Dynamic Compositions: Recombining Search User Interface Features for Supporting Complex Work Tasks Hugo C. Huurdeman University of Oslo Norway h.c.huurdeman@ub.uio.no ABSTRACT of optimizing results displayal to singular queries, we propose a Due to the tremendous advances in information retrieval in the fundamentally different approach, involving support for a user’s past decades, search engines have become extremely efficient at information seeking process. acquiring useful sources in response to a user’s query. However, The non-trivial question which follows is how to concretely for more sustained and complex information seeking tasks, these achieve this enhanced process support. In the context of this paper, search engines are not as well suited. During complex information we focus on the presentation of results from search engines via seeking tasks, various search stages may occur, which imply varying their constituent SUI features. Creating compositions of interface support needs for users. However, the implications of theoretical features with a high usability is no easy task. As Oddy [13] already information seeking models for concrete search user interfaces argued in 1977, the “art” of information system design is to “find (SUI) design are unclear, both at the level of the individual features the form and timing of information presentation which will best and of the whole interface. Guidelines and design patterns for con- aid the system user in whatever task he has in hand.” In this paper, crete SUIs, on the other hand, provide recommendations for feature we focus on the timing and form of SUI features, assessing how design, but these are separated from their role in the information they fit in different stages of the information seeking process, and seeking process. This paper addresses the question of how to design how they can potentially be recombined in dynamic ways. SUIs with enhanced support for the macro-level process, first by To this end, we first discuss background literature related to reviewing previous research. Subsequently, we outline how three process support for complex tasks (Section 2). Based on previous types of SUI features can be recombined to form a supportive frame- research, we then outline our supportive framework for designing work for complex tasks. We provide concrete recommendations for task support in terms of SUI features (Section 3), followed by the designing more holistic SUIs which potentially evolve along with a discussion and conclusion (Section 4). user’s information seeking process. 2 BACKGROUND KEYWORDS This paper focuses on cognitively complex tasks, during which information seeking; search user interfaces; search stages search systems may act as a mediator between user & information. 2.1 Complex Tasks 1 INTRODUCTION Unlike simple lookup tasks, complex work tasks [16] may involve Tremendous advances in information retrieval technology have learning and construction, understanding and problem formulation occurred during the past decades. We now have arrived at the point [3]. These tasks can be performed by topic novices, but also by where systems may actually solve problems for users. For instance, more experienced actors. For instance, a student may perform a via common search engines on the web we get ‘instant answers’ task involving a topic she knows little about, but this knowledge for factual questions ranging from the weather in the next week- advances over time, or a researcher may start with a loose research end to the birthdate of the current prime minister. Information question, which becomes more focused after interaction with a seeking in the context of more complex tasks, however, is not as set of information. Besides their obvious occurrence in a work straightforward: broader inquiries cannot be directly answered in and study contexts, complex tasks are also performed in leisure a succinct snippet of information. For instance, gaining novel ideas settings, e.g. shopping for products which are inherently complex. for research, or finding the appropriate sources for writing an essay The complexity of information seeking and searching has been requires intensive interaction with information sources. captured in a wide variety of models (see e.g. [18]). During the process of information seeking and use, as occurring in complex research-based tasks, the needs and understanding of 2.2 Information Seeking Models a user may evolve, moving from broad conceptualizations to a In this paper, we focus on models looking at information seek- focused perspective. To create more supportive systems for complex ing as a temporal process. Kuhlthau’s Information Search Process tasks featuring sustained information interaction, current ad-hoc model is an influential model [10], based on several longitudinal approaches to search-based interaction should be rethought. Instead studies. A key aspect of the model is that it looks at information CHIIR 2017 Workshop on Supporting Complex Search Tasks, Oslo, Norway. searching as a process of knowledge construction, during which a Copyright for the individual papers remains with the authors. Copying permitted user’s uncertainty fluctuates. The model focuses on the evolution for private and academic purposes. This volume is published and copyrighted by its editors. Published on CEUR-WS, Volume 1798, http://ceur-ws.org/Vol-1798/. of users’ thoughts, feelings and actions across six broad stages. These include early stages of initiation and topic selection, as well as CHIIR 2017 Workshop on Supporting Complex Search Tasks, March 11, 2017, Oslo, Norway. Hugo C. Huurdeman exploration. At a certain point, a focus is formulated, after which in- a resultset may decrease. Personalizable features tailor