ABSTRACTWe describe a case study of a complex, ongoing, collabora-tive - PDF document

ABSTRACTWe describe a case study of a complex, ongoing, collabora-tive work process, where the central activity is a series ofmeetings reviewing a wide range of subtle technical topics.is not well-versed in all the topics. We provided tools to cap-ture the meeting discussions and tools to ÒsalvageÓ the cap-tured multimedia recordings. Salvaging is a new kind ofactivity involving replaying, extracting, organizing, and writ-ing. We observed a year of mature salvaging work in the casestudy. From this we describe the nature of salvage work (theconstituent activities, the use of the workspace, the affor-dances of the audio medium, how practices develop and dif-ferentiate, how the content material affects practice). We alsodemonstrate how this work relates to the larger work pro-vage with capture, the inßuence on the people being reportedon and reported to). Salvaging tools are shown to be valuablefor dealing with free-ßowing discussions of complex subjectKEYWORDS activity capture, audio recording, multimedia,LiveBoard, meeting support tools, notetaking, salvaging,work process supportMEETING CAPTURE AND SALVAGEOur interest is how computational tools can support the natu-ral, informal activities that are inherent in human collabora-(1) to support and capture the free-ßowing activities of meet-records effectively in the larger work processes in which theMeetings are productive because of their interactional char-acter. The rapid give-and-take of conversational exchangesproduces insights and shared understandings. However, it isoften difÞcult to document the content and process of meet-ings. The result is that there are lost opportunities to makespectives, needs, decisions, reasons, caveats, and so forthÑexpressed at meetings, as well as the modes of expression(enthusiasm, caution, etc.). We are exploring one approachto this problem: to capture not only the written artifacts ofthe meeting, such as prepared materials and notes taken atthe meeting, but also audio and video records of the courseof activity of the meeting. Multimedia records provide a richresource with which to revisit the course of the meeting, toreexperience and reinterpret its details, tenor, and tone. Ourgoal is to understand how multimedia records can be used.reporters who routinely use audio recorders. But new tech-nologies for capturing, indexing, and accessing multimediarecords can expand the ways with which recorded materialscan be worked. We call the new activity of working with cap-tured records Òsalvaging.Ó The research challenge is tounderstand the nature and efÞcacy of this new kind of activ-ity and how it can be evolved into effective work practices.Much of the emerging research in multimedia capture seemsto take a cognitive view of capture and salvage as a memoryaid, such as studies testing the retrieval of answers to speciÞcal of answers to speciÞc)research issues of human memory, our view is that the cap-turing and salvaging of meetings needs to be understoodwithin the social and organizational context of the largerwork processes. The development of effective practices ofcapturing and salvaging meetings must be done by interrelat-ing them with other work practices.Consistent with this orientation, we have devoted much ofour research to a particular case study, where we create andexplore the use of capture and salvage tools in the context of 1.We want to emphasize that we are exploring multimedia inorder to help people cope with the complexity of freeßowing dis-cussions. Media recordings are commonly used for quite differentreasons: for accountability, such as in legal processes (Òanythingyou say can be held against youÓ) and security (surveillance). Thisis the very opposite of what we are trying to achieve. We donÕt wantto inhibit people by making them feel accountable for every state-ment, but rather we want them to feel free to discuss ideas and beable to revisit them. Thus, trust and understanding of how captured ÒIÕll Get That Off the AudioÓ:A Case Study of Salvaging Multimedia Meeting RecordsThomas P. Moran, Leysia Palen,* Steve Harrison,Patrick Chiu, Don Kimber, Scott Minneman, William van Melle, Polle ZellwegerXerox Palo Alto Research Center, 3333 Coyote Hill Road, Palo Alto, CA 94304*Department of Information and Computer Science, University of California, Irvine, CA 92697{moran,harrison,chiu,kimber,minneman,vanmelle,zellweger}@parc.xerox.com, palen@ics.uci.edu a real ongoing work process. The value of this approach isthat we can participate in the co-evolution of tools and