UBC Faculty Research and Publications

VisArchive: a time and relevance based visual interface for searching, browsing, and exploring project… Hu, Keyun; Staub-French, Sheryl; Tory, Melanie; Nepal, Madhav P Mar 22, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-40327_2016_Article_36.pdf [ 2.14MB ]
JSON: 52383-1.0308650.json
JSON-LD: 52383-1.0308650-ld.json
RDF/XML (Pretty): 52383-1.0308650-rdf.xml
RDF/JSON: 52383-1.0308650-rdf.json
Turtle: 52383-1.0308650-turtle.txt
N-Triples: 52383-1.0308650-rdf-ntriples.txt
Original Record: 52383-1.0308650-source.json
Full Text

Full Text

RESEARCH Open Accessleihavoffer simple and inexpensive ways to store digital informa-tion and documents of a construction project and enableto all stakeholders or even integrated into a database man-agement system for higher-level data processing, the in-and exploringing and time-be of practicalble to rapidlyfrom the largeroject archivesHu et al. Visualization in Engineering  (2016) 4:6 DOI 10.1186/s40327-016-0036-8* Correspondence: ssf@civil.ubc.caUniversity of British Columbia, Vancouver, Canadaally archived in a shared digital storage repository. (Steed et al. 2012; Strotgen and Gertz 2012; Tory et al.2008).acity to access project information remotely from any-where. Construction documents such as meeting agendas,meeting minutes, schematic diagrams and computer-aideddesign (CAD) drawings, cost data, project schedule, designspecifications contain rich information about differentfacets of a construction project. This information is typic-complexity of its structure make searchinginformation in the project archive challengconsuming. In order for such repositories touse, construction professionals need to be aretrieve and manipulate relevant informationand diverse collection of documents within pproject team members or software tools the ease and cap- creasing amount of information and the increasingBackground: Project archives are becoming increasingly large and complex. On construction projects in particular,the increasing amount of information and the increasing complexity of its structure make searching and exploringinformation in the project archive challenging and time-consuming.Methods: This research investigates a query-driven approach that represents new forms of contextual informationto help users understand the set of documents resulting from queries of construction project archives. Specifically,this research extends query-driven interface research by representing three types of contextual information: (1) thetemporal context is represented in the form of a timeline to show when each document was created; (2) thesearch-relevance context shows exactly which of the entered keywords matched each document; and (3) the usagecontext shows which project participants have accessed or modified a file.Results: We implemented and tested these ideas within a prototype query-driven interface we call VisArchive.VisArchive employs a combination of multi-scale and multi-dimensional timelines, color-coded stacked bar charts,additional supporting visual cues and filters to support searching and exploring historical project archives. Thetimeline-based interface integrates three interactive timelines as focus + context visualizations.Conclusions: The feasibility of using these visual design principles is tested in two types of project archives:searching construction project archives of an educational building project and tracking of software defects in theMozilla Thunderbird project. These case studies demonstrate the applicability, usefulness and generality of thedesign principles implemented.Keywords: Visual analytics, Access history, Relevance-ranked search, Visualization, Construction project, SoftwaredefectsBackgroundElectronic data storage and database management systemsUnfortunately, even though this rich information can bechronicled and archived in a common repository accessibleVisArchive: a time and revisual interface for searchand exploring project arcKeyun Hu, Sheryl Staub-French*, Melanie Tory and MadhAbstract© 2016 Hu et al. Open Access This article is diLicense (http://creativecommons.org/licenses/bmedium, provided you give appropriate creditlicense, and indicate if changes were made.vance basedng, browsing,ivesPrasad Nepalstributed under the terms of the Creative Commons Attribution 4.0 Internationaly/4.0/), which permits unrestricted use, distribution, and reproduction in anyto the original author(s) and the source, provide a link to the Creative CommonsHu et al. Visualization in Engineering  (2016) 4:6 Page 2 of 16Several recent studies, however, have shown that find-ing the right documents from a project archive in atimely manner can be very difficult. A ‘project archive’, inthe context of this research is defined as a collection offiles or information being generated or recorded histor-ically throughout the project and stored in a commonshared repository. For example, on several of the con-struction projects we have studied, design and construc-tion teams used Autodesk Buzzsaw® (Buzzsaw 2014), athird-party application that is often used as a central re-pository for project information archiving, sharing andretrieval. Tory et al. (2008) found that design and con-struction professionals using these existing tools had adifficult time searching and locating project files unlessthey were already familiar with the hierarchy structureand the name of the item they were searching for. Simi-larly, Demian and Fruchter (2006a) reported that con-struction professionals often asked colleagues forinformation rather than searching project archives. Theyargue that project archives may provide insufficient con-textual information to help people understand the mean-ing of retrieved documents.Documents in construction project archives are typicallyorganized and stored in hierarchical directories similar toother types of project archives. This allows individuals toaccess files by browsing directories and searching filesusing the meta-data (keywords, date, authors, etc.).Demian and Balatsoukas (2012) argue that there are twounique issues in the design of information retrievalsystems for the construction domain: (1) Engineers andconstruction professionals are unique in terms of their in-formation needs and information-seeking behavior, and(2) construction project archives are organized differentlythan other types of document collections. Consequently,they conclude that there are unique design challenges forconstruction project archives and a need for research intothe design of query-driven systems for this domain.Research into the design of such query-driven systems hasshown the importance of revealing contextual informationabout the results of a user’s textual query rather than sim-ply a relevance-ranked list of resulting documents(Demian and Balatsoukas 2012; Demian and Fruchter2006b). Specifically, users need to see resulting items inthe hierarchical context of their ancestors, descendants,and siblings within the file hierarchy, and understand thegranularity of the item within that hierarchy (Demian andBalatsoukas 2012). Our work extends query-driven inter-face research by providing two new forms of contextualinformation to help users understand the set of docu-ments resulting from their query: 1) temporal context, inthe form of a timeline to show when each document wascreated and 2) search-relevance context, by showingexactly which of the entered keywords matched eachdocument. In addition, we provide a new approach toreveal usage context for a particular item, so a user can ex-plore which other users have accessed or modified a file.We introduce these ideas within a prototype query-driveninterface,VisArchive.VisArchive employs visualization techniques to revealthe various types of contextual information, thereby off-loading cognitive effort onto the perceptual system(Card et al. 1999). It employs a