the experi- formation seeking changes, and stages of collection and presentation ence to a user, based on her actions. Contrary to input and control follow. Based on other longitudinal studies, Vakkari [15] observed features, personalizable features became more useful over time [8]. implications for information sought, assessed relevance and search tactics, terms and operators. He grouped Kuhlthau’s stages into 3 TOWARDS A HELPFUL FRAMEWORK FOR three stages: pre-focus, focus formulation and post-focus. COMPLEX TASKS 2.3 Search User Interfaces As illustrated by the information seeking models discussed in the previous section, a searcher’s conceptual framework about a topic Already in the 1970s, researchers looked at challenges in design- may evolve over time. During a novice user’s information journey, ing interfaces for (bibliographic) search systems [2], including the knowledge structures evolve, just as during a scholars’ research pro- characteristics of searchers, the search environment and feedback cess, conceptualizations of a topic may undergo changes. Keeping to searchers. However, even though various early experiments re- this evolution in mind, the system should form a “helpful frame- sulted in “intelligent intermediary systems” [9, p.137], this research work within which the user can make problem-solving decisions” in the 1990s gave way to streamlined IR systems, often focusing [13]. However, current search interfaces typically do not evolve on query formulation and inspection. Motivations behind the sim- with a user’s knowledge – to become truly ‘helpful’, a system should ple design are multifold: search tasks are usually part of larger ideally support the information seeking process of a user, moving work tasks, and the interface should distract as less as possible [6]. from exploratory pre-focus, to focus formulation and final post-focus Notwithstanding the apparent simplicity of current search inter- stages. Our proposed framework is visualized by Figure 1, and con- faces, the “art” of designing them is still complex. Over the years, sists of three dimensions. As context, we use SUI features listed in however, a number of frameworks, guidelines and design pattern [17], augmented with more recently introduced features. libraries have been created [14]. Despite the immediate value of those frameworks for creating appropriate search user interfaces, they mainly focus on designing the functionality of SUI elements 3.1 First Dimension in the best way1 . It is unclear at which moments of complex tasks The first dimension of a system constituting a ‘helpful framework’ these features are most useful, and how they can be combined to consists of features offering automatically generated suggestions support (and not impede) complex searches. A higher-level system to users. This support typically takes place at Bates [1]’s search perspective has been provided by Bates [1]. The “degree of user vs. activity level of the ‘move’ (e.g. entering search terms), and ‘tactic’ system involvement in the search” encompasses a continuum, rang- (e.g. choosing a broader term). For instance, a word cloud feature ing from fully manual search activities to fully automated searches. may suggest keywords for a query, or a query suggestion feature Furthermore, she distinguishes various levels of search activities. may propose a broader formulation of a query. The need for this The lower level activities are moves (simple actions) and tactics (one low-level support, embodied in various input and control features, or more moves to further a search), while higher level activities generally decreases over time. When a user’s conceptualization include stratagems (a complex set of tactics and moves), and strate- of a topic grows, she becomes increasingly able to express herself gies (a plan for the entire information search). Bates’ work may precisely in the context of that topic [8, 10], and support at the level provide inspiration for a better understanding of system support of moves and tactics becomes more superfluous. across stages. An SUI designer has a wide variety of features at her disposal to provide low-level support for searching. First of all, at the level 2.4 From Stages to Interfaces of the query, Query Corrections, Query Autocomplete, and As we argued in [7, 8], there are issues in the translation from the Query Suggestions (a) can provide help in formulating the right rich stages in the information seeking literature to concrete sup- query, and suggesting alternative queries. Especially in initial stages, port in terms of search system features. These papers looked at the Facets and Filters (b) can be useful to delineate resultsets, and stages in which SUI features would provide support, also taking into adapting Results Ordering (c) may initially help to find the right account previous literature [4, 12, for example]. Huurdeman et al. items. Word Clouds (d), even though their effectiveness in infor- [8] used a feature categorization from Wilson [17] to more broadly mation searching has shown fluctuating results, may also provide group different types of SUI features, and assessed their value over inspiration. Finally, current search interfaces often contain Entity time using a multistage task design. Informational features, showing cards (e), an information panel with brief information and related search results or information about results, were naturally useful in entities for an intended query target. all information seeking stages. Input and control features, to express needs and modify input, on the other hand, could be categorized 3.2 Second Dimension as search stage sensitive features. The value of these features was The second dimension of a ‘helpful framework’ is formed