they did improve the effectiveness of the refinding tasks.
E-mail categorization using partially related training examples BIBAFull-Text 86-95
  Maya Sappelli; Suzan Verberne; Wessel Kraaij
Automatic e-mail categorization with traditional classification methods requires labelling of training data. In a real-life setting, this labelling disturbs the working flow of the user. We argue that it might be helpful to use documents, which are generally well-structured in directories on the file system, as training data for supervised e-mail categorization and thereby reducing the labelling effort required from users. Previous work demonstrated that the characteristics of documents and e-mail messages are too different to use organized documents as training examples for e-mail categorization using traditional supervised classification methods.
   In this paper we present a novel network-based algorithm that is capable of taking into account these differences between documents and e-mails. With the network algorithm, it is possible to use documents as training material for e-mail categorization without user intervention. This way, the effort for the users for labeling training examples is reduced, while the organization of their information flow is still improved.
   The accuracy of the algorithm on categorizing e-mail messages was evaluated using a set of e-mail correspondence related to the documents. The proposed network method was significantly better than traditional text classification algorithm in this setting.
Investigating the potential impact of non-personalized recommendations in the OPAC: Amazon vs. WorldCat.org BIBAFull-Text 96-105
  Simon Wakeling; Paul Clough; Barbara Sen
Recent research into the functionality of Online Public Access Catalogues (OPACs) has led to a call for such systems to incorporate functionality to facilitate resource discovery, and replicate the information search experience users encounter elsewhere on the Web. Recommendations represent one such feature. Developments so far in this area indicate that non-personalized or item-level recommendations are most suited to the OPAC environment. Whilst a number of such systems have been developed and implemented, research has yet to investigate fully the impact of such recommendations on user performance, search behavior, and system perceptions. This paper presents the results of an exploratory laboratory-based study comparing user behavior in Amazon, which offers non-personalized recommendations, and WorldCat.org, which does not. An analysis of task performance and participant interactions with the systems reveals that the presence of non-personalized recommendations improves resource discovery, search efficiency, and perceived usability.
Social priors to estimate relevance of a resource BIBAFull-Text 106-114
  Ismail Badache; Mohand Boughanem
In this paper we propose an approach that exploits social data associated with a Web resource to measure its a priori relevance. We show how these interaction traces left by the users on the resources, which are in the form of social signals as the number of like and share, can be exploited to quantify social properties such as popularity and reputation. We propose to model these properties as a priori probability that we integrate into language model. We evaluated the effectiveness of our approach on IMDb dataset containing 167438 resources and their social signals collected from several social networks. Our experimental results are statistically significant and show the interest of integrating social properties in a search model to enhance the information retrieval.
On choosing an effective automatic evaluation metric for microblog summarisation BIBAFull-Text 115-124
  Stuart Mackie; Richard McCreadie; Craig Macdonald; Iadh Ounis
Popular microblogging services, such as Twitter, are engaging millions of users who constantly post and share information about news and current events each day, resulting in millions of messages discussing what is happening in the world. To help users obtain an overview of microblog content relating to topics and events that they are interested in, classical summarisation techniques from the newswire domain have been successfully applied and extended for use on microblogs. However, much of the current literature on microblog summarisation assumes that the summarisation evaluation measures that have been shown to be effective on newswire, are still appropriate for evaluating microblog summarisation. Hence, in this paper, we aim to determine whether the traditional automatic newswire summarisation evaluation metrics generalise to the task of microblog summarisation. In particular, using three microblog summarisation datasets, we determine a ranking of summarisation systems under three automatic summarisation evaluation metrics from the literature. We then compare and contrast this ranking of systems produced under each metric to system rankings produced through a qualitative user evaluation, with the aim of determining which metric best simulates human summarisation preferences. Our results indicate that, for the automatic evaluation metrics we investigate, they do not always concur with each other. Further, we find that Fraction of Topic Words better agrees with what users tell us about the quality and effectiveness of microblog summaries than the ROUGE-1 measure that is most commonly reported in the literature.
Diversifying contextual suggestions from location-based social networks BIBAFull-Text 125-134
  M-Dyaa Albakour; Romain Deveaud; Craig Macdonald; Iadh Ounis
In this paper, we study the emerging Information Retrieval (IR) task of contextual suggestion in location-based social networks. The aim of this task is to make personalised recommendations of venues for entertainments or activities whilst visiting a city, by appropriately representing the context of the user, such as their location and personal interests. Instead of only representing the specific low-level interests of a user, our approach is driven by estimates of the high-level categories of venues that the user may be interested in. Moreover, we argue that an effective model for contextual suggestion should not only promote the categories that the user is interested in, but it should also be capable of eliminating redundancy by diversifying the recommended venues in the sense that they should cover various categories of interest to the given user. Therefore, we adapt web search result diversification approaches to the task of contextual suggestion. For categorising the venues, we use the category classifications employed by location-based social networks such as FourSquare, urban guides such as Yelp, and a large collection of web pages, the ClueWeb12 corpus, to build a textual classifier that is capable of predicting the category distribution for a certain venue given its web page. We thoroughly evaluate our approach using the TREC 2013 Contextual Suggestion track. We conduct a number of experiments where we consider venues from the closed environments of both FourSquare and Yelp, and the general web using the ClueWeb12 corpus. Our empirical results suggest that category diversification consistently improves the effectiveness of the recommendation model over a reasonable baseline that only considers the similarity between the user's profile and venue. The results also give insights on the effectiveness of our approach with different types of users.