combination of multi-scale and multidimensional timelines, color-codedstacked bar chart, additional supporting visual cues andfilters to support searching and exploring historical pro-ject archives. The timeline-based interface integratesthree interactive timelines as focus + context visualiza-tions (Steed et al. 2012; Pirolli et al. 2001). It implementsand extends Dynamic Queries (Ahlberg et al. 1992) andVisual Information Seeking principles (Ahlberg andShneiderman 1994). The feasibility of using these visualdesign principles is tested in searching construction filearchives of an educational building project. We alsoapply our tool to a software defects tracking system inthe open-source software development domain to dem-onstrate its ability to search and visualize larger amountsof unstructured data.Motivation, Methods, and NeedsThis section describes the unique challenges of workingwith construction project archives and the specific userneeds for visualizing archived construction project infor-mation. The visualization design espoused in this sectionis primarily motivated by a common problem encoun-tered in the construction domain. However, the conceptscould also be applied to other domains that have non-spatial, metadata-based and time-oriented project data,such as source files of a software project.Construction projects generate voluminous data setswith significant heterogeneity of data files (structured,semi-structured and unstructured) (Russell et al. 2009)and types (Knowledge and Information Management2006; Rezgui 2001). Electronic documents for a construc-tion project are archived and stored over time in a centralrepository which we refer to as a ‘project archive’. The filescan be moved, modified and accessed by different individ-uals from diverse backgrounds and from different organi-zations. Design and construction practitioners frequentlysearch for relevant project information within the projectarchives on a regular basis as part of their intra- or inter-organizational decision making processes. As such, theability to search, browse and explore project archivesmore easily and effectively is critical for the success of aconstruction project. Specifically, construction expertsneed an effective and efficient way to find and classify in-formation (Caldas et al. 2002) and, more importantly, toexplore relevant information in the project archive(Demian and Balatsoukas 2012).Previous research (Demian and Balatsoukas 2012) andour own studies of construction projects (Tory et al.2008)) have demonstrated that information seeking andretrieval from large, shared construction project archivesis a very difficult and time-consuming process. In par-ticular, these studies found that it is difficult for projectmembers to search for and locate relevant documents,and that there is a lack of support to browse and exploresearch results in construction project archives. Practi-tioners can spend significant amounts of time searchingfor documents and in many cases, fail to find what theyare looking for. For example, in one meeting we ob-served, the mechanical consultant spent over 10 minsearching for images of the water filtration systems onhis laptop, which significantly interrupted workflow inthe meeting. In another meeting, the project team wasunable to find the sustainability goals for the project andended up spending significant amounts of time trying toidentify the goals from memory and writing them out byhand on a white board.The digital files of the construction project we studiedwere archived into different directories organized in ahierarchy and stored in a central shared repository usingAutodesk Buzzsaw® (Fig. 1). Project files could bedesignated space. Project participants could view theaudit logs to identify who and when other members ofthe team accessed particular documents. Although thesetools enabled users to track and manage file versions, itwas extremely difficult for users to group the activitiesand visually get a clear picture of how the files had beenaccessed and modified. It was also challenging to explorethe project archive and search for information that pro-ject participants were not familiar with. These challengesimpacted meeting productivity, disrupted group discus-sion, and impeded decision-making, all of which can becostly to a construction project.The challenges associated with searching, browsing,and exploring construction project archives illustratesthe important contextual information needed to helpusers understand the set of documents resulting fromtheir queries. Specifically, we identified the followingcontextual information that informed the design require-ments we considered in the development of VisArchive: Search Relevance Context: Relevance-rankedsearching of project archives and effectivevisualization of search results is critical for projectparticipants to: (i) generate search results with dif-Hu et al. Visualization in Engineering  (2016) 4:6 Page 3 of 16accessed by browsing the hierarchical directories andproject members had some flexibility in saving thedigital files in different directories or in their ownFig. 1 Snapshot of Autodesk Buzzsaw document repository (Source: www.ferent levels of relevance to the search keywords andfilter unnecessary information; (ii) provide inter-active visualizations and supporting visual cues toautodesk.com)Hu et al. Visualization in Engineering  (2016) 4:6 Page 4 of 16visualize the project archive and search results; and (iii)help users distinguish the different search results.VisArchive implements features that allow users tovisually find relevant information and prioritize the in-formation to view. Temporal Context: Practitioners need the ability tobrowse, explore and access project archives forspecific time periods or across different timehorizons. VisArchive provides usable componentssuch as visual timeline displays, multi-scale displays,and scroll bars so that users can easily explore andinteract with the project archive. Usage Context: Users need the visual capacity toview the access log of particular files as additionalinformation in order to track actions undertaken byothers and to visualize archive access history.VisArchive represents this contextual informationusing a combination of usable, visual, and interactivecomponents to better support searching, browsing andexploring project archives. While the design of thecurrent prototype is based on the requirements gatheredfrom the construction domain, we believe that most ofthese problems and requirements are common to otherdomains as well.Relevant visualization techniquesAs a visualization-based interface,VisArchive utilizes andextends various visualization design ideas. In this sec-tion, we summarize existing design contributions relatedto VisArchive’s three context-specific design features:timeline based visualizations; visual indications of searchrelevance and matched keywords; and visual representa-tions of user activity in shared repositories.Timeline-based visualizationsTimelines have been widely used in variety of applica-tions and domains to visualize and present historical andtemporal data. They provide aids in uncovering import-ant relationships of searched results or entities, cues forfiltering of information, and assistance to identify spe-cific patterns of search results (Kwon et al. 2012).They’ve been used to explore temporal metadata and re-lationships in digital libraries (Kumar et al. 1998), visu-ally browse and explore a blog archive by using a timeslider (Indratmo et al. 2008), visualize email archive con-tent (Viegas et al. 2006), visualize distributed softwaredevelopment consisting of code repositories and projectcommunication (Gilbert and Karahalios 2007), exploretemporal patterns of events within medical histories(Fails et al. 2006), visualize digital collections of webresources (Padia et al. 2012) and visualize designprocess with