workpractices that are effective in the work setting and can thusunderstand how these kinds of tools Þt into the social contextof an extensive work process.WHAT IS SALVAGING?the board, taking notes, moving around, coming and going,laughing, pointing to materials, and so on. These activitiescan produce various, such as drawings from thewhen the meeting is captured become additional artifacts. It thatcan be used during salvaging to access the speciÞc activitiesof the meeting for replay. In our case, we use timestampedrepresentations resulting implicitly from the meeting activi-speaker changes, etc.The central activity discussed in this paper is what we havesalvaging (in the dictionary sense of Òsaving for fur-ther useÓ). It involves culling through the artifacts of a meet-ing to dig up useful pieces, indexing them, relating them,rechecking them, organizing them, and creating new materi-als with them. The notion of salvaging is richer than justretrieval or browsingÑit is an active process of process ofmeeting more easily accessible to potentialconsumerssimply want to ÒreadÓ about the meeting.The concept of salvaging, versus simple consuming, is valu-able when the artifacts of capture are not themselves easily In many important meetings, the process and/or content is too complex, subtle, inarticulate, and/or chaoticto allow the creation of accurate or understandable meetingartifacts in real time. The best the meeting participants cando is create adequate indices into the activity to support latersalvaging.The salvagerÕs goals can vary considerably, depending on theanticipated consumers and their needs. The salvager mightwant to simply Òclean upÓ the meeting artifacts to serve asmeeting minutes. Other salvaging goals include clarifyingthe rationale for meeting decisions, clarifying the attributionsof ideas, putting meeting statements in context, and Þlteringout sensitive parts of the meeting record.Salvaging can produce a wide range of artifacts that addressdifferent kinds of consumers. The simplest artifact is a textreport, which makes for quick consumption. Or the salvagermight want to create an index to allow a consumer to selec-tively replay parts of the meeting. In this case, the salvagerthe notes from the meeting to serve as good entry points for 2.Part of our current research is exploring tools that automati-cally identify and present artifacts from meetings on the Web withlittle or no intermediate salvaging effort.3.It should be pointed out that we do not claim that all meet-ings are worth salvaging. An interesting phenomenon we haveexperienced is that it is often difÞcult to know until after a meetingwhether there might be something worth salvaging.playback. Or the salvager might want to go all the way andor set of meetings. In this extreme case, the activity of sal-vaging is like multimedia authoring. In general, however, webelieve that most salvaging is oriented more to sense-makingartifacts than to high-production-value presentations.Salvaging performance varies substantially depending onmultiple factors, such as whether the salvager was present inthe captured session, the salvagerÕs expertise with the tech-nology, the salvagerÕs familiarity with the material being dis-cussed, the salvagerÕs goals, and the type of event or meetingcaptured. We will report on our studies of these dimensionsof use and how they shape different salvaging behaviors in afuture paper. The present paper focuses on a case studyinvolving one salvager over a long period with evolvingexpertise and utilizing materials of differing familiarity.RELATION TO OTHER RESEARCHWe have been working on capture and salvage tools for manyyears, beginning with theWhereWereWe system [8] (whichfocused on playback during the captured event itself). Wetools to achieve effective capture in meeting settings [9]. Wereal setting in [10]. However, in [10] we focused on thevaging.Schmandt and his colleagues have been working for manyyears on making audio usable. In [4] they explore ubiqui-tously capturing audio in the work environment, and theyshow techniques for accessing the audio. Using the strokeson a sketching application to index captured audio (andvideo) was Þrst reported in theNoTime system [7]. TheFilochat system [15] used stroke indexing for notes taken atAudio Notebook [14] is a paper-based portablenotetaker that indexes captured audio. All of these are ori-More elaborate collaborative systems are also beingexplored. The BellcoreStreams system [2] focuses on cap-torium setting. Various automatic video and audio indexingtechniques are used to create timeline-like views of the pre-Jabber