Exploring knowledge graphs for exploratory search BIBAFull-Text 135-144
  Bahareh Sarrafzadeh; Olga Vechtomova; Vlado Jokic
In order to provide the user with more support in performing exploratory activities, recent research has been focused on identifying the types of tasks users perform, and understanding the nature of these tasks. However, most of the proposed models focus on either traditional document retrieval or the use of linked data for finding relevant information. We believe neither of these two types of information resources can offer sufficient support for complex search tasks on their own. We propose that a hybrid approach that combines the coherent content of text with the organized structure of graphs should be taken to better support information finding and sense making.
   Currently, there is limited insight into the types of information seeking activities performed when a knowledge graph is combined with document retrieval to support exploratory search. This paper describes a general framework that provides the first step towards examining users' exploratory search behaviour when interacting with knowledge graphs and their corresponding documents. We conducted a user study that suggests searchers perform different information seeking activities for a complex search task compared with a simple search task. These findings provide insights that can be used to inform the design of a new search framework, which enables more effective information finding and analysis.
From multistage information-seeking models to multistage search systems BIBAFull-Text 145-154
  Hugo C. Huurdeman; Jaap Kamps
The ever expanding digital information universe makes us rely on search systems to sift through immense amounts of data to satisfy our information needs. Our searches using these systems range from simple lookups to complex and multifaceted explorations. A multitude of models of the information seeking process, for example Kuhlthau's ISP model, divide the information seeking process for complex search tasks into multiple stages. Current search systems, in contrast, still predominantly use a "one-size-fits-all" approach: one interface is used for all stages of a search, even for complex search endeavors. The main aim of this paper is to bridge the gap between multistage information seeking models, documenting the search process on a general level, and search systems and interfaces, serving as the concrete tools to perform searches. To find ways to reduce the gap, we look at existing models of the information seeking process, at search interfaces supporting complex search tasks, and at the use of interface features over time. Our main contribution is that we conceptually bring together macro level information seeking stages and micro level search system features. We highlight the impact of search stages on the flow of interaction with user interface features, providing new handles for the design of multistage search systems.
Stuck in traffic: how temporal delays affect search behaviour BIBAFull-Text 155-164
  David Maxwell; Leif Azzopardi
In this paper we investigate how query response delays and document download delays affect user interactions within a search system. Guided by Information Foraging Theory and Search Economic Theory, five competing hypotheses relating to the behaviours of searchers in the presence of delays are considered and examined in the context of ad-hoc topic retrieval. A between-subjects laboratory study with 48 undergraduate subjects was conducted. Subjects were randomly assigned to one of four conditions that varied the type of delay experienced. When faced with query response delays, subjects did not examine more documents per query as expected. However, when the total amount of time spent per query (a combination of delay and querying time) increased, subjects did examine more documents per query. When faced with document download delays, subjects did not spent more time within documents. Subjects however did spend longer within documents when subjected to both query and document delays. We found a strong and significant correlation between query time (independent of delay) and the interactions of subjects in terms of the number of queries posed, the number of documents examined, and the depth to which subjects went. These findings contrast with previous works on how delays affect search behaviour, and suggest that the theory needs to be refined to make more credible predictions relating to search behaviours.
The effect of cognitive abilities on information search for tasks of varying levels of complexity BIBAFull-Text 165-174
  Kathy Brennan; Diane Kelly; Jaime Arguello
Although web search engines are designed as one-size-fits-all tools, people do not come in one size, but instead vary across many different attributes. One such attribute is cognitive ability. Because information search is primarily a cognitive activity, understanding the extent to which variations in cognitive abilities impact search behaviors and outcomes is especially important. We describe a study in which we explore how people's cognitive abilities affect their search behaviors and perceptions of workload while conducting search tasks with different levels of complexity. Twenty-one adults from the general public completed this study. We assessed participants' associative memory, perceptual speed, and visualization abilities and also measured workload. To evaluate the relationship between cognitive ability, task complexity and workload, we conducted three separate mixed factor ANOVAs corresponding to each of the abilities. Our results suggest three important trends: (1) associative memory ability had no significant effect on search behavior and workload, (2) visualization ability had a significant effect on search behavior, but not workload, and (3) perceptual speed had a significant effect on search behavior and workload. Specifically, participants with high perceptual speed ability engaged in more search activity in less time and experienced less workload. While the interactions were not significant, the differences were more pronounced for more complex tasks. We also found a significant relationship between task complexity and workload, and task complexity and search behaviors, which corroborates previous research.