evolving building information data (Kimet al. 2011), among other examples. Timelines can alsobe used as an interactive filter for information indexed bytime (e.g., Wu and Tory 2009; Jones et al. 2013).More relevant to us are search and query interfacesthat provide some temporal context to the search re-sults. Previous research in information retrieval hasshown that enabling a user to see temporal context oftheir search results and to sort or filter by time can bevery useful (Alonso et al. 2009, Dumais et al. 2003).Commercial search tools (e.g., Google Scholar) alsomake use of temporal context interfaces. However, muchof this previous research has been done in a generalinternet search context and has focused on extractinguseful temporal information from the documents to sup-port temporal clustering and queries (e.g., Alonso et al.2009, Dumais et al. 2003, Hoffart et al. 2011, Jones andDiaz 2007). Temporal feedback interfaces in these sys-tems are typically limited to a simple facet where a usercan filter a textual results list by entering or selecting atime range, plus the ability to sort results by time. Jonesand Diaz (2007) do present a timeline visualization tocomplement the textual results list, but at a very abstractlevel showing only key temporal clusters, not actual doc-uments. We suspect that temporal context may be evenmore important for information seeking in project arch-ive interfaces because participants in a project may beable to recall the approximate date when an item wascreated or used.Visual indication of search relevance and matchedkeywordsVarious techniques have been developed to reveal therelevance of retrieved documents to search keywords.Veerasamy and Heikes (1997)) designed a visualizationthat displayed the relevant documents and assisted usersto effectively reformulate queries based on the searchedkeywords in the first stage. Foo and Hendry (2007) cre-ated and evaluated a suite of visualizations for searchingone’s desktop. Relevant results to the search keywordsand filters were categorized by using different colors,shapes, etc. so that users could effectively identify anddistinguish the results relevant to different searched key-words and filters. However, these visualizations were notspace efficient. Neither of these tools integrated tem-poral context into the search process.Cambiera (Isenberg and Fisher 2009), a tabletop visualanalytics tool, supports information foraging activities inlarge text document collections. It allowed users to visu-ally connect different groups of data or activities and up-dates in different visualizations for a visual analyticstask, particularly using color-coded search keywords.However, Cambiera focused on providing awareness ofusers’ analytical activities to others in a collaborativesearch task. VisGets (Dork et al. 2008) used color-codedweighted brushing to indicate search results withHu et al. Visualization in Engineering  (2016) 4:6 Page 5 of 16different relevance mapped to the keywords. However,the visualization did not visually distinguish results withthe same relevance ranking but different matched key-words. Jones et al. (2013) suggested a process for creat-ing data visualizations in collaborative engineeringprojects by constructing a text visualization task tax-onomy and creating visual mappings of the text data.However, the visualization design process does not in-clude interactive searching, browsing and retrieval ofinformation.Visual representation of user activity in sharedrepositoriesVisualization of time-based human activities has a long his-tory. For example, Lifelines (Plaisant et al. 1998) visualizedmedical records of a patient such as past symptoms, diag-noses and medications through an interactive timeline-based interface. Within the context of project repositories,the vast majority of this work has focused on software de-velopment repositories, with greater emphasis on changesto files and source code rather than activities of the devel-opers (Storey et al. 2005). For example, Augur (Froehlichand Dourish 2004) visualized software artifacts and devel-opment activities with color-coded indications over thesource code. Those projects that have focused on activitiesof developers have tended to support understanding theoverall project evolution (e.g., Ogawa and Ma 2008), ratherthan information seeking tasks such as finding all filesmatching given criteria.The Timespace (Krishnan and Jones 2005)visualization system provided overviews of user activityon multiple projects and detailed views of user activitywithin a selected project, allowing users to explore theactivities on the projects. However, the tool focused onpersonal activities and did not support exploration ofgroup activities (e.g., who has modified a file on a spe-cific date). PragmatiX (Walk et al. 2013) provided avisualization of collaborative change logs, to help man-agers monitor progress, tracking and exploring quality-related issues such as overrides and coordination amongcontributors. It focused on change log analysis and ex-ploration. There is a limited previous work that focusesspecifically on visualizing file access logs, which can as-sist users in searching and exploring temporal sharedproject archives.Results: System Design and ImplementationThis section describes the development of, VisArchive, avisualization tool that implements a combination of stand-ard visualizations and interaction techniques to solve thespecific problem of searching project archives. We firstprovide a detailed description of the visualizations imple-mented, and then describe the relevance algorithm andthe implementation details.Overview of the visualization toolVisArchive consists of the following visual and interactivecomponents: search bar (Fig. 2(a)), information browser(Fig. 2 (b) and (c)), interactive Timelines (Fig. 2 (d)), ad-vanced filters and access history viewer (discussed in sec-tion “Evaluation”). The search bar allows users to inputmultiple keywords as well as the option for advanced filters.The information browser (Fig. 2 (b)) including descriptionviewer (Fig. 2 (c)) allows users to browse the items withinan archive and to view the meta-information and descrip-tion of a selected item in detail. Two interactive timelinesat the bottom of the interface (Fig. 2 (d)) visualize informa-tion of the archive including one full-range timeline for theoverall project archive and one scalable timeline for viewinga detailed portion of the file archive within a selected timeinterval.The information shown in the timelines and informa-tion browser will be updated simultaneously while usersare performing different search tasks and/or moving thetime slider to view the archive in a different time range(Fig. 2(f )). By performing a search task, search resultswill be generated behind the scenes by relevance-ranking algorithm (described in “Relevance algorithm”)and the relevance information related to the search key-words will be visualized in the timelines and informationbrowser with additional visual representations to helpusers identify the most relevant search results and ex-plore other related information in the file archive.Interactive timelines and visualization of the search resultsThe items in the archive, by default, are arranged in thetimeline based on creation time. The time range selectorprovides the visualization of the project archive within acustomizable time range or interval (Fig. 2(g)). Userssearch the file archives based on keyword/s. VisArchiveimplements the concept of dynamic queries (Ahlberget al. 1992), which allows users to formulate search quer-ies dynamically and get feedback immediately throughmanipulation of the time slider and information browser.The search results are assigned with different levels ofrelevance to the search query based on the relevance-ranking algorithm.The levels of relevance for search results are repre-sented by a color