Complex Tools for Complex Tasks Elaine G. Toms University of Sheffield Information School, Regent Court Sheffield, UK S1 4DP e.toms@sheffield.ac.uk ABSTRACT restrictive conditions that were combined with many Boolean How do we design interfaces to better support complexity in the operators to create a lengthy intricate search command with the search process? The intent of this paper is to illustrate that the objective of achieving the needed response in one single albeit search is not the complex piece, it is the search within its context expensive operation. This was a learned skill performed by expert that makes the process complex. searchers who used complex queries. With the emergence of direct manipulation interfaces and the internet/web, search became a Categories and Subject Descriptors task/tool of the masses, with diminished economic costs and with H.5.2 [User Interfaces]: H.3.3 [Information Search and the physical actions required to search now accessible to/usable Retrieval] by all. While we can consider search now to be ‘easy,’ there are General Terms limitations. The former sophistication is no longer possible Design, Human Factors. leaving the user struggling to make sense of the process. What Keywords was once a single interaction now takes multiple steps to achieve a similar result. At the same time, the sophistication in search Search tools; Search interface features. algorithms means that a query now retrieves mostly relevant 1. INTRODUCTION results. In a study of over 400 searchers, Toms et al (2014) found How do we design interfaces to better support complexity in the that the system delivered relevant results, but the users continued search process? First I argue that the act of searching is not the searching as they appeared not to know when to stop, leaving the problem, nor what really represents the complexity of this authors to conclude that the interface failed the users by not process. The complexity emerges from the work task and its guiding them through the process beyond search to use. Over 30 context (Jarvelin & Repo, 1982; Bystrom & Jarvelin, 1995; years ago, Jarvelin and Repo (1982) concluded that user “seems to Vakkari, 1999;), challenging how the user approaches the search happen just by itself,” which remains equally true today. task, finds all of the essential information, extracts what is Arguably, complex search tasks do not exist (which suggests that required, and consequently uses that information (likely in I am contravening the very objective of this workshop). The combination with multiple searches to multiple systems) to complexity of search resides with the work tasks (Jarvelin & resolve the work task. The concept of complexity with respect to Repo, 1982; Bystrom & Jarvelin, 1995) from which search tasks search has had a long history (see discussion Bystrom & Jarvelin emanate. A search only exists to support that task (Vakkari, (1995) as well as Wildemuth et al.’s (2014) critical analysis of the 1999). While a search task may range from a simple Q&A to one concept). intended to compare multiple options, each of which requires an Secondly, I propose a set of tools that are complementary to, and independent search, the search task itself remains a single tightly integrated with the search box to facilitate users’ work task dimensional element. completion. The tools are essentially apps or dynamic interface In general, a work task has a defined objective or goal that can features that theoretically should reduce the complexity of the readily be mapped to an at-the-moment unknown outcome or process, providing “all the tools that are commonly used for a result. There may be known conditional/unconditional specific task or workflow” (Feldman & Reynolds, 2010). Notably requirements. To complete the task requires a series of activities despite the number of amount of discussion that complexity and each of which uses actions performed with a set of tools. These search have received, we still do not know what design features tools, one of which is a search tool, consume information/data might alleviate complexity (Fidel, 2012), and the user’s from diverse sources (see model of the process augmented and perception of mental effort required to do the job. Perhaps that is expanded from Bystrom & Jarvelin (1995) in Toms (, 2011)). because we are addressing complexity with respect to the wrong This is a generic description of how search fits within the larger concept. work task. But this is not a single process; there may be multiples of these and within each, search may (or may not) be required. Think about the process used to piece together a response to a 2. SEARCH AND WORK TASKS request, write a report, make a decision, or plan for a future event. In the 1960s to 1980s, search was complex; a search string was There may be multiple sub processes in addition to search that composed of multiple units of synonymous terms sometimes with requires extraction, analysis, comparison, prediction, modification, and/or manipulation, for which the found Copyright   2015   for   the   individual   papers   by   the   papers'   authors.   information is a critical ingredient. Thus a work task is usually Copying  permitted  for  private  and  academic  purposes.  