Investigating people: a qualitative analysis of the search behaviours of open-source intelligence analysts BIBAFull-Text 175-184
  Sean McKeown; David Maxwell; Leif Azzopardi; William Bradley Glisson
The Internet and the World Wide Web have become integral parts of the lives of many modern individuals, enabling almost instantaneous communication, sharing and broadcasting of thoughts, feelings and opinions. Much of this information is publicly facing, and as such, it can be utilised in a multitude of online investigations, ranging from employee vetting and credit checking to counter-terrorism and fraud prevention/detection. However, the search needs and behaviours of these investigators are not well documented in the literature. In order to address this gap, an in-depth qualitative study was carried out in cooperation with a leading investigation company. The research contribution is an initial identification of Open-Source Intelligence investigator search behaviours, the procedures and practices that they undertake, along with an overview of the difficulties and challenges that they encounter as part of their domain. This lays the foundation for future research in to the varied domain of Open-Source Intelligence gathering.
Itinerary recommenders: how do users customize their routes and what can we learn from them? BIBAFull-Text 185-194
  Richard Schaller; David Elsweiler
Itinerary recommenders provide tourists with personalized routes connecting several Points of Interest (POIs). Therefore transit times and users' preferences have to be considered to generate optimal plans. Nevertheless users might appreciate routes being customised to their liking, e.g. based on further contextual factors the system does not know of. Additionally new knowledge on the go, e.g. an unexpectedly overcrowded POI, might make it necessary to adapt plans.
   In this paper we present a system that is able to recommend itineraries and allows users to customize them via manual editing. We investigate, via 2 large-scale naturalistic studies (n=1235 and n=2649), how these editing operations were performed. To this end logs of user interactions with the system were collected. The results of the analysis of these data reveal some surprising usage patterns and point to how such systems can better serve users' needs. Our main conclusion is that itinerary recommendations can benefit from incorporating additional knowledge about users' preferences derived from how users modifiy their route. Moreover, assistance on the go can be improved by suggesting better route alternatives in case of unexpected incidents by imitating the modifications users would manually perform.
A qualitative exploration of secondary assessor relevance judging behavior BIBAFull-Text 195-204
  Aiman L. Al-Harbi; Mark D. Smucker
Secondary assessors frequently differ in their relevance judgments. Primary assessors are those that originate a search topic and whose judgments truly reflect the assessor's relevance criteria. Secondary assessors do not originate the search and must instead attempt to make relevance judgments based on a description of what is and is not relevant. Secondary assessors may be hired to help in the construction of test collections. Currently our knowledge about secondary assessors is largely limited to quantitative measurements of the differences between judgments produced by secondary and primary assessors. In order to better understand the behavior of secondary assessors, we conducted a think-aloud study of secondary assessing behavior. We asked secondary assessors to think-aloud their thoughts as they judged documents. The think-aloud method gives us insight into how relevance decisions are made. We found that assessors are not always certain in their judgments. In the extreme, secondary assessors are forced to make guesses concerning the relevance of documents. We present many reasons and examples of why secondary assessors produce differing relevance judgments. These differences result from the interactions between the search topic, the secondary assessor, the document being judged, and can even apparently be caused by a primary assessor's error in judging relevance. To improve the quality of secondary assessor judgments, we recommend that relevance assessing systems allow for the collection of assessor's certainty and provide a means to help assessors efficiently express their judgment rationale.
Time well spent BIBAFull-Text 205-214
  Charles L. A. Clarke; Mark D. Smucker
Time-biased gain provides a general framework for predicting user performance on information retrieval systems, capturing the impact of the user's interaction with the system's interface. Our prior work investigated an instantiation of time-biased gain aimed at traditional search interfaces utilizing clickable result summaries, with gain realized from the recognition of relevant documents. In this paper, we examine additional properties of time-biased gain, demonstrating how it generalizes effectiveness measures from across the field of information retrieval. We explore a new instantiation of time-biased gain, applicable to systems where the user judges the quality of their experience by the amount of time well spent. Rather than the single number produced by traditional effectiveness measures, time-biased gain models user variability and produces a distribution of gain on a per-query basis. With this distribution, we can observe performance differences at the user level. We apply bootstrap sampling to estimate confidence intervals across multiple queries.

Short papers

Simulated work tasks: the case of professional users BIBAFull-Text 215-218
  Tanja Svarre; Marianne Lykke
This paper investigates simulated work tasks as a tool for information retrieval (IR) evaluation in a work-based, specialized setting. It has been shown that simulated work tasks must be tailored toward the group of study participant to ensure that the depicted situations are realistic and interesting from the participants' point of view [3]. Specifically, we investigate what characterizes an effective simulated work task in a professional government setting and how to design workable tasks for the evaluation of in-house information systems such as a corporate Intranet. The findings reveal that the test participants adopt the tasks. To understand and apply the simulated work tasks, the participants draw on different types of experience and knowledge: Topical, related, structural, and common knowledge. The study also shows that the knowledge types identified are more important for successful retrieval of information than similarity of the simulated work tasks with the participants' daily work tasks.
Visual information search based on knowledge schema modeling BIBAFull-Text 219-222
  Seulki Lee; Achal Shah; Wan C. Yoon
This paper suggests a new interactive system based on visualization of the user's knowledge schema to aid the process of information search and knowledge discovery. The system, inspired from the human memory model, constructs an external representation of the user's conceptual knowledge structure from the user's existing information such as the folder structure and associated tags. In contrast to existing systems based on user models, the system allows users to specify their information needs by selecting a part of their knowledge schema, understand their search results in the context of their existing knowledge organization, and store new information based on their knowledge schema. Through a preliminary evaluation, we show that the visualization of the extracted knowledge structure and the presented relevance of new information was perceived to be useful, even with a basic model of text processing. The participants were able to search for relevant information effectively and expand their information collection with consistency.