scale. Grey color represents the archiveitems with zero level of relevance (i.e. none of the searchkeywords match the meta-data of the archive items).The continuum of lighter yellow to dark red indicatesthe increasing level of relevance. This color-coding is ap-plied to the stacked bar charts in the timelines as well asin the information browser.Blue arrows shown at the bottom of the bar charts inthe timelines indicate the most relevant files created onparticular dates that match all the search keywords andare considered to be one of the most relevant searchHu et al. Visualization in Engineering  (2016) 4:6 Page 6 of 16results. Users are thus able to identify the most relevantsearch results and their creation dates from the timelineusing the visual cues of the blue arrows.Information browser, visual cues and advanced filterssupporting the search resultsInformation items with relevance-ranked visual informa-tion are updated and displayed synchronously in the in-formation browser as users adjust the time range in thetimeline. The information browser lists the informationitems vertically and shows all the information itemswithin the same time range that is selected in the time-lines (Fig. 2 (c)). Users are able to scroll, browse and se-lect the file of their interest and identify its meta-information including the access history information ofthe selected file.For consistency, the color-coding used for visualizingthe relevance-ranked search results in the timelines isused in the information browser to potentially help usersidentify the relevant items more easily and effectively.Moreover, the rectangles representing information itemsFig. 2 Overview of VisArchive’s main interface: (a) Search bar; (b) Informatioslider; (f) Time range; (g) Time range selector[b][a][d][g][c]with scaled-colors allow users to explore other relevantfile items in the archive with different relevance levelsmatching the search keywords.VisArchive allows users to distinguish search resultsmatching different search keywords. We applied techniquessimilar to visual brushing and linking (Buja et al. 1991) toestablish relationships and to distinguish between eachgroup of data and provide focused + contextual informationwith multi-scale timeline views. In VisArchive, searchedkeywords are colored with randomly assigned distinctcolors and linked to each of the search results in the infor-mation browser when users perform a search task (Fig. 2(b)). The supporting visual cue (color-coded panes for eachitem) allows users to distinguish between search results andexplore their relevance details in the archive.The use of filters (Fig. 3) helps users narrow down thesearch results to be visualized and displayed based on filecontents and properties such as by file types (Fig. 3 (a),created users (Fig. 3 (b), and keyword exception (Fig. 3(c)).In general, custom filters should be developed for eachdomain in order to conform to the information in the[e][f]n browser; (c) Description viewer; (d) Interactive timelines; (e) Timeproject archive and the searching preferences of users.le tHu et al. Visualization in Engineering  (2016) 4:6 Page 7 of 16For example, software developers might be interested insearching software defects for specific software compo-nents, release versions, etc.[b][a]Fig. 3 Advanced filters designed for the construction file archives: (a) FiAccess history visualization viewerUsers can view access history information of a selected fileitem using “View Access Info” tab. A pop-up window withsimilar visualization representation and interaction to thatof VisArchive’s main screen consists of a timelinevisualization (Fig. 4(a)) to visualize summary informationabout the access history, an access history browser (Fig. 4(b))to display the details of access history, and a user filter (Fig. 4(c)) to filter the access history by access user name. The ac-cess history visualization viewer uses distinct colors to in-dicate or distinguish visually the different types of access,so that users can recognize when and how a file wasaccessed.Relevance algorithmThe relevance algorithm generates the relevance-rankedsearch results, which are represented visually on theinteractive timelines and information browser by apply-ing visualization representations and supporting visualcues. Figure 5 schematically shows the process. It inte-grates relevance-ranked search results with a visual rep-resentation to enable users to visually search andexplore the archives more easily and intuitively. This al-gorithm could be easily replaced by any other rankingalgorithm if different relevance criteria were desired.To generate the relevance-ranked search results, thealgorithm calculates a relevance ranking based on thesearch terms and assigns the ranking to each informa-tion item in the project archive. The prototype first ex-tracts the meta-information of each item from theproject archive database (e.g. the meta-data of the files[c]ypes; (b) User who created the file; (c) Keyword exceptionin the construction project archive contains filename, filepath, file keywords and description). The algorithm thenmatches this extracted information with the search key-words input by the user to compute the relevance levelsfor each item. At the end, the prototype prepares thesearch results with the assigned level of relevance fordata visualization that is presented to the users. Higherrelevance level will be assigned if the meta-data of theitem matches more search keywords. The level of rele-vance will be increased by one if any one of the searchkeywords is found in the meta-data of the item regard-less of the number of the times that keyword appeared.Level 0 will be assigned if none of the search keywordsis matched. For example, we assign the level 0 of relevanceto the items if none of the searched keywords was foundin the extracted meta-data of the item. We assign level 5to the item if five of the search keywords were matched.Therefore, every time users input keywords to perform asearch task, all the items in the archive will be assignedlevels of relevance from zero up to the number of searchkeywords. The relevance-ranked search results are proc-essed with visualization techniques and visually repre-sented to users in the user interface. While the currentimplementation of VisArchive can support archives withthousands of files without system performance issues, forVisArchive was implemented as a desktop applicationviHu et al. Visualization in Engineering  (2016) 4:6 Page 8 of 16using Java and the JFreeChart (2013) visualization tool-kit. Most of the charting and visualization used in thisvery large archives, some adaptations to the relevancy al-gorithm and the user interface may be needed.Implementation detailsFig. 4 Visualization showing access history of a selected file: (a) Timelineprototype were generated by using the JFreeChart APIwith modifications and customizations. The prototyperequires a database to store the project archive as infor-mation records, file access history and/or a central filerepository to store the electronic files of the archive ifFig. 