system [5] captures video conferenc-ing meetings. The research is focused mostly on indexing thespeech recognizer and then applying a lexical analyzer tobuild a tree of keywords that provide an index structure intoClassroom 2000 [1] is a preliminaryeffort at capturing the presentation materials and the audio inthe classroom, using a LiveBoard for presentation and palm- 4.Salvaging is also quite different from the activity of sequen-tial data analysis [13], such as video analysis. A data analystÕs goalis to create a detailed account of the course of activity. Salvagers aremore interested in the content and in extracting the valuable partsand the context necessary for understanding those parts. brief study of people searching for speciÞc events in video-tapes of real meetings using a VCR, showing some of thestrategies they employ. Whittaker et al. [15] ran an experi-Filochat, showing that indexed audio sup-ported better recall, but took longer (there is a similar resultut took longer (there is a similar result)Filochatferent people found it useful to ÒsalvageÓ the meetings.However, the evidence in the current literature about thenature and utility of salvaging is scant.THE CASE STUDYWe have put our capture and access tools to use by engagingwith an ongoing work process at Xerox PARC, the manage-ment of intellectual property. This is not a controlled study,but a naturalistic, longitudinal study where the use of thetools was determined by the demands of the work process.The Work ProcessPARCÕs management of intellectual property is a complex,x,)Researchers are encouraged to submitInvention Proposals(IPs) describing their inventions, which are evaluated by peerreview viaTechnology Assessment Panels (TAPs) of techni-cal experts from the labs. There are several TAPs covering awide range of technical domains, such as solid state devices,large area electronics, image processing, software architec-tures, software applications, and user interfaces. TAPs meeton a regular basis to review submitted IPs, and the results arereported to inventors, managers, and patent attorneys.. His job is to keep the process run-ning smoothly, report to management on the status of intel-lectual property, and look for ways to improve the qualityThere are two settings on which our project focuses: the TAPmeeting and CÕs ofÞce, where he writes his summaries of themeetings. C calls the TAP meetings, in which an average of 6members review an average of 6 IPs. The members bringtheir annotated copies of IPs to be evaluated. In the meetingthey consider each IP in turn, discussing its value alongmany dimensions. There is much give-and-take in the dis-cussion, and membersÕ views can change considerably in theprocess of trying to reach a consensus. The discussions aver-age about 15 minutes. They give each IP a summary rating,but there are always caveats, suggestions, and actions. Cmanages the agenda, facilitates the IP discussions, brings thegroup to consensus, and takes notes on the discussion.CÕs most critical task is then to create reports summarizingeach IP and its assessment. These reports provide feedbackto the inventors, inform the managers and attorneys aboutnew intellectual property, and help TAP members who peri-reports is an extremely challenging task for two reasons. Theject matter in the TAP discussions. Although C, a physicist,was formerly a researcher and is thus knowledgeable aboutsome of the technologies, he is a novice at others, such assoftware. Second, because of his workload, C is often notable to work on a report until a month or more after the TAP,by which time it is difÞcult to rely on his memory and hisnotes are difÞcult to interpret.Capture and Salvage ToolsIn this study, capture took place in the group setting and sal-vage in the individual setting. For the group, we set up a reg-ular PARC meeting room for the TAPs. There was a largetable with audio microphones and a LiveBoard [3]. C satnear the end of the table, so that the TAP members could eas-ily interact directly with each other. We set up the followingÒconfederationÓ of activity capture and salvage tools [9,10]:The LiveBoard ran theTivoli whiteboard application [11] todisplay pages of materials to support each TAP meeting. TheÞrst page held the agenda of IPs, followed by a review pagefor each IP on the agenda. The review page provided a formfor notes of the discussion of the IP. A Tivoli review page isshown in the upper left window of Figure 1. Every action inTivoli, such as switching a page (signaling that the TAPÕsattention was moving to another IP) or making a stroke withthe LiveBoard pen, was timestamped