Shedding light on a living lab: the CLEF NEWSREEL open recommendation platform BIBAFull-Text 223-226
  Torben Brodt; Frank Hopfgartner
In the CLEF NEWSREEL lab, participants are invited to evaluate news recommendation techniques in real-time by providing news recommendations to actual users that visit commercial news portals to satisfy their information needs. A central role within this lab is the communication between participants and the users. This is enabled by The Open Recommendation Platform (ORP), a web-based platform which distributes users' impressions of news articles to the participants and returns their recommendations to the readers. In this demo, we illustrate the platform and show how requests are handled to provide relevant news articles in real-time.
What does time constraint mean to information searchers? BIBAFull-Text 227-230
  Chang Liu; Fan Yang; Yu Zhao; Qin Jiang; Lu Zhang
In this paper, we explore the relationship between time constraints and users' assessment of their search. A user experiment was conducted. Participants were asked to search under two conditions: with time constraint (TC) and with no time constraint (NTC). The results showed that time constraint did not significantly influence participants' assessment of task difficulty, but significantly influenced users' search confidence and their evaluation of search performance. Particularly, participants were less confident and considered their search performance worse in TC than in NTC conditions. We also found users acquired more new knowledge and had more positive affective states after searching in NTC than in TC conditions. Interestingly, we found time constraints also affect participants' anticipation of time needed to complete the task; participants thought they would need significantly less time to complete the search task when they were given time constraints than without time constraints. These preliminary results suggested that time constraints had remarkable influence upon users' perception of search tasks and their search experience.
Multilingual interface preferences BIBAFull-Text 231-234
  Maria Gäde; Vivien Petras
The most common level of multilinguality in information systems is the adaptation of the interface language. In this paper, the usage of multilingual interfaces is investigated and a comparison between interface language preferences for daily versus occasional usage as well as automatic versus user-triggered interface language change settings is drawn. The study presents the results of a log file analysis of 10 months of Europeana usage data, the digital library for Europe's cultural institutions such as libraries, audio-visual archives, and museums. In total, 1,071,872 sessions from 21 countries are analyzed with respect to their browser, Google referrer and Europeana interface language preferences. Both browser and search engine result page referrer language indicate a strong preference for native language use. In contrast, the analysis of the Europeana interface language use and interface language change indicates weaker preferences for native languages and a stronger acceptance of the default English version. Instead, language information from the query, facet usage as well as objects viewed could reveal language preferences.
Users' criteria of video digital libraries from a public affairs context BIBAFull-Text 235-238
  Boryung Ju; Dan Albertson
This paper reports on a user-centered analysis of video digital libraries. Video digital libraries enable "in the loop" retrieval and playback from centralized and organized collections. As a time-based and multi-channeled format, video digital library systems warrant different considerations for design and information delivery. The purpose of the present study is to collect and initially analyze users' criteria of video digital libraries as part of their interactive experiences. Fifty-two journalism and political science college majors were surveyed, resulting in a total of 242 individual collected responses. Content analysis was performed on the survey responses, and the emergent coding method produced 28 criteria (subcategories) under 5 major categories. Criteria corresponding to Retrieval functions of video digital libraries emerged as the highest priority of the participants, based on its frequency across the responses for the major categories. Criteria corresponding to the User Interface, Collection Quality, User Support, and Organization of Collection followed respectively, in terms of frequency. Cohen's Kappa was .87, indicative of high-level of inter-coder reliability. Findings of the present study provide an initial baseline for design and evaluation of video digital libraries and motivate further research.
Modeling decision points in user search behavior BIBAFull-Text 239-242
  Paul Thomas; Alistair Moffat; Peter Bailey; Falk Scholer
Understanding and modeling user behavior is critical to designing search systems: it allows us to drive batch evaluations, predict how users would respond to changes in systems or interfaces, and suggest ideas for improvement. In this work we present a comprehensive model of the interactions between a searcher and a search engine, and the decisions users make in these interactions. The model is designed to deal only with observable phenomena. Based on data from a user study, we are therefore able to make initial estimates of the probabilities associated with various decision points.
   More sophisticated estimates of these decision points could include probabilities conditioned on some amount of search activity state. In particular, we suggest that one important part of this state is the amount of utility a user is seeking, and how much of this they have collected so far. We propose an experiment to test this, and to elucidate other factors which influence user actions.
Evaluating a tool for the exploratory analysis of usability information using a cognitive walkthrough method BIBAFull-Text 243-246
  Ben Heuwing; Thomas Mandl; Christa Womser-Hacker
Results of empirical usability evaluations in large software-development organizations constitute a valuable asset for these companies. Information needs of usability professionals in these organizations are diverse, and involve both qualitative findings and quantitative data from diverse research methods and sources. Therefore, usability specialists need support for organizing, retrieving, assessing, and analyzing the internal results of usability research. This paper focuses on a method used to evaluate a prototype of a faceted retrieval tool that specifically supports usability specialists accessing a collection of usability results. The evaluation (n=11) was conducted using a primarily qualitative, scenario-based approach. Because of this, it was possible to direct evaluation towards conceptual issues instead of examining details of the surface of the interface. In addition, a survey collected answers to standardized items on the usefulness and ease of use of the system in combination with more domain specific questions. Together, these results provide a valid foundation for the assessment of the usefulness and the relative priority of features.