5 Diagram of generating relevance-ranked orientedsearch resultsdigital files are part of the project archive. In order tomake the archive content searchable, the extraction oftextual information as meta-data for keyword-basedsearching from the electronic files is needed. For easeand efficiency of generating a dataset to demonstrate theconcept used in the interface, this information was ex-tracted and created manually from the existing archives.From the construction project archive, electronic files[b] [c][a]sualization; (b) Access history viewer; (c) User filterwere indexed by extracting all necessary meta-information regarding each document and integratingthis information into the database for demonstrationpurposes. The meta-data that were extracted from thefiles consisted of file name, file description, date of cre-ation, related keywords and file path. File access historydata was stored separately in a different table from filemeta-data in the database.VisArchive is a front-end desktop client that communi-cates with the database and file repository and generatesthe search results to support the archive search and datavisualization. The data to be visualized and used by theprototype are stored as entries in a database. Since theconstruction file archives were stored as electronic filesin a central file repository, a file parser could be devel-oped in the future to extract the meta-data from the fileand parse this information into the database automatic-ally. The repository management system may allow usersto tag related keywords as meta-data to a file manuallywhen they create or modify the files.Evaluation and DiscussionIn order to examine the feasibility of using VisArchivefor searching, browsing and exploring information inthe details (e.g. file description or file path) whenpear at the top of the list, the PM would not knowHu et al. Visualization in Engineering  (2016) 4:6 Page 9 of 16they click on a file from the information browser.The information browser for the software projectcase study was modified to display the summary foreach software defect. Users can identify a softwaredefect item by viewing its summary.Case study 1: construction project archivesIn this case study, VisArchive was used to search con-struction project archives of an educational building pro-ject. The project archive contained more than 800 filesthat were created during its design development phaseby different individuals involved in the project. We cre-ated a project archive using 300 files that were selectedbased on the information available for testing the proto-type. These files were stored and shared as digital copiesin a central hosting server with a variety of file typessuch as PDF, DOC and TXT. The information, such asID, name, path, and description of each file, was ar-chived as searchable meta-data into a database systemfor archive searching, data processing and visualizationby VisArchive.This case study provides access history visualizationcomponents for visualizing file access history in thearchive. Since prior file access history data was not pro-vided by the construction project archive, a synthetic fileaccess history for a number of files in the database wascreated for demonstration purposes. The testing focusedon searching and exploring the files in the archive thatproject archives of different domains, two case studieswere conducted. The case studies involved (1) a con-struction project archive, and (2) a software projectarchive (software defect tracking). Specifically, the designprinciples implemented in VisArchive were examined inrelation to the three context-specific design features: (1)the temporal context shows timeline based visualiza-tions; (2) the search-relevance context shows visual indi-cations of search relevance and matched keywords; and(3) the usage context shows visual representations ofuser activity in shared repositories. The evaluation fo-cused on the feasibility of the prototype to resolve com-plex use scenarios, rather than simple search usingknown file names or IDs.The interface of VisArchive was revised and modi-fied based on these case studies, but the core fea-tures of the tools described in Section “Systemdesign details, reasoning structure and implementa-tion” remained stable. Since a file contains more in-formation than users often need, the interface forthe construction project case study was designed toinclude a description viewer that allows users to viewusers might have never accessed before, or that userslacked specific information about the files.clearly which keywords and how many of them werematched. Accordingly, the PM might need to open eachfile to evaluate how relevant it is to the search keywords.It would also have been difficult for the PM to under-stand how these files had been produced along the way.This is important because the PM needs to find files thatwere produced in a certain period in the project’shistory.For the scenario described above, VisArchive enablesthe PM to easily identify the files matching all the searchkeywords (“electrical,” “mechanical,” “structural,” and“Mike”) on ten different dates. These files are consideredthe most relevant to the PM’s search and contain all thesearch keywords entered. Thus,VisArchive helps the PMto view the most relevant files first — more quickly thanotherwise possible when searching manually. The effi-ciency of finding the most relevant information is verycritical particularly in a large-scale construction projectarchive. The blue arrows in the timelines not only indi-cate the relevant files, if any, in the archive — matchingall the search keywords — but also provide users a visualoverview of when these files were created during thespecific project stage (Fig. 6). In order to help the PM toretrieve the most relevant file out of the search results,the information about each file in the informationbrowser is very useful as it allows the PM to view andaccess detailed information about the files (Fig. 7).Exploring files relevant to the search keywordsIn VisArchive, the color-coded stacked bar charts in thetimelines and visual support in the information browserare designed to enable users to explore files with differ-ent levels of relevance to the search keywords. For ex-ample, the PM in the above-mentioned scenario, besidessearching for the files that match all the search keywords(“electrical,” “mechanical,” “structural,” and “Mike”), wasfurther interested in exploring other files that matchedto one or more of these keywords. For example, the PMneeded information about other “electrical” related filesSearching files that match all the search keywordsThe testing considered the real scenario in which a pro-ject manager (PM) wanted to find all “electrical”, “mech-anical” and “structural” documents that an engineer(named hereafter as “Mike”) had worked on. The PMhad to share them with another engineer that came on-board after Mike left the company. Traditional searchmethods present search results as a list of files from topto bottom, with information such as file name, size, last-modified date, etc. Although existing search solutionssuch as Buzzsaw® enable the most relevant files to ap-which Mike was also involved in. Existing search solu-tions for construction project archives make it difficultFig. 6 Timeline visualization while searching keywords: electrical, mechanical, structural, and MikeHu et al. Visualization in Engineering  (2016) 4:6 Page 10 of 16for individuals to explore the files by their relevance tothe search keywords as they generate a long list of files.Since there is no visualization of the search results, indi-viduals must view the textual meta-information of a fileto identify its creation date and matched keywords.While the file list can be arranged by either “Date” or“Relevance”, individuals cannot easily explore andbrowse the relevant files in the project timeline andquery for information such as the following: Are thereany files that match a certain number of the search key-words? When were these files created in the projecttimeline? Which month contains more relevant filesthan the others? The color-scaled visual support both inthe timeline and the information browser of VisArchive,allows the PM to identify these relevance-ranked filesand to identify the level of relevance for each file in theinformation browser. For example, two files (Fig. 7 (a)Fig. 7 Information browser displaying the files with color-coded visual supand (b)) are shown as less relevant than the most rele-vant file (Fig. 7 (c)), but they are highlighted as morerelevant than the other files found in the search results.The associated color-coded visual panes for searchkeywords in the information browser