by Tivoli, providingindex points into the audio record. C preferred to take discus-sion notes with a keyboard at the table. So we provided a lap-top with our own notetaking application [9], whichtheir timestamps onto the Liveboard, where the TAP mem-in Tivoli as aclock iconconsisting of 1-5 lines, about every 2-3 minutes.For the individual setting, C was provided with a ÒsalvagestationÓ in his ofÞce. The salvage station interface, shown inFigure 1, presented the Tivoli display, a set of playback con-trols (play, stop, and forward or backward 10 seconds), and aword processor for creating the IP reports. The salvage sta-tion provided C with random access into the audio at theindex points. The main indices were the pages. By going tothe Tivoli page for a particular IP, C could play the discus-Figure 1.Salvage Station Display. sion of that IP. Within the page, C could use the beamed andhandwritten notes as indices. Gesturing on any clock (infront of a beamed note) or stroke caused audio to play.C could also use Tivoli as a workspace to create furtherannotations and indices during salvaging. When a clock wascreated during salvaging, it indexed the current playbacktime. Annotations could be typed or sketched anywhere onthe workspace, since Tivoli is a freeform editor. Figure 2shows an example of a salvage artifact created by C. Theand annotations to the right were added during salvaging. of salvaging tools, not with the details of their userinterfaces. User interface issues will be discussed elsewhere.We studied CÕs salvaging process under natural-istic working conditions. The volume and complexity of CÕsjob as manager of the IP evaluation process kept himextremely busy. We wanted him to Þt our tools into helpingwith the demands of his job; we could not afford to impose anew routine on him under the guise of experimental control.C managed his own schedule. The salvage tools were madeavailable to him for report writing in his own ofÞce.Although the use of these was left to his discretion, he didincorporate them into his routine work practice.CÕs report writing for a given TAP was distributed across ses-sions, locations, and tools. He would work when he couldÞnd the time, he sometimes took work home, and he some-times worked on the text of reports without the need for sal-vage tools. Our study focused only on the sessions where heused the salvage tools.We used four sources of data in this study: (1) We col-lected all the meeting and salvage artifacts (e.g., Figure 2).(2) The salvage tools were instrumented to produce time-stamped logs of all user interface operations. (3) We inter-viewed C several times in his ofÞce (so he could refer tomaterials he used); these interviews were audio taped. Wealso had frequent informal discussions with C. (4) Finally,we videotaped his salvage sessions. Cameras were mountedin his ofÞce, and we provided a dedicated VCR. C was verycooperative and started the VCR every time he salvaged.Time Span.We Þrst installed the capture and salvage tools in1/94 [10], and they have been in continual use since then.The present study of CÕs salvaging work focused on theCASE STUDY FINDINGSWe present Þndings showing the intricate nature of salvagingwork, how CÕs practices developed over time, and how Cresponded to different content domains.CÕs salvaging practices were fairly stable at the beginning oftime C already had over a yearÕs experience with the tools).However, new practices emerged near the end of the studyperiod, which we call his ÒlateÓ phase. We present here threesalvaging sessions representative of these phases and of thediverse content domains he works with. The differences inCÕs work practices can be described in terms of his strategiesand pattern of activities.Salvaging StrategiesC developed multiplesalvaging strategies for working with1.Writing the Þnal report while listening to the cap-2.Creating annotations and audio indices in the3.Writing the Þnal report from textual notes only(meeting notes and any additional annotations).C employed these strategies in response to different condi-tions. For example, he used strategy 3 when he had a particu-larly simple TAP (on a familiar topic). Strategy 1 wasemployed mainly in his early phase, and he began to usestrategy 2 in his late phase. Strategy 3 was used to follow onthe results of strategy 2. 5.We provided a separate Sun workstation as the salvage sta-tion. C used a PC for most of his work. Although the integrationwas not ideal, C was able to get data from one workstation to thatother. Most important was getting data into Lotus Notes, whichFigure 2.Sample of Salvage Annotations.added during the salvage session are grey (red in theoriginal). Text is ÒencryptedÓ to mask proprietary data.