Modeling the interactive patent retrieval process: an adaptation of Marchionini's information seeking model BIBAFull-Text 247-250
  Julia J. Jürgens; Christa Womser-Hacker; Thomas Mandl
This paper presents an adapted information seeking model for the patent retrieval process, providing a theoretical ground for improvements of information retrieval systems in this special domain. The generation of the model is based on insights gained from the literature and from interviews with patent searchers.
Categorising search sessions: some insights from human judgments BIBAFull-Text 251-254
  Tony Russell-Rose; Paul Clough; Elaine G. Toms
The session is a common unit of interaction that is used in search log analysis. By analysing sessions, it is possible to identify distinct classes of searcher behaviour that can be used to design search applications that better support groups of users based on their expected behaviours. This paper describes an online card sort experiment to investigate how people distinguish between search sessions (i.e., how they group them) to gain insights into their organising principles and to inform the future use of automated approaches, such as clustering. Results show patterns of user behaviour to be the most common way of grouping sessions.
Health information seeking using smartphones among low SES hispanic adults in the U.S.A BIBAFull-Text 255-258
  Henna Kim; Yan Zhang
The Internet-enabled smartphones are readily enabling ubiquitous and continuous access to information. Recent reports showed that Hispanics are more likely to own smartphones and use the mobile Internet than other racial groups in the U.S.A. However, little is known about the mobile access and use of smartphones in seeking health information for this group. This study conducted semi-structured interviews with 20 low SES (socioeconomic status) Hispanics in the U.S.A. Mobile context and situations prompting the adoption of smartphones for health information seeking were explored. The results shed light on how smartphones could help the underserved Hispanics search for health information, narrowing a gap in health disparity. Furthermore, this exploratory study contributes to a more in-depth understanding of mobile context and situations in mobile health information seeking behavior.
Contextual modeling content-based approaches for new-item recommendation BIBAFull-Text 259-262
  Victor Codina; Luis Oliva
The new-item cold-start problem is a well-known limitation of context-free and context-aware Collaborative Filtering (CF) prediction models. In such situations, only Content-based (CB) approaches can produce meaningful recommendations. In this paper, we propose three Context-Aware Content-Based (CACB) models that extend a linear CB prediction model with context-awareness by including additional parameters that represent the influence of context with respect to the users' interests and rating behaviour. The precision of the proposed models has been evaluated using a contextually-tagged rating data set for journey plans in the city of Barcelona (Spain), which has a high number of new items. We demonstrate that, in this data set, the most sophisticated CACB model, which exploits the contextual information at different granularities and also the distributional similarities between contextual conditions during user modeling, significantly outperforms a context-free CB model as well as a state-of-the-art context-aware approach.
Users' reading habits in online news portals BIBAFull-Text 263-266
  Cagdas Esiyok; Benjamin Kille; Brijnesh-Johannes Jain; Frank Hopfgartner; Sahin Albayrak
The aim of this study is to survey reading habits of users of an online news portal. The assumption motivating this study is that insight into the reading habits of users can be helpful to design better news recommendation systems. We estimated the transition probabilities that users who read an article of one news category will move to read an article of another (not necessarily distinct) news category. For this, we analyzed the users' click behavior within plista data set. Key findings are the popularity of category local, loyalty of readers to the same category, observing similar results when addressing enforced click streams, and the case that click behavior is highly influenced by the news category.
Does the perceived usefulness of search facets vary by task type? BIBAFull-Text 267-270
  Kristof Kessler; Luanne Freund; Richard Kopak
This research addresses the need for faceted search systems that can support task-based searching. We report on a systems review carried out to identify the most prevalent facets in current use across three domains and an online questionnaire with 83 responses conducted to assess the perceived usefulness of search facets for different types of search tasks. Results include a ranked list of commonly used search facets. Facets are perceived to be more useful for search tasks motivated by learning goals than those with functional goals (doing tasks). Usefulness scores for specific facets were quite consistent across tasks, so findings do not support the concept of dynamic, task-based display of search facets.
A review of users' search contexts for lifelogging system design BIBAFull-Text 271-274
  Ying-Hsang Liu; Ralf Bierig
The development of mobile and wearable technology has made it possible for people to collect and retrieve large amounts of data about their daily activities. We reviewed selected literature from four related research areas that actively engage in the investigation and modelling of users' search contexts. We discuss their similarities and their potential use for lifelogging. This paper represents a first step toward the conceptualisation of search contexts from an interdisciplinary perspective.