allow users to dis-tinguish the files with same relevance but differentmatched keywords. For example, when the PM wantedto explore electrical documents with which Mike was in-volved (files containing “electrical”, and “Mike”), otherfiles relevant to other search keywords are also shownwith the same relevance as shown in Fig. 7, (e.g. filescontaining “structural” and “Mike”). With existing solu-tions, users would need to read extra meta-informationof each file in order to differentiate between files withthe same relevance level.Besides searching and exploring files in the projectarchive, the PM wanted to explore other information,[a][c][b]ports: (c) most relevant file; ((a) and (b)) second most relevant filesfor example, identifying the time periods in which theproject archive was more active (i.e. when more fileswere created). The color-coded stacked bar charts onthe timelines show the density of file creation, andthe density of files relevant to the search keywordsthroughout the life of the project. Because the time-lines convey information about the various activitiesand file types created during the project (e.g. docu-ments such as different layout plans may have beencreated most frequently earlier on in the project),VisArchive can help to narrow down the time periodsand the intensity of project activities in order to findthe most relevant documents.Exploring file access historyThe access history information of files in the con-structiion archive is extremely important for designcoordination and development. The architect on theproject had placed the latest and updated version ofCase study 2: defects tracking of mozilla thunderbird projectThe second case study expands the application ofVisArchive beyond the construction project archive toexplore software defects tracking in the open-sourcesoftware development domain. Unlike the constructionproject archives, software defects in this case study werenot structured into directories as digital files. Comparedto the previous case study, this case study project pro-vides a test environment for searching and visualizinglarger amounts of unstructured data.The Mozilla project was started in 1998 and wasintended to develop open-source software projects usingthe power of thousands of programmers all over theworld (Mozilla project 2014). Thunderbird is the MozillaFoundation’s next-generation email client. As the soft-ware is being used all over the world by thousands ofusers, software defects and issues can be found and re-ported by using a web-based defect tracking tool calledBugzilla (Bugzilla 2014) that also allows developers to trackHu et al. Visualization in Engineering  (2016) 4:6 Page 11 of 16the architectural design into a shared constructionarchive but was not sure whether the consultantshad accessed it. The color-coded visualization of ac-cess history provides much of this information, suchas the number of times the file was accessed and the typeof access (e.g., opening or modifying a file). In the accesshistory viewer (Fig. 8), each type of access isassigned a color-code to improve the users’ ability toidentify the file they are looking for (e.g., the onethey accessed the day before or the one that wasmodified most recently). The access records can befiltered by individuals who have created, accessed,or/and modified the file.Fig. 8 Visualization of access history filtered by selected access usersthese issues and, eventually, fix them. The Thunderbirdproject archived more than 5000 software defects in Bug-zilla from the beginning of the project in 2004. Around1000 defect records from Bugzilla under the Thunderbirdproject between 2006 and 2007 were used for testing. Amodified and simplified version of the VisArchive interfacewas used compared to the construction case study. Foreach software defect, the defect ID, date, and summarywere used as meta-data within VisArchive.Bugzilla allows users to search the defect archives byentering keywords and using advanced filters that aresimilar to the search mechanism of VisArchive. Findingrelevant information over thousands of defect records inBugzilla is a tedious process. Moreover, the search re-sults are represented in a conventional list of defect in-formation (Fig. 9). Although users can reorder searchresults alphabetically by attributes, it is difficult for usersto view the relationships among different defects, andespecially, to explore defects that are partially relevant tothe search keywords.The defect summary was described by the defectfinder, and contained crucial information needed by asoftware developer or tester to recreate the defect. Thiskind of meta-data is hard to categorize and filter usingthe original Bugzilla interface. Therefore, emphasis inthis case study was on finding relevant software defectsby searching and exploring through the summary ofsoftware defects.Searching defects in the software defect archivesarchive, the blue arrows always provide awareness of themost relevant results and accordingly offer visual cues tothe users.Exploring the defect archive and relevant software defectsIf none of the defects matches the user’s requirementsor users want to explore other software defects with lessrelevance, users may refer to the stacked bar charts inthe timeline and visual panes to identify the most rele-vant defects in the archive and where these defects occuron the archive timelines.Figure 11 shows the example in which a software de-veloper searches the software defect archive with morekeywords than the example in 4.2.1 (e.g. searching “com-pose,” “window,” “file” and “attachment”). The developercan easily identify the results (Fig. 11 (a)) matching allthe search keywords (e.g. defects might be about “file at-Hu et al. Visualization in Engineering  (2016) 4:6 Page 12 of 16In order to fix issues and improve the quality of soft-ware, developers need to search and find the softwaredefects from the archives that correspond to their ex-pertise or responsibility. In addition, time information,such as when the issues were filed, is also useful for de-velopers to prioritize them and fix. Figure 10 shows anexample of search results and visualization support thatis provided by VisArchive for the software defect project.The software developer can take advantage of the time-line visualization of search results to easily identify soft-ware defects along with the date that these defects werecreated. With VisArchive, the developer can navigate tothe time range containing the earliest and most relevantdefects found (Fig. 10(a)) in the timeline and view thedefect summary of the most relevant defect (Fig. 10(b))in the information browser. Regardless of the size of theFig. 9 Conventional list of search results provided by Bugzillatachment” in “compose window”) by finding the bluearrow in the timeline and the blue highlighted tickets inthe defect browser. The developer may also want to ex-plore other defects that are partially relevant to the searchkeywords. For example, the developer may be interestedin other defects relevant to “compose window” or “file at-tachment” (e.g. Email “compose window” might haveother issues besides in the “file attachment” function, andthe developer may want to fix those as well). Other key-word combinations (e.g. “compose attachment”, “windowfile”) are not the terms that the developer is concernedwith in this case, and thus these software defects can beignored. By glancing at the stacked bar chart over thetimeline, the developer can easily perceive how the defectsin the archive are relevant to the search keywords andhow these defects distribute over the timeline in the[a][b]omtheHu et al. Visualization in Engineering  (2016) 4:6 Page 13 of 16archive. The developer can browse and explore the defectsthat partially match keywords by visually scanning theFig. 