(Salvage on 11/17/95 of TAP meeting on 10/23/95.) Listening ProÞlesDifferent salvaging strategies result in different patterns ofactivities, such as reviewing notes from the meeting, creatingnew text, choosing which part of the meeting to listen to, andso on. We have found it useful to represent the pattern ofactivities on asalvage proÞle, which plots the salvagerÕsactivities on a graph with two timelines, the timeline of thesalvage session (horizontal axis) and the timeline of themeeting being salvaged (vertical axis). A characteristic sal-vaging activity is listening to a playback of the recordedaudio. A proÞle showing the listening activities is called alistening proÞle. Listening activities are shown as 45 degreeFigure 3 shows how listening proÞles portray different pat-terns of salvaging activity. The baseline pattern is to listen tothe audio straight through from start to Þnish. Salvagersoften stop and restart the audio, they skip segments of audio,and they relisten to segments of audio. Note that these four: the salvager is basically conform-Listening proÞles of three of CÕs salvage sessions are pre-sented in Figures 4Ð6. These proÞles have additional infor-mation. The dashed horizontal lines show how the TAPmeetings were divided into segments discussing differentIPs. The grey areas highlight the fact that CÕs listening pat-tern shows that he systematically attended to one IP at a timewhile salvaging. On the right of the proÞles of the Þrst twoÞgures are graphs plotting the number of times each segmentof the meeting audio was played during the salvage session.Early Salvaging PhaseThe two salvage sessions in Figures 4 and 5 are typical ofCÕs early phase. The Þrst session deals with a software TAPand the second with a physics TAP. Because of CÕs heavyworkload at that time, both of these salvage sessions laggedtheir TAP meetings by about two months.salvages the IPs in the order they were discussed in the TAPmeetings. Within IPs, he proceeds through the meeting audiosequentially, with no skipping around. A sequential strategyis the safest way to navigate through the audio data. It makesit easy to keep track of what has and what hasnÕt been sal-vaged. Jumping around the audio in a non-sequential fashionrequires other props for keeping track.CÕs strategy in both of the early sessions is to type the reportin the word processor while simultaneously listening to thetext by to the audio and using the meetingnotes and his own recall. Other features:Features of the Early Software Salvage Session (Figure 4):2. The listening pattern shows stops and starts. For example,at around salvage minute 30, C stops for about six min-then consulted the hardcopy of the IP and completed thetext of his report before going to the next IP.3. C replays the audio, but only a little. For example, at aboutsalvage minute 10, C relistens to the Þrst part of the audio.some better phrases to improve his text.4.While he types the report, he lets the audio play continu-background, as can be seen at salvage minutes35-50. Continuous play allows him to listen for importantpoints while also ensuring that he gets total coverage ofFeatures of the Early Physics Salvage Session (Figure 5):1.In contrast to the software session, C does not listen to allof the audio. For example, he skips a large segment in the2.He repeatedly listens to a particular segment of the meet- 6.A detailed examination of the video of this session revealedthat C did attempt to salvage the IPs in a different order, using ahardcopy agenda as a guide. However, the meeting did not followthe order of the hardcopy agenda; and C was very confused. Heaudio medium to order his work.Figure 3.ProÞles of Listening Patterns. op / startSalvage TimeMeeting Time Figure 4.Listening ProÞle of an ÒEarly SoftwareÓSalvage Session.(Salvage on 5/19/95 of TAP meeting on 3/22/95.) 