Classifying the influence of negative affect expressed by the financial media on investor behavior BIBAFull-Text 275-278
  Andy Moniz; Franciska de Jong
Prior text mining studies have documented a causal link between human emotions and stock market patterns, yet relatively little research exists into what triggers these emotions. This paper aims to bridge the gap by empirically testing a social psychology theory of human behavior. Underlying our approach lies Attribution Theory, which addresses how observers form causal inferences and moral judgments to explain human behavior, particularly those with negative outcomes. The system presented here works in three stages. The first phase computes a measure of media pessimism by counting negative terms from the General Inquirer dictionary to detect acts of corporate irresponsible behavior. The second phase extends the term-counting approach to capture contextual information. Emotion topic priors are incorporated in a Latent Dirichlet Allocation (LDA) model to infer the financial media's expression of negative affect. Finally, the system combines the two components in an ensemble tree to classify the impact of financial media allegations on a company's stock market patterns. The paper underlines the potential benefit of text mining technology for the support of investor strategies, and more generally demonstrates the power of combining multiple methods for applications in specific domains.
Curiosity driven search: when is relevance irrelevant? BIBAFull-Text 279-282
  Juan D. Millan-Cifuentes; Ayse Göker; Hans Myrhaug; Andrew MacFarlane
Classical information search behaviour models based on work-task scenarios fail to explain common leisure search scenarios motivated by a hedonistic need rather than a defined information need. This paper presents work into such unstructured search driven by curiosity. In order to explore this hedonistic catalyst, a social media search application was designed in which the search experience is triggered by the user's spatio-temporal context during their exploration rather than query-response based information retrieval. We report a study with real users and a simulated casual-leisure search task where results indicated that relevance is not relevant for some searches.
User intent behind medical queries: an evaluation of entity mapping approaches with Metamap and Freebase BIBAFull-Text 283-286
  João R. M. Palotti; Veronika Stefanov; Allan Hanbury
This work focuses on understanding the user intent in the medical domain. The combination of Semantic Web and information retrieval technologies promises a better comprehension of user intents. Mapping queries to entities using Freebase is not novel, but so far only one entity per query could be identified. We overcome this limitation using annotations provided by Metamap. Also, different approaches to map queries to Freebase are explored and evaluated. We propose an indirect evaluation of the mappings, through user intent defined by classes such as Symptoms, Diseases or Treatments. Our experiments show that by using the concepts annotated by Metamap it is possible to improve the accuracy and F1 performances of mappings from queries to Freebase entities.
Tweets I've seen: analysing factors influencing re-finding frustration on Twitter BIBAFull-Text 287-290
  Florian Meier; David Elsweiler
While social networking and microblogging platforms have received considerable research attention, little work has been done to understand how users preserve, manage and reaccess content they acquire from these sources. In this paper we present initial analyses of a large-scale survey (n=606) to understand Personal Information Management (PIM) practices with social networking and microblogging systems, and Twitter in particular. Our results indicate that re-finding information in tweets is a common Twitter activity and can be frustrating. Using questionnaire responses, we investigate the influence of several factors, including how the user tends to preserve tweets of interest, and the level of frustration involved in re-finding such tweets when they are required later.
Using artefacts to investigate children's information seeking experiences BIBAFull-Text 291-294
  Emma Nicol
Pieces of work or "artefacts" produced by children in the form of posters were used in a semi-structured interview to gain insights into children's experience of information seeking in a classroom setting. By referring to information on the poster, children were able to articulate their perceptions of the task, evaluate their success in completing it and reveal which aspects of the task they preferred doing. They were able to say where, and in some cases how information had been found but were generally less able to explain why they had chosen particular pieces of information. The paper concludes that artefacts such as posters can provide a useful entry point for interviewing children about their information behaviour, avoiding some of the known challenges in this.
A social bookmarking system to support cluster driven archival arrangement BIBAFull-Text 295-298
  Marc Bron; Titia van der Werf; Shenghui Wang; Maarten de Rijke
Cultural heritage materials are increasingly being made available through standard search facilities. However, it is challenging to automatically organize these materials in a way that is well aligned with users' specific interests. We report on the development of a social bookmaking system to collect human annotations that are used to measure the performance of three different clustering algorithms. We find that there is a discrepancy between the latent structure present in the data and the clusters annotated by humans. However, it is difficult to detect such discrepancies explicitly.
Browsing patterns in retrieved documents BIBAFull-Text 299-302
  Jaana Kekäläinen; Paavo Arvola; Sanna Kumpulainen
The paper reports a test exploring how retrieved documents are browsed. The access point to the documents was varied -- starting either from the beginning of the document or from the point where relevant information is located -- to find out how much browsing and context the users need to judge relevance. Test results reveal different within-document browsing patterns.
Using sensor graphs to stimulate recall in retrospective think-aloud protocols BIBAFull-Text 303-307
  Gabriele Pätsch; Thomas Mandl; Christa Womser-Hacker
The method of stimulated recall, also called retrospective think-aloud protocol, may draw subjects' attention rather towards their actions and thoughts than to their emotional experience of the search process. This report presents a variation of this method, which aims at eliciting a more extensive verbalization of feelings. In a qualitative study the search activities of seven participants were recorded by a screen recording software, and their electrodermal activity was measured by a skin conductance sensor. Subsequently, a guided interview was conducted, which was followed by the retrospective think-aloud protocol. The protocol consisted of two phases with different stimuli: At first, only the screen recording was displayed to the participants, and afterwards, it was supplemented by a synchronous video of the physiological sensor data. This paper gives a full description of the technique, including recommendations for its practical application. Overall, the presented results show that more events were reported using this method.