10 Visualization of search results provided for searching “message cindicated in the zoomed timeline; (b) Most relevant defect indicated incolor panes of keywords in the defect browser, instead ofreading the summary of defects (e.g. the defects contain-ing “compose window” are interesting items (Figs. 11 (b)[a][b]Fig. 11 Visualization of search results provided for searching “compose winrelevant defect; (b) Defects matching “compose window”and 12(a)), whereas the defects containing “window file”may be disregarded (Fig. 12(b)).pose window” in the software defect archive: (a) Most relevant defectinformation browserSimilar to the construction case study, the timelines ofthe defect archive show a picture that conveys to the de-veloper how many defects and how the defects in thedow file attachment” in the software defect archive: (a) the mostfecHu et al. Visualization in Engineering  (2016) 4:6 Page 14 of 16archive match the search keywords. The developer isable to determine — visually — the dates that containthe defects that are more or less relevant to the searchkeywords. When used for logging new software issues,VisArchive enables software testers to search and explorewhether there are similar or related issues existing in thearchive. If relevant defects matching the same keywordsare found in the archive, timelines enable users to deter-mine and explore their relationship over time. As an ex-ample, a tester searches for a system bug in which a“Removed account stays in ‘recent’ folder view,” evenafter it has been removed. The search keywords for thisbug include “folder” and “preferences,” and the resultsshow the bug was logged in August. The timelinevisualization reveals that another defect, labelled“Unmarking folder as favourite in ‘Favourite Folders’view doesn’t remove folder,” is also associated with thekeywords of “favourites,” “folder” and “preferences.” Thetester sees the bug was logged in May and that it hassince been resolved. This information could be helpful[b][a]Fig. 12 Sample information displayed in information browser: (a) the deto the developer to explore whether the resolution tothe earlier bug logged in May could help to resolve thesimilar bug detected in August.ConclusionsThis research extends query-driven interface research byproviding three context-specific design features: timelinebased visualizations; visual indications of search rele-vance and matched keywords; and visual representationsof user activity in shared repositories. These design fea-tures were implemented in an interactive visualizationtool called VisArchive that integrates multiple commonlyused visualization and user interaction techniques to fa-cilitate searching, browsing, and exploring informationin historical project archives. VisArchive visualizesrelevance-ranked search results with color-coded stackedbar charts in project timelines and uses additional sup-porting visual cues to distinguish search results based onsearch keywords. Two case studies from the construc-tion and software project domains were used to demon-strate its applicability, usefulness and generality.Currently, VisArchive allows users to view the accesshistory of a single item (e.g. a file or a defect record) inthe project archive. It might be useful for users to alsoexplore the access history of multiple items in the ar-chives. Different visual representations might be used toachieve this goal in future research (e.g. multiple time-lines could be used to represent the access history of dif-ferent files in one display, or they could be aggregatedinto one timeline and use color-scale to distinguish dif-ferent items). The interactive timeline visualization couldalso be more flexible to allow for different time frames.The current implementation represents each bar of thetimeline in terms of the number of items created in1 day. However, if the project archive covers a longperiod of time (e.g. the lower timeline of the softwaredefect archive, visualizing around 1000 records over2 years), the bar chart will be compacted, and the usert matching “compose window” (b) the defect matching “window file”will have difficulty clearly seeing the color-codedvisualization in the lower timeline. Furthermore, therelevance-ranking algorithm for generating search re-sults is based on keywords, which may limit the qualityand reliability of the search results in other domains.The current prototype used manually extracted meta-data for demonstration purposes. However, key designideas of VisArchive should be able to seamlessly inte-grate with existing archive management systems thatprovide well-designed content management and textsearch capabilities. For example, the defect tracking sys-tem Bugzilla allows users to create and edit defects withsearchable content and meta-data such as related key-words. Since all the content and meta-data have beenstored in database when the defect was created, they be-came searchable by the system. VisArchive could be em-bedded to the system in place of the conventional textbased search results to provide better visualization andHu et al. Visualization in Engineering  (2016) 4:6 Page 15 of 16interaction capabilities. However, in a file or documentbased archive such as the construction archive, searchingcan be more challenging. To overcome this, searchabledata such as textual content or meta-data needs to beextracted from the file or inserted by users when creat-ing them (this is especially important if the file is nottext based, such as image files). Well-designed archivemanagement systems should be able to extract the textcontent automatically from the text-based files and makethe keywords searchable in the system. The quality ofthe search results will depend on the searchable meta-data and the user’s analytical ability.While the visual cues for both the historical contextand the matching-level searches should make search effi-cient for users, the cues will only be successful if theusers notice and understand them. To this end, more re-search is needed to confirm whether users can indeedintuitively understand the meaning of visual indicatorsimplemented in VisArchive. In addition, user studies arealso needed to examine the effectiveness, usability anduser experience of VisArchive in various domains. Webelieve that the design principles implemented in VisArc-hive can be applied to different domains as long as the in-formation items in the archive have temporal information,such as creation date, and the relevant meta-data. VisArc-hive simply requires the access information (e.g. creationdate, access date, and modification date) and searchablemeta-data (e.g. summary, description, keywords, tags) tofunction. For example, research papers in the IEEE on-line archives (IEEE 2014) are associated with a publi-cation date and meta-data for users to search. Insteadof presenting the search results in a conventional listof papers, VisArchive could visualize a timeline-basedoverview of the published papers. Advanced filtersand an information browser could be customized andmodified based on users’ needs in these differentdomains.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsIn preparing this manuscript, all the listed authors have made significantcontribution. KH carried out the gap analysis, literature review, researchdesign, implementation, and evaluation studies. MT and SS-F conceived theresearch idea, defined the problem and assisted in the design,implementation and evaluation of the research. MN extended the literaturereview and gap analysis, and helped to refine the concepts and draft themanuscript. All authors have read and approved the final manuscript.Received: 8 August 2015 Accepted: 22 February 2016ReferencesAhlberg, C., & Shneiderman, B. (1994). Visual information seeking: tight coupling ofdynamic query filters with starfield displays (pp. 313–317). Boston: Pro. of theSIGCHI Conference on Human Factors in Computing Systems.Ahlberg, C., Williamson, C., & Shneiderman, B. (1992). Dynamic queries forinformation exploration: an implementation and evaluation (pp. 619–626).Monterey: CHI’92 Proc. of the SIGCHI Conference on Human Factors inComputing Systems.Alonso, O., Gertz, M., & Baeza-Yates, R. (2009). Proc. CIKM’09, November 2–6, 2009(pp. 97–106).Bugzilla. (2014). https://bugzilla.mozilla.org/. [Accessed 6 Mar 2016].Buja, A., McDonald, J. A., Michalak, J., & Stuetzle, W. (1991). Interactive datavisualization using focusing and linking (pp. 156–163). San Diego: IEEEConference On Visualization ′91.Buzzsaw. (2014). http://www.autodesk.com/products/buzzsaw/overview.