1020304050 ing audio in the Þrst IP. For example, at salvage minute 10he relistens in order to check the text he just typed. Healso creates a clock to index the replayed segment.3.There is an interesting case of faulty navigation in theaudio in the second IP. C starts on the second IP by typingsome text (salvage minutes 17-21), and then he wants tosimply presses the play button, which plays the audiothe Þrst IP. It takes C a few seconds to realize that audio isfrom the Þrst IP. He decides to let the audio play out whilehe occupies himself by organizing papers and by typingsome initial text for the second IP. By the time he hearsLate Salvaging PhaseFigure 6 presents a listening proÞle of CÕs salvaging a soft-ware TAP meeting 8 months later. This proÞle reveals anadvanced salvaging practice. From the video data, it is alsoapparent that CÕs behavior is quick and skilled, with a densemix of activities.C was very focused, as he approached this session only 3days after the TAP with the issues he wanted to deal withfresh in his mind. CÕs strategy was to work carefully atunderstanding particular segments of the meeting and withcompiling additional notes, which he added to the workspaceon Tivoli (as in Figure 2). He would then use these in a laterFeatures of the Late Software salvage session (Figure 6):1.In contrast to his early phase, C does not follow the meet-ing order in salvaging IPs. He skips from the Þrst to thefourth IP, apparently because this was the order of IPs onhis hardcopy of the meeting notes.2.C listens to only a small fraction of the audio record.3.C replays relevant segments of audio over and over. Hebacks up and relistens to short segments while transcrib-ing the spoken words next to the meeting notes. C alsoscriptions begin. For example, in the third IP, C very care-software logic.4.C has evolved special marks in his meeting notes to locatethe relevant audio segments. The mark is ÒHAÓ (for ÒhearaudioÓ). (See Figure 2 for examples of HA marks.) Dur-ing the meeting, C would type HA to signal points hethought were important, well-articulated or difÞcult tounderstand, anticipating that he would want to revisitthem during salvaging. Figure 6, on the right, showsseen that C uses the HA marks to focus his listening activ-ity. In some cases, the points he marked are no longer rel-evant to him. In other cases, he uses the approximatelocation of HAs to localize his salvaging activities.In a later session, C used these salvage artifacts to write theactual IP reports. We examined his Þnal reports and foundthat about half the notes he transcribed in the salvage sessionwere incorporated fairly literally into his Þnal IP reports andconstituted a signiÞcant portion of the resulting text.Salvaging Different ContentDuring interviews, C told us that he felt much more at easedealing with the content material of the physics TAPs thanthe software TAPs. Familiarity with content appears to be astrong determinant of the kind of salvaging strategies Cemploys. These strategy differences became more pro-nounced over time. In the early phase (Figures 4 and 5), wesee that he felt the need to listen to the audio of softwareTAPs more completely. In his late phase, where he was try-ing to be more prompt and focused in his salvaging, the dif- 7. C also used HAs to mark parts of the meeting that went toorapidly to take notes on. For example, in this session C marked atime when a list of speciÞc items was being rattled off; this list wastranscribed during salvaging and included in the report.Figure 5.Listening ProÞle of an ÒEarly PhysicsÓSalvage Session.(Salvage on 5/12/95 of TAP meeting on 3/8/95.) 3040 Figure 6.Listening ProÞle of a ÒLate SoftwareÓSalvage Session.(Salvage on 12/2/95 of TAP meeting on 11/29/95.) 2030405060 ferences have become more pronounced. C engaged in moreelaborate indexing and salvaging and made more use of thesalvage artifacts in the software TAPs. Figure 6 shows C dig-stand the technical arguments, during which he createsfurther annotations on the salvage artifact. These activitiesdo not occur while salvaging the physics TAPs.We have not presented an example of a late physics salvagesession. Instead, we show the differences between the twoThe differences are clearly revealed by comparing measuresof various activities across a number of late phase TAPs. Themeasures are: the number of explicit written references to theaudio C creates during a TAP meeting; the number ofÒchunksÓ of textual annotation C adds to the salvage artifactwhile salvaging; and the number of clocks he creates on thesalvage artifact. We counted these features in CÕs salvageartifacts for a span of ten months (9/95Ð6/96) in four differ-ent TAPs, two physics TAPs and two software TAPs (bal-anced for frequency and dates). The table in Figure 7conÞrms that C hardly ever engages in the more elaboratesalvage activities when dealing with the familiar physicscontent domain. In fact, C claims that he often does not needto use the audio to report on the familiar physics TAPs. Thesalvage tools are more valuableÑindeed crucialÑfor deal-ing with difÞcult and unfamiliar materials.FURTHER OBSERVATIONS ON SALVAGINGThe case study raises a multitude of general issues; many ofthese are concerned with the relationship of salvaging toother people and work in the larger work setting.Evolving Practices for EfÞciency.We saw in the Late Soft-ware salvage session that C uses the audio in a very focusedway to, as C says, Òmine for gold.