Demos

X-REC: cross-category entity recommendation BIBAFull-Text 308-311
  Dragan Milchevski; Klaus Berberich
We demonstrate X-Rec, a novel system for entity recommendation. In contrast to other systems, X-Rec can recommend entities from diverse categories including goods (e.g., books), other physical entities (e.g., actors), but also immaterial entities (e.g., ideologies). Further, it does so only based on publicly available data sources, including the revision history of Wikipedia, using an easily extensible approach for recommending entities. We describe X-Rec's architecture, showing how its components interact with each other. Moreover, we outline our demonstration, which foresees different modes for users to interact with the system.
Interactive summarization of social media BIBAFull-Text 312-315
  Wen Li; Carsten Eickhoff; Arjen P. de Vries
Data visualization and exploration tools are crucial for data scientists, especially during a pilot study. In this paper, we present an extensible open-source workbench for aggregating, summarizing and filtering social network profiles derived from tweets. We briefly demonstrate its range of basic features for two use cases: geo-spatial profile summarization based on check-in histories and social media based complaint discovery in water management.
YASFIIRE: yet another system for IIR evaluation BIBAFull-Text 316-319
  Xing Wei; Yinglong Zhang; Jacek Gwizdka
We present a system that supports Interactive Information Retrieval user studies on the Web. Our system provides support for user and task management, for processing web-based task specific interfaces and for Web-event logging. It also offers functionality useful to IIR studies that capture eye-movement on Web page elements. The system complements logging functionality offered by a typical usability/eye-tracking software packages and is designed to act in concert with such software.
Mobile tourist guides: bridging the gap between automation and users retaining control of their itineraries BIBAFull-Text 320-323
  Richard Schaller
We present a mobile tourist guide for planning and conducting sightseeing day trips. Users are provided different means to access and select the available sights, events and other points of interest (POIs): Via a hybrid recommender system, via browsing by sight category, via searching over descriptions of POIs or via browsing on a map. Based on user's selection a route planner for time-constrained activities generates route suggestions taking additional constraints for public transport connections into account.
   A novelty of the implemented approach compared to existing solutions for tourists is that the user retains full control over the tour by diverse interaction possibilities: Before route generation different means for selecting POIs are provided, during route generation multiple route variants are suggested and after route generation users are able to directly edit any detail at any time, even if there are existing constraints that hinder the direct execution of an edit operation. Moreover, recommender, planner and editing are closely interconnected: Recommendations are used by the planner to fill-up unavoidable gaps. This may also be initiated manually during editing where also parts of the planner are involved to permit only those edits that keep the route feasible. The app is currently tailored to the city of Nuremberg but can be extended for other cities as well.

Doctoral consortium

Web search for instruction-related information: Why? Where? How? BIBAFull-Text 324-326
  Angela Vorndran; C. Womser-Hacker; M. Rittberger
In this paper I would like to outline the theoretical concept, methods and research questions of my dissertation project. Following the theory of Information Use Environment by Taylor (1991) and its interpretation in the context of Web search by Detlor (2003), this investigation looks into the information behavior of teachers when searching the Web for instruction-related information. Typical problem situations of the work context are identified as well as relevant information traits and approaches to information search and use. Furthermore, the use of several search environments with different degrees of interactive character are analyzed comparatively. Taking into consideration a variety of data sources to gain a multi-perspective view on the issue, usage data of the German Education Server (GES) are analyzed in addition to discussion fora on teacher-specific websites and complemented by qualitative interviews.
Searcher heterogeneity in collaborative information seeking within the context of work tasks BIBAFull-Text 327-329
  Stefanie Elbeshausen; Christa Womser-Hacker; Thomas Mandl
In this paper we present an overview over a doctoral thesis project emphasizing the planned research and methodology. The thesis project focuses on Collaborative Information Seeking (CIS) with a core interest in searcher heterogeneity. A special field of interest is the way how different personalities interact in a collaborative search scenario and if heterogeneous groups perform better or worse than homogenous groups. The goal is to provide a description of group performance in CIS and suggestions for the design of collaborative search tools.
Primary school children's single topic search over multiple search sessions: does search evolve? BIBAFull-Text 330-332
  Sophie Rutter
Children's search behavior has mainly been studied in single search sessions where the search task is new to the children. In this proposal it is suggested that an investigation into children's search on a single topic across multiple search sessions may reveal different search behaviour to what is already known.
Measuring and improving data quality of media collections for professional tasks BIBAFull-Text 333-335
  Myriam C. Traub
Carrying out research tasks on data collections is hampered, or even made impossible, by data quality issues of different types, such as incompleteness or inconsistency, and severity. We identify research tasks carried out by professional users of data collections that are hampered by inherent quality issues. We investigate what types of issues exist and how they influence these research tasks. To measure the quality perceived by professional users, we develop a quality metric. This allows us to measure the suitability of the data quality for a chosen user task. For a chosen task, we study how the data quality can be improved using crowdsourcing. We validate our quality metric by investigating whether professionals perform better on the chosen research task.