[Accessed 6 Mar 2016].Caldas, C. H., Soibelman, L., & Han, J. (2002). Automated classification ofconstruction project documents. Journal of Computing in Civil Engineering,16(4), 234–243.Card, S. K., Mackinlay, J. D., & Shneiderman, B. (1999). Readings in informationvisualization: using vision to think. San Francisco: Morgan KaufmannPublishers.Demian, P., & Balatsoukas, P. (2012). Information retrieval from civil engineeringrepositories: importance of context and granularity. Journal of Computing inCivil Engineering, 26(6), 727–740.Demian, P., & Fruchter, R. (2006a). An ethnographic study of design knowledgereuse in the architecture, engineering and construction industry. Research inEngineering Design, 16, 184–195.Demian, P., & Fruchter, R. (2006b). Methodology for usability evaluation ofcorporate memory design reuse systems. Journal of Computing in CivilEngineering, 20(6), 377–389.Dork, M., Carpendale, S., Collins, C., & Williamson, C. (2008). VisGets:coordinated visualizations for web-based information exploration anddiscovery. Visualization and Computer Graphics, IEEE Transactions on,14(6), 1205–1212.Dumais, S., Cutrell, E., Cadiz, J. J., Jancke, G., Sarin, R., & Robbins, D. C. (2003). Proc.SIGIR’03 (pp. 72–79).Fails, J. A., Karlson, A., Shahamat, L., & Shneiderman, B. (2006). A visual interface formultivariate temporal data: finding patterns of events across multiple histories(pp. 167–174). Baltimore: IEEE Symposium on Visual Analytics Science andTechnology.Foo, S., & Hendry, D. (2007). Desktop search engine visualisation and evaluation(10th International Conference on Asian Digital Libraries, pp. 372–382). Hanoi:Springer Verlag.Froehlich, J., & Dourish, P. (2004). Unifying artifacts and activities in a visual tool fordistributed software development teams (pp. 387–396). Edinburgh: Proc. of the26th International Conference on Software Engineering.Gilbert, E., & Karahalios, K. (2007). CodeSaw: A social visualization of distributedsoftware development (11th IFIP TC 13 International Conference on Human-Computer Interaction, INTERACT, pp. 303–316). Rio de Janeiro: SpringerVerlag.Hoffart, J., Suchanek, F. M., Berberich, K., Lewis-Kelham, E., de Melo, G., & Weikum,G. (2011). Proc. WWW 2011 (pp. 229–232).IEEE. (2014). http://www.ieee.org/index.html. [Accessed 6 Mar 2016].Indratmo, Vassileva, J. Gutwin, C. (2008). Exploring blog archives with interactivevisualization. Naples: Proc. of the Working Conference on Advanced VisualInterfaces.Isenberg, P., & Fisher, D. (2009). Collaborative brushing and linking for Co-locatedvisual analytics of document collections. Computer Graphics Forum, 28(3),1031–1038.JFreeChart (2013). http://www.jfree.org/jfreechart/. [Accessed 6 Mar 2016].Jones, R., & Diaz, F. (2007). Temporal profiles of queries. ACM Trans. InformationSystems, 25(3), 31.Jones, S., Payne, S. J., Hicks, B. J., & Watts, L. (2013). Visualization of heterogenoustext data in collaborative engineering projects. Atlanta: The 3rd IEEE Workshopon Interactive Visual Text Analytics.Kim, S.-A., Choe, Y., Jang, M., & Seol, W. (2011). Design process visualization systemintergrating BIM data and performance-oriented design information. Seoul:28th International Symposium on Automation and Robotics in Construction.Knowledge and Information Management (KIM). (2006). KIM grand challengeproject. http://www.ukoln.ac.uk/projects/grand-challenge/. [Accessed 6 Mar2016].Krishnan, A., & Jones, S. (2005). TimeSpace: activity-based temporal visualisation ofpersonal information spaces. Personal and Ubiquitous Computing, 9(1), 46–65.Kumar, V., Furuta, R., & Allen, R. B. (1998). Metadata visualization for digital libraries:interactive timeline editing and review (pp. 126–133). Pittsburgh: Proceedingsof the 1998 3rd ACM Conference on Digital Libraries.Kwon, B. C., Javed, W., Ghani, S., Elmqvist, N., & Ji, S. Y. (2012). Evaluating the roleof time in investigative analysis of document collections. IEEE Transactions onVisualization and Computer Graphics, 18(11), 1992–2004.Mozilla Project. (2014). https://bugzilla.mozilla.org. [Accessed 6 Mar 2016].Ogawa, M., & Ma, K.-L. (2008). StarGate: A unified, interactive visualization ofsoftware projects (pp. 191–198). Kyoto: IEEE Pacific Visualization Symposium.Padia, K., Alnoamany, Y., & Weigle, M. C. (2012). Visualizing digital collections atarchive-it (pp. 15–18). Washington: 12th ACM/IEEE-CS Joint Conference onDigital Libraries.Pirolli, P., Card, S. K., & Van Der Wege, M. (2001). Visual information foraging in afocus + context visualization (pp. 506–513). Seattle: Proc. of the SIGCHIConference on Human Factors in Computing Systems.Plaisant, C., Mushlin, R., Snyder, A., Li, J., Heller, D., & Shneiderman, B. (1998).LifeLines: using visualization to enhance navigation and analysis of patientrecords (Proceedings of the AMIA Symposium, pp. 76–80).Rezgui, Y. (2001). Review of information and the state of the art of knowledgemanagement practices in the construction industry. Knowledge EngineeringReview, 16(3), 241–254.Russell, A. D., Chiu, C.-Y., & Korde, T. (2009). Visual representation of constructionSubmit your manuscript to a journal and benefi t from:7 Convenient online submission7 Rigorous peer review7 Immediate publication on acceptance7 Open access: articles freely available online7 High visibility within the fi eld7 Retaining the copyright to your articleHu et al. Visualization in Engineering  (2016) 4:6 Page 16 of 16management data. Automation in Construction, 18(8), 1045–1062.Steed, C. A., Symons, C. T., DeNap, F. A., & Potok, T. E. (2012). Guided text analysisusing adaptive visual analytics. Burlingame: Proc. SPIE 8294, Visualization andData Analysis.Storey, M. D., Čubranić, D., & German, D. M. (2005). On the use of visualization tosupport awareness of human activities in software development: A survey and aframework (Proc. of the 2005 ACM Symposium on Software Visualization, pp.193–202).Strotgen, J., & Gertz, M. (2012). Event-centric search and exploration in documentcollections (pp. 223–232). Washington: Proc. of the 12th ACM/IEEE-CS JointConference on Digital Libraries.Tory, M., Staub-French, S., Po, B. A., & Wu, F. (2008). Physical and digital artifact-mediated coordination in building design. Computer Supported CooperativeWork, 17(4), 311–351.Veerasamy, A., & Heikes, R. (1997). Effectiveness of a graphical display of retrievalresults (pp. 236–245). Philadelphia: Proc. of the 20th Annual InternationalACM SIGIR Conference on Research and Development in InformationRetrieval.Viegas, F. B., Golder, S., & Donath, J. (2006). Visualizing email content: portrayingrelationships from conversational histories (pp. 979–988). Montreal: Proc. of theSIGCHI Conference on Human Factors in Computing Systems.Walk, S., Pöschko, J., Strohmaier, M., Andrews, K., Tudorache, T., Noy, N. F., et al.(2013). PragmatiX: an interactive tool for visualizing the creation processbehind collaboratively engineered ontologies. International Journal onSemantic Web and Information Systems, 9(1), 45–78.Wu, F., & Tory, M. (2009). PhotoScope: visualizing spatiotemporal coverage ofphotos for construction management (pp. 1103–1112). Boston: 27thInternational Conference Extended Abstracts on Human Factors inComputing Systems.    Submit your next manuscript at 7 springeropen.com


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items