Ó He localized his searchfor valuable pieces of information by ßagging those ideasHAs is a practice that has evolved over time. Early in his useof capture tools, C would often say during a meeting, ÒIÕllget that off the audioÓ, which signalled to the TAP membersthat C knew something important was being said, eventhough he didnÕt take full notes on it at the moment (he wasfrequently in the middle of typing an earlier note). However,this utterance was not something he could later use as anindex into the audio. To create a useful index, he eventuallybegan writing explicit audio references, such as Òhear audio,Óin his notes during the meeting. One meeting later, he short-ened this to ÒHAÓ and added HAs to his regular note-takingvocabulary. Other factors converged to make this practicework. C was making a special effort to be efÞcient and tovage soon after a TAP meeting, C was more prepared tointerpret the signiÞcance of the various HAs and hence touse them effectively in salvage sessions. The development ofthe HA convention and the strategy for using it is a primeexample of his evolving practices. Over time he developedmany specialized strategies to deal with the particularities ofthe different TAPs and different circumstances he encoun-tered in the work process.The Salvage Artifact as a Workspace.C deals with manyartifacts during salvaging: both in hardcopy (the IP, the meet-ing notes, email) and on the workstation display (the meetingnotes, the draft text of the report), as well as the audiohis salvaging task. One way to organize the various resourcesis to move any relevant information from all sources into thereport. However, C found that it was better to have an inter-mediate place to assemble information during salvaging. Hethe meeting notes. From an interview with C, we can discernthat he did this to limit the complexity of having to simulta-neously deal with two different places, two different organi-zations, and two different tasks.C uses the meeting notes to index into the audio. He isfocused on this display window and not on the report win-dow. The meeting notes are ordered by the discussion in themeeting, and he follows the discussion order to salvage,whereas the draft outline of his report is often in a differentorder. C says he is ÒÞshing for informationÓ when salvaging,and it is better not to try to organize the information at thesame time. Further, C claims that as Òyou listen to the mate-rial, you change the organization it is Þnally going to go intowork with them.ÓSpeed/Accuracy Trade-Off.spends on his writing and the thoroughness and accuracy ofhis reportsÑthe classic speed/accuracy trade-off. Captureand access tools allow, even encourage, greater accuracy.WhittakerÕs study [15] showed that retrieval from capturedaudio was more accurate, but his subjects stopped short of100% accuracy (retrieval just took too long). The larger workcontext is a strong determinant of where to strike the bal-ance. For example, for a period of time C experimented withÒquick reportsÓ (brief reports produced without salvaging),but these were not acceptable to the inventors, who wantedreporting. Salvage tools do not determine a particular placeon the speed/accuracy trade-off; it is up to the user to man-age this. Good tools can alter the trade-off curve somewhat,but perhaps more important is that they allow users a greaterrange of possible places on the trade-off. The utility of cap-ture and salvage tools in a given work setting must be under-stood in part by the speed/accuracy demands.Figure 7.Three Measures of Salvaging in Different(The measures are averagesper IP. A-D represent different TAPs.)PhysicsSoftwareABCD.29.401.511.29created during salvaging.00.00.63.95salvaging.00.00.49.29 We saw in theLate Software salvage session that C sometimes carefullytranscribes the words of TAP members, especially on topicsless familiar to him; and he uses these transcriptions in hisreports. The practice of borrowing words from the audio pro-duces an accurate report of the TAP discussion, which isappreciated by the inventors. But it also has interestingeffects on the TAP process. We reported in [10] that there isan awareness of the audio in TAP meeting and that TAPmembers often seem to Òspeak to the record,Ó that is, theyaddress C in his future role as salvager as much as in his cur-rent role as meeting facilitator. They are conÞdent that C willÒget itÓ later, even if he doesnÕt at the