by in- highest in the initial pre-focus stage, and decreased over time. This formational features. These features provide the actual results, or reflects a user’s increasing understanding of a topic, during which information about encountered result items. For instance, a search the value of features to help formulating a query and delimiting system may show the title of a document, a short snippet and basic 1 For instance, how to design a ‘pagination control’ feature for a search metadata. As evidenced in previous experiments (e.g. [8]), these engine, https://developer.yahoo.com/ypatterns/navigation/pagination/search.html features may be useful throughout the process. They provide low- (accessed: 01/08/16) level support at the move and tactic level, for instance selecting Dynamic Compositions relative importance CHIIR 2017 Workshop on Supporting Complex Search Tasks, March 11, 2017, Oslo, Norway. high-level support 3 strategems & strategies h i j k personalizable features low-level support 2 moves & tactics f g h informational features pre-focus focus post-focus 1 task progress a b c d e input & control features Figure 1: Schematic overview of a supportive framework for designing ‘stage-aware’ search user interfaces for complex tasks: low-level support for moves and tactics gradually gives way to higher level support for stratagems and strategies. and opening information sources, but also higher level support (e.g. [11]. Other tools which may be useful, sometimes only in passive offered by visualizations of result sets). ways [8] are Query History (j) features. Finally, External tools Informational features may provide both low and high-level sup- (k) may provide high-level support, such as word and data process- port. These features contain the Search Results (f) themselves ing, as well as reference management. (commonly shown by their title and short textual snippet). Espe- Summarizing, more dynamic support for complex research-based cially in e-commerce systems, also Thumbnails (g) depict resultset tasks may be achieved by differentiating SUI feature categories items. Visualizations (h) provide more insights into retrieved re- and their levels of support. In particular, functionality providing sultsets. These may initially be useful for a researcher to explore a low-level support (i.e. input and control features), are useful in the set of data, but also to visualize a gathered set of focused results. initial stages of a complex research-based task. Searchers with low domain knowledge, but also researchers exploring a new topic and 3.3 Third Dimension collection may utilize this functionality to bootstrap their searches. The third dimension of a ‘helpful framework’ consists of features Features providing high-level support (in particular personalizable which can support seeking at a higher level. While these types features), may invite searchers to explicitly reflect and interact with of features may include automated functionality, the main aim is results, as well as seeing how these results fit in their process and to provide insights into a user’s process through her actions. As strategy. Kuhlthau’s model has indicated, processes of hypothesis genera- tion, data collection, information organization and the preparation 4 DISCUSSION AND CONCLUSION of a personalized synthesis of a topic take place during processes of The road towards designing optimal search user interfaces for com- knowledge construction [10, p.194]. This reflects the highly person- plex tasks is long and winding. Indeed, the design of SUIs can be alized nature of such complex activities, meaning that automated seen as an “art”, involving numerous thorny issues and trade-offs support may not suffice. Instead, the aim of personalizable features in usability. For instance, combining excessive sets of features may should be to aid users in performing their task. In different experi- overload the user, while a streamlined approach can be too limiting ments, demand for and use of annotation, saving and organization for supporting user needs in different stages of complex tasks. At features by both students and graduate researchers has been evi- each stage of a task, an optimal combination of features may ex- denced. As opposed to low-level features, these higher-level features ist. This paper provides initial handles to determine the relative may support Bates’ ‘stratagems’ and ‘strategies’ (planning in the importance of features when designing SUIs, thus connecting the- context of an entire search). On the one hand, through logging oretical information seeking models and more concrete search user user’s actions and potentially gathering data about the actors’ do- interface design. main knowledge or task at hand, they provide a trail of activities, which may (passively) aid users in locating where they are in the At the level of the whole SUI, various approaches for the provision of process. On the other hand, they also allow a user to ‘work with dynamic support for information seeking stages can be suggested. results’, and thus encourage reflection on encountered results. As First of all, a totally open approach is possible – searchers are such, they become increasingly useful throughout a task. free to choose a custom set of SUI features at any point of the More high-level support throughout the process may be offered process (“build your own SUI”). Second, predefined interface panels by Results Saving (h) features, alternatively embodied in e.g. shop- combining features can be offered to a user (e.g. for exploration and ping carts and wishlists. Interfaces may also offer Personal results focused search), and a user can choose a panel she needs at any Organization opportunities. Furthermore, especially