ENCOURAGING A CULTURE CHANGE IN TASK

MANAGEMENT WITHIN PIM TOOLS

Roger Tagg

1,2

, Leonard Koh

and Tamara Beames

1,2

(Formerly at)

University of South Australia, Mawson Lakes SA 5095, Australia

Keywords: Personal Information Management, Task Management, User Interface, Adoption of Technology.

Abstract: Personal Information Management (PIM) software tools are becoming increasingly important in easing the

high information workloads of today’s knowledge workers. However despite the efforts of both commercial

vendors and research teams, significant improvements have yet to find their way into mainstream

commercial tools and common usage. This paper focusses on two particular aspects: improved support for

task management and better user interface metaphors. We also address the issue of how current knowledge

work culture affects the way in which PIM tools are utilised, and of adoption of these tools.

1 INTRODUCTION

There continues to be pressure on knowledge

workers, especially since the recent global financial

crisis, to achieve more and more in less time and

with fewer resources. This situation has been

exacerbated by increasing administrative controls

and tighter competition for business. IT assistance

seems to have hit a road block; users are still using

much the same everyday tools for Personal

Information Management (PIM) as they were a

decade ago.

This paper looks at possible ways round this road

block from two angles: firstly though suggested

improvements to today’s PIM tools in the areas of

task management and user interfaces; and secondly

through addressing issues of culture and adoption

that limit the pace at which any improvement can be

achieved.

The ideas in the paper are based on experience

with our VPS (Virtual Private Secretary) project

(Tagg et al 2009). This project aims to ease

information overload by providing software support

similar to that which a human secretary might

provide.

Section 2 of this paper summarizes user feedback

to some demonstrations and presentations of VPS

prototypes. Section 3 describes certain extensions

proposed for VPS (and PIM tools in general) in the

task management and user interface areas. Section 4

reviews some other notable PIM projects worldwide.

Section 5 raises issues of adoption of new

technologies, both in general and for PIM in

particular. Section 6 offers some concluding

thoughts and ideas for future work.

2 THE VIEWS OF USERS

In (Tagg and Beames 2009) we raised the question

“what will it take for a critical mass of users to

change from the toolkits they currently use for PIM

(e.g. MS Outlook or Lotus Notes) to a new, more

intelligent and more task-oriented interface for

everyday information and communication

management?” We suggested that, as well as a major

software improvement, a change in human-to-human

etiquette will be needed. For example, current

culture is to send messages to other users without

any contextual clues for the receiver.

In early 2009, two of the current authors had

held detailed interviews with a small sample of

potential users, including an academic, an

administrator and a marketing manager.

To follow this up, the other author carried out

two informal surveys of user habits and cultural

attitudes. Respondents were attendees at “project

fairs” at which his Honours project (involving

extensions to VPS) was demonstrated – along with

other student projects - to industry sponsors,

academics and other students.

137

Tagg R., Koh L. and Beames T. (2010).

ENCOURAGING A CULTURE CHANGE IN TASK MANAGEMENT WITHIN PIM TOOLS.

In Proceedings of the 12th International Conference on Enterprise Information Systems - Human-Computer Interaction, pages 137-142

DOI: 10.5220/0002889201370142

 SciTePress

2.1 First user Survey (June)

45 respondents were asked about their level of use of

email, their experience of overload, their use of

email for task management and their priorities for

improvement in email tools.

A majority (60-65%) check their emails at least

every morning, and file emails in folders. 82%

admitted experiencing some degree of information

overload, half of these putting the blame largely on

email. Respondents’ task management was not

sophisticated, 82% leaving task-containing messages

in their inbox as a reminder. However only 40%

thought that they needed specific task management

facilities, and only a similar number were even

aware that MS Outlook contained a Tasks

subsystem. Only 18% saw themselves as ever taking

a “tasks first” view rather than a “read all the

messages first” view.

2.2 Second Survey (November)

In the second survey respondents were asked only

about their task management practices and how they

would react if they were required to provide more

clues when requesting other people to perform tasks,

or to change to using PIM tools that were more task-

oriented. 29 responses were received (17 students, 8

industry and 4 academics).

Asked how they currently manage their “to do”

list, responses were widely scattered (see Table 1).

Table 1: “To-do list” approaches used by survey

respondents.

To-do list method Respondents

MS Outlook Tasks

My mobile phone

Special “to-do list” software

Set up tasks as calendar appointments

Mark the emails as “unread”

Write in a diary or filofax

Write on a loose sheet of paper

Keep it in my head

Other (inky palm, whiteboard, QuickNote)

Many respondents said they integrated their total list

of “to-do” items through email. However, when

requesting tasks from others, they gave phoning,

texting or even asking face to face equal preference.

The reaction to changing to using more task-oriented

tools was that if asked to do so, over half the

respondents would conform – but with 50% of

respondents being students, this may not be

significant.

2.3 Comments on the Surveys

Most people we surveyed seem to have built up their

PIM practices over time by experience of working

with whatever tools they have been given. An

informal comment was noted that said “it’s not my

job as an email sender to categorize for the receiver

the messages I send”.

3 PROPOSED IMPROVEMENTS

TO PIM TOOLKITS

3.1 Task-oriented Assistance to

Receivers of Messages

Work on VPS in 2008 (Tagg et al 2009) led to the

development of a prototype system for pre-

categorizing incoming emails through the use of an

ontology and lexicon controlled by the receiver.

In trials of this prototype, it was found that while

the system could adequately recognize the work

topic to which an email belonged, it was less good at

recognizing “taskiness”, i.e. how likely it is that the

message sender is actually asking the receiver to do

something important, and what sort of action (e.g.

acknowledgement, reply, vote casting or sending of

an attachment) is required.

It also became clear that our potential users did

not want “yet another software tool” such as our

VPS prototype would have been. Therefore in 2009

we re-implemented it as a potential “myTasks”

extension to the university’s existing web portals

used by staff and students, using the same ASP.NET

technology as in that portal. However we later

discovered that the portal appears to not be widely

used as a starting point for users’ general PIM work,

and the portal development team did not encourage

our efforts.

3.2 Encouraging Message Senders to

Give More Clues

Given the problems in identifying taskiness, we have

conjectured, along with other researchers, that a

change in the current culture, in which senders

would be encouraged to make any task requests

more specific, is perhaps needed.

In an earlier prototype (Tagg and Mahalingam

2005), we attempted to require task requesters to

complete mandatory fields on a web form, but the

feedback was that most message senders would not

be prepared to do this in the current culture.

ICEIS 2010 - 12th International Conference on Enterprise Information Systems

138

We have instead proposed a “privileged senders”

approach whereby a receiver can say to senders “if

you give me better clues, I can attend to your

requests more quickly”. A possible choreography is

shown in Table 2 below.

Table 2: Suggested choreography for a “Privileged

Senders” add-on.

Receiver (Message) Sender

(Install

“privileged

senders” add-on)

(Store sender

details)

(Receiver’s PIM

agent categorizes

the message and

infers any task)

(Review task

status)

(Complete the

task)

Invitation to be

one of my privil-

eged senders--->

<---------Accept

<-----Send

marked up

message

Acknowledge

and agree/refuse-

---->

<----Reminders

and changed

deadlines

Progress report

or extended

deadline request

----------->

“Done” message

(manual or

application-

generated)---->

<--Accept or

reject the result

(Install “privileged

senders” add-on)