moment. The result ofis no longer a need for TAP members to critique drafts of theIP reports (something they needed to do frequently beforethe use of our tools [10]). One TAP member commented tous that CÕs reports using our tools better represent the that is expressed in TAP meetings. This isimportant not only for the richness of the reports, but also forthe satisfaction of individual TAP members, who see theircontributions being explicitly used.By being fairly literal in his reporting, C is attempting to usethe language that both the TAP members and the inventors He is acting as a channel of communication betweenthat is, that salvaging could produce a multimedia report,appropriately Þltered, that directly and vividly communi-indeed one of the goals of our current and future work.istic salvaging context reported here (together with otherstudies to be reported in a later paper) begins to provideinsights into important characteristics of salvage tools. Let usnote two design implications (among many) for salvagetools. First, tools should provide a workspace to collect andintegrate information. The workspace should serve as aÒstaging areaÓ for organizing the information, but not requirethat the information be organized as it is assembled. Second,tools should provide ways for a salvager to manage, struc-ture, and control the salvage process itself. In particular, bet-ter tools for orienting and navigating within the audiochannel are needed. We have prototyped an improved time-ed time-ing activities (e.g., notetaking, page turning). It would alsobe valuable to indicate which portions of a meetingÕs audiowere played in this (or previous) salvage sessions.We have demonstrated a real-world use of tools to capturefree-ßowing discussions of complex subject matter and sal-vaging tools to produce high quality documentation. Thesetools are still in use and have become a vital part of the workprocess. We have described the nature of salvaging in prac- 8.In fact, reports in the software area are now longer and moretice and how the practices evolve; we have observed somegeneral features of salvaging activity; we have suggestedsome design implications for supporting salvaging; and wehave shown that salvaging cannot be thought of as an iso-lated task, but that it is intertwined with the capture setting.ACKNOWLEDGMENTSWe would like to acknowledge and thank colleagues who workedHebel, Gordon Kurtenbach, Bill Janssen, and Sara Bly. We thankthe many TAP members for their supportive collaboration andAnnette Adler for many discussions and comments on this paper.[1]Abowd, G. D., et al. (1996). Teaching and learning as multi-Proceedings of[2]Cruz, G., & Hill, R. (1994). Capturing and playing multime-dia events with Streams.Proc. of MultimediaÕ94[3]Elrod, S., Bruce, R., et al. (1992). LiveBoard: A large interac-tive display supporting group meetings, presentations, andProceedings of CHIÕ92[4]Hindus, D., & Schmandt, C. (1992). Ubiquitous audio: cap-Proceedings of CWCWÕ92[5]Kazman, R., Al-Halimi, R., Hunt, W., & Mantei, M. (1996).Four paradigms for indexing video conferences.[6]Kimber, D., Wilcox, L., Chen, F., & Moran, T. P. (1995).Speaker segmentation for browsing recorded audio.Proceed-[7]Lamming, M., & Newman, W. (1992). Activity-based infor-mation retrieval: technology in support of personal memory.Personal Computers and Intelligent Systems: InforemationProcessing Ô92[8]Minneman, S., & Harrison, S. (1993). Where Were We: Mak-ing and using near-synchronous, pre-narrative video.Pro-[9]Minneman, S., Harrison, S., Janssen, B., Kurtenbach, G.,Moran, T. P., Smith, I., & van Melle, W. (1995). A confedera-tion of tools for capturing and accessing collaborative activity.Proceedings of MultimediaÕ95[10]Moran, T. P., Chiu, P., Harrison, S., Kurtenbach, G., Minne-man, S., & van Melle, W. (1996). Evolutionary engagement inan ongoing collaborative work process: a case study.Proceed-[11]Pedersen, E., McCall, K., Moran, T. P., & Halasz, F. (1993).Tivoli: An electronic whiteboard for informal workgroupProceedings of INTERCHIÕ93[12]Russell, D. M., SteÞk, M. J., Pirolli, P., & Card, S. K. (1993).Proceedings of Inter-[13]Sanderson, P. M., & Fisher, C., (Eds.) (1994). Special issue onexploratory sequential data analysis.[14]Stifelman, L. J. Augmenting real-world objects: a paper-basedConference Companion to CHIÕ96[15]Whittaker, S., Hyland, P., & Wiley, M. (1994). Filochat:Handwritten notes provide access to recorded conversations.Proceedings of CHIÕ94,[16]Wolf, C., Rhyne, J., & Briggs, L. (1992). Communication andinformation retrieval with a pen-based meeting support tool.Proceedings of CSCWÕ92,