in a research stage (as evaluated in [5]). Third, a totally adaptive approach may context, Annotations (i) are used at different points in the process be followed: using evidence from usage data, interface features are CHIIR 2017 Workshop on Supporting Complex Search Tasks, March 11, 2017, Oslo, Norway. Hugo C. Huurdeman automatically offered or disabled. Hence, the potential adaptation [3] K. Byström and K. Järvelin. Task complexity affects information of interfaces for complex tasks spans a continuum, ranging from seeking and use. IP&M, 31(2):191–213, 1995. fully manual to entirely automatic approaches. [4] A. Diriye, A. Blandford, A. Tombros, and P. Vakkari. The role of search It would be valuable to gain further insights into the influence of interface features during information seeking. In TPDL, LNCS vol dynamic presentation of search stage-sensitive SUI features on user 8092, pages 235–240. Springer, 2013. [5] M. Gäde, M. Hall, H. Huurdeman, J. Kamps, M. Koolen, M. Skov, satisfaction (i.e. the features within the first and third dimension E. Toms, and D. Walsh. Overview of the Interactive Social Book of the framework discussed in Section 3). In the CLEF Interactive Search Track. In CLEF 2015 Notes, volume 1391. CEUR, 2015. Book Search Track, users were able to select interface panels repre- [6] M. Hearst. Search User Interfaces. Cambridge University Press, 2009. senting different search stages, suggesting positive effects on user [7] H. C. Huurdeman and J. Kamps. From multistage information-seeking engagement [5]. Future studies should further look at the impact models to multistage search systems. In Proc., IIiX ’14, pages 145–154, of dynamic and adaptive presentation of SUI elements, especially 2014. ACM. since this influences the consistency of an interface. This may be [8] H. C. Huurdeman, M. L. Wilson, and J. Kamps. Active and Passive tested by adaptively enabling and disabling SUI features in exper- Utility of Search Interface Features in Different Information Seeking imental systems with rich functionality in a (simulated) complex Task Stages. In Proc., CHIIR ’16, pages 3–12, 2016. ACM. work task setting. [9] P. Ingwersen and K. Järvelin. The Turn - Integration of Information Seeking and Retrieval in Context. Springer, Dordrecht, 2005. At the level of atomic SUI features, this paper briefly outlined [10] C. C. Kuhlthau. Seeking meaning: a process approach to library and feature utility during the information seeking process, based on information services. Libraries Unlimited, Westport, Conn., 2004. Bates [1] levels of search activities (i.e. moves, tactics, strategies and [11] L. M. Melgar, M. Koolen, H. C. Huurdeman, and J. Blom. A Process strategems). Further research is needed to allow for making more Models of Scholarly Media Annotation. In Proc., CHIIR, 2017. ACM. conscious choices of which features to include in an interface, based [12] X. Niu and D. Kelly. The use of query suggestions during information on the purpose they serve in the process. For instance, we may use search. Inf. Process. Manage., 50:218–234, 2014. [13] R. N. Oddy. Information retrieval through man-machine dialogue. J. Bates’ levels of search activities as a ‘lens’ for analyzing existing Doc., 33(1):1–14, Jan. 1977. SUI features. [14] B. Shneiderman and C. Pleasant. Designing the user interface: strategies Furthermore, as suggested in [8], individual features could be for effective human-computer interaction. Pearson Education, 2005. improved by taking previous user interactions as a basis and thus [15] P. Vakkari. A theory of the task-based information retrieval process: becoming more personalizable. For instance, query suggestions can a summary and generalisation of a longitudinal study. J. Doc., 57(1): lose their value over time due to a user’s increased knowledge 44–60, Feb. 2001. [8], but may provide more “intelligent” suggestions by taking into [16] B. Wildemuth, L. Freund, and E. G. Toms. Untangling search task account previous user interactions. complexity and difficulty in the context of interactive information retrieval studies. J. Doc., 70(6):1118–1140, 2014. The presented framework is just an initial step towards a more holis- [17] M. L. Wilson. Search User Interface Design. Synthesis Lectures on tic approach for SUI design. First of all, it needs further grounding Information Concepts, Retrieval, and Services, 3(3):1–143, Nov. 2011. in actual SUI design practice, in particular with respect to current [18] T. D. Wilson. Models in information behaviour research. J. Doc., 55: systems ‘in the wild’, and with respect to previous research studies 249–270, 1999. and observations. Further research on the utility of SUI features, as well as more high-level SUI functionality in search systems is needed. For instance, explicit support for Bates’ strategems and strategies is still rare, 27 years after her seminal paper. However, the ubiquitous presence of search engines in diverse manifestations may allow for more inclusive views on user activities in consecutive stages of complex search processes. By adapting low and high-level support, thus creating dynamic SUI compositions, we may be able to arrive at a more “intellectual symbiosis” between user and system as envisioned by Bates [1]. 5 ACKNOWLEDGMENTS Related to this paper’s topic, the author wishes to thank Jaap Kamps for invaluable discussions and advice, as well as Max Wilson for earlier collaborations. REFERENCES [1] M. J. Bates. Where should the person stop and the information search interface start? Inf. Proc. Man., 26(5):575–591, Jan. 1990. [2] J. L. Bennett. Interactive bibliographic search as a challenge to interface design. In Interactive bibliographic search: The user/computer interface, pages 1–16. AFIPS Press, 1971.