This  volume   complex; the searching is for discrete pieces of information, is  published  and  copyrighted  by  its  editors.     analogically like the ingredients brought together for a recipe, and ECIR  Supporting  complex  Search  Task  Workshop,  Vienna,  Austria.   like a recipe, search is just one of the many elements required to Published  on  CEUR-­‐WS:  http://ceur-­‐ws.org/Vol-­‐1338/   get the job done. 3. INTERFACE TO SUPPORT SEARCH provided at this point. At each step in that search procedure a user Interfaces to applications intended to support those complex work performs a series of actions but at present very little support is tasks are very limited. In fact, arguably they are non-existent. provided by the system. Spell check, query word suggestion, some Most of the typical complex work tasks in knowledge work use of relevance feedback, and some filtering are perhaps the environments are completed using the lowly word processor that extent. Even this set is somewhat incomplete. is partnered with perhaps a browser, with one of more windows/tabs used to access the essential information, and Missing from these features are those intended to assist the user possibly additional applications such as spreadsheets, and concept with the intensive cognitive processin in the pre- during and post- mapping tools which are also used to manipulate or analyse some search phases of each sub process of each work task. One can of that found information. At present, search remains mostly un- envisage a series of tools or software applications along the lines integrated with work task and tends to be treated as an of those suggested in Table 1. Not unlike the toolbox used to independent activity. create images, or those used to analyze a bunch of numbers for patterns and relationships, this list identifies a distinct set of Instead we look to the emerging work on re-thinking the generic specialty tools to support primarily user cognitive limitations at search interface with a view to the proposition that it may be that point in the process. The list includes an extension of directly integrated with a workplace digital solution in the future. potential and existing features (or tools or apps or modes) some of Initially started as command-line interface, the implementation of which have been previously identified by Hearst (2009), Russell- search evolved into a form-fill-in style of interaction enhanced Rose and Tate (2011), Wilson (2011) and White (2011). Such a with direct manipulation elements – a search box and command set would remove some of the existing complexity from how the button, or else a faceted style interface implemented as a series of current search process is implemented. menus with filtering options. This approach to thinking of the search interface as query formulation/reformulation (or concept The set is divided into four grouping with some repetition from selection), search results display and/or some visualization of group to group, and with variation in specificity as the type of these (Marchionini, 1997; Hearst, 2009) predominates action or procedure desired varies with the point in the process. developments even today, although multiple variations have Prior to searching, a user sets a goal to be achieve; is the appeared in research projects. information needed as evidence, to find out how to do something, i.e., a procedure, or simply to gain an overview of the topic. These Perhaps the first systematic approach to search interface design is tools clearly would need to relate to relevance and use, targetting Wilson’s four key features: any search output to deliver on the goal, and only on that goal. 1. Input: components for a user to indicate the search intent; The tools to aid query development have perhaps been the most 2. Control: components that enable query expansion, modifying robust and established to date. At present the user is left to his or and augmenting the input; her own devices over the course of a search, relying on 3. Informational: components that output the results, or recognition more so that (cognitive) recall, and with an emphasis information about the results; on memory and mental storage. Much could be done at this phase 4. Personalisable: components that relate to characteristics of the to assist. For example, we spend significant efforts in identifying user and/or their past interactions. similarity with academia, but we fail to supply the same technique to the user in identifying similarity amongst a set of research In addition, White (2011) identified a set of techniques that papers. We do little to assist the user in temporary storage such as specified how some of those components can be rendered. For the bookbag used in wikiSearch (Toms et al, 2009). Similarly we example, Search box, QBE, form fillin, faceted metadata, and do little to assist the user in digest the contents of a webpage or categories are types of Input; while Corrections, sorting, IQE + document. Ideally the tool sets need to be in situ, and readily assistance, filters, grouping; relevance feedback are types of available without user hunting through a file system or interface. Control; Text snippets, images, thumbnails, shortcuts, techniques In addition to these, one could also consider Bates’ (1990) search for visualizing