Develop, implement, and improve a web session detection model BIBAFull-Text 336-338
  Chaoyu Ye; Max L. Wilson; Tom Rodden
More research in web and Information Retrieval is turning towards session-based retrieval rather than single item or query investigation. However, most of the session detection attempts only used simplistic rules (e.g. "30 mins inactivity creates a new session"). Up to this point, there are various fuzzy definitions of session, but no general consensus about it in the literature [3]. Whilst comparably little work has involved the mental model about the "web session" from real users. In response to these, my research focuses on web session detection involving real users with a comprehensive set of factors identified by them rather than the "simple fixed timeout". My objective is to develop a session detection model with corresponding rules for each factor, and then embedded them into a Chrome Extension to automatically detect more accurate web sessions from log data.
Personal information management and social networks re-finding on Twitter BIBAFull-Text 339-341
  Florian Meier; David Elsweiler
The following PhD project argues for analyzing Personal Information Management (PIM) behaviour on social networks and microblogging platforms, such as Twitter, as such services go far beyond their expected usage and have grown a viable source of information to many users. Of the three main information interaction activities that characterise PIM (finding -- keeping -- re-finding), only the retrieval of information has received researcher attention on such services. However, recent literature suggests that other PIM related activities occur, that should be further investigated. In addition to motivating the project, this position statement outlines the main research questions and approach, details the current status, which sheds light on how people perform PIM activities in the context of social networks and microblogging platforms. Here we detail some of the key findings and describe our tentative plans for the future. It is this plans that I wish to discuss in the doctoral consortium.
Too little time?: time constraints and time pressure in information search BIBAFull-Text 342-344
  Anita Crescenzi
People often experience time pressure when searching, yet the impact of time pressure on search behaviors and outcomes has not been studied extensively by the information science and information retrieval communities. In addition, experimental interactive information retrieval studies often include researcher-imposed time limits that may influence participants' search behavior and outcome measures in ways that are not well understood. This research proposes a systematic investigation of imposed time constraints on frequently used interaction measures and participants' perceptions of the search process and outcome.
Combining document retrieval with knowledge graphs for exploratory search BIBAFull-Text 345-347
  Bahareh Sarrafzadeh; Olga Vechtomova
With the massive increase in information availability, it gets more and more difficult to make sense of the available information. The Web has provided the opportunity to browse and navigate through the extensive information space by utilizing the modern search engines. This in turn has led to increasing expectations to use the Web as a source for learning and exploratory discovery. Although current Information Retrieval (IR) methods satisfy simple and straight-forward needs, they do not offer enough support for the users with complex search tasks which involve learning and investigation activities. In my PhD research I aim to support different aspects of information seeking that are observed in exploratory activities. I propose a new framework based on combining knowledge graphs with document retrieval in order to effectively improve search breadth and quality.
Designing autotelic searching experience for casual-leisure by using the user's context BIBAFull-Text 348-350
  Juan D. Millan Cifuentes; Ayse Göker; Andrew MacFarlane
Which is more important: the journey or the destination? Classical Interactive Information Retrieval (IIR) based on work-task scenarios usually puts the emphasis on the destination of the search (the results) with metrics such as precision and recall rather than the search journey. But social media, mobile devices and other pervasive technologies have made information accessible to people in leisure scenarios and open up casual-leisure search behaviours motivated by hedonistic need such as having fun, or relaxing instead of a well-defined information need. During search sessions users might find irrelevant information but they may keep exploring because the IR system satisfies their current leisure need. This research aims to understand better casual-leisure search behaviour and design new IR systems to support autotelic search experiences.
Investigation of information behavior in Wikipedia articles BIBAFull-Text 351-353
  Barbara Rösch
This work aims to explore information behavior in selected Wikipedia articles. To get insights into users' interaction with pictorial and textual contents eye-tracking experiments are conducted. Spread of information within the articles and the relation between text and images are analyzed.
Adaptive search systems for web archive research BIBAFull-Text 354-356
  Hugo C. Huurdeman
The wealth of digital information available in our time has become indispensable for a rich variety of tasks. We use data on the Web for work, leisure, and research, aided by various search systems, allowing us to find small needles in giant haystacks. Despite recent advances in personalization and contextualization, however, various types of tasks, ranging from simple lookup tasks to complex, exploratory and analytical ventures, are mainly supported in elementary, "one-size-fits-all" search interfaces.
   Web archives, keepers of our future cultural heritage, have gathered petabytes of valuable Web data, which characterize our times for future generations. Access to these archives, however, is surprisingly limited: online Web archives usually provide a URL-based Wayback Machine interface, sometimes extended with rudimentary search options. As a result of limited access, Web archives have not been widely used for research so far. For emerging research using Web archives, there is a need to move beyond URL-based and simple search access, towards providing support for complex (re)search tasks.
   In my thesis, I am exploring ways to move beyond the "one-size-fits-all" approach for search systems, and I work on systems which can support the flow of complex search, also in the context of archived Web data. Rich models of search and research can be incorporated into adaptive search systems, supporting search strategies in various stages of complex search tasks. Concretely, I look at the use case of the Humanities researcher, for which the large, Terabyte-scale Web archives can be a valuable addition to existing sources utilized to perform research.