relevance or content are types of Informational moves and tactics, e.g., monitoring and how each might be components. Notable about these components is that they map to supported as tools. primarily physical user interface actions, on par with the typical physical actions in office applications, e.g., cut and paste, insert, Figure 1, however, is not intended to be a complete set, and may format, track changes and so on. not be the best set.. It is instead hypothesized as an approach to aiding the user, and thereby reducing the complexity present in 4. SEARCHER’S TOOLBOX work tasks. While these tools are very much about simplifying the Missing from most approaches to date is assistance for the truly search process, they also engage with the work task more directly complex piece: work-in-head. Searching and digesting so that the tool is helping the user to work toward that work task information is a complex cognitive or mental process, and one goal. This suggests that we need a systematic approach that builds could speculate about whether searchers need “cognitive on the work of Wilson (2011) to augment those operational prostheses,” aids that assist the user with that level of intensive components with those that support the ‘real’ information work, cognitive activity (Toms, Villa & McCay-Peet, 2014). These the work that takes place in head. include assistance with creating queries, providing relevance feedback, and decision making throughout the process (White, Given the conceptual nature of these tools, they will be non-trival 2011). to develop, as each will require its own search and analysis algorithms that are tailored to the purpose of the tool. The search box itself is a coarse approach to the entire Comparatively and anologically speaking, do we need an Allan information search problem, although it lends itself to a wrench or a flathead screwdriver, a simple chopping knife or a procedural method. Other than some very simple options, e.g., mandolin slicer? auto correct, spell check, auto suggest, etc., very little assistance is The challenge we face is how to move beyond the searchbox, and [8] Russell-Rose, T. & Tate, T. (2013). Designing the Search beyond the naysayers who claim that the searchbox is the Experience. Morgan Kaufman. optimum solution. Part of this challenge is also the user [9] Toms, E.G. (2011). Task-based information searching and community who have been trained to input a few words and retrieval. In: Ruthven I and Kelly D (eds) Interactive receive some results (regardless of the level of usefulness), and Information Seeking, Behaviour and Retrieval. London: have been reluctant to use additional interface features. Thus a Facet, 2011, pp. 43–59. dramatic change may be required to influence human behaviour, a change that may be a disruptive technology. [10] Toms E.G., McCay-Peet L and Mackenzie T. (2009). wikiSearch: from Access to Use. ECDL (European Conference on Digital Libraries) 2009: 27-38. [11] Toms, E.G., Villa, R., & McCay-Peet, L. (2013). How is a REFERENCES search system used in work task completion? Journal of Information Science 39(1), 15-25. [1] Bates, M. J. (1979), Information search tactics. Journal of the [12] Vakkari, P. (1999). Task complexity, problem structure and American Society for Information Science 30, 205–214. information actions: integrating studies on informaton [2] Bystrom, K. & Jarvelin, K. Task complexity affects seeking and retrieval. Information Processing & information seeking and use. Information Processing & Management 35(6), 819-837. Management 31(2), 191-213. [13] White, R.W. (2011). Interactive Techniques. In Ruthven & [3] Feldman, S & Reynolds, H. (2010). Worldwide search and Kelly. Interactive Information Seeking, Behaviour & discovery 2009 vendor share and update on market trends. Retrieval. Facet. IDC 223926. [14] Wildemuth B, Freund L & Toms E.G. (2014). Untangling [4] Fidel, R. (2012). Human Information Intertion: An search task complexity and difficulty in the context of Ecological Approch to Information Behavior. MIT Press. interactive information retrieval studies. Journal of Documentation, 70(6), 1118-1140. [5] Hearst, M. (2009). Search User Interfaces. Cambridge. [15] Wilson, M. L. (2011). Search User Interface Design. [6] Jarvelin, K. & Repo, A.J. (1982). Knowledge work Synthesis Lectures on Information Concepts, Retrieval, and augmentation and human information seeking. Journal of Services, Morgan & Claypool. Information Science 5, 79-86. [7] Marchionini, G. (1997). Information Seeking in Electronic Environments published by Cambridge University Press Pre-­‐Search  Goal   Creating  a  Query   Over  Course  of  Search   Viewing  a  Page   Acquire  Evidence   Auto  correct   Analyse   Comprehend   Define   Auto  complete   Compare   Define   Explain   Auto  suggest   Differ   Discriminate   Find  Instructions   Define   Diverge   Explain   Gain  an  Overview   Explain  concept   Evaluate   Filter   Identify  Scope   Spell  check   Filter   Manipulate   Orient   Translate   Monitor   Personalise   Monitor   Word  selection   Organise   Relate/Connect       Personalise   Simplify       Retain,  i.e.,  book  bag   Synthesize       Sort         Stimulate         Suggest         Verify     Figure  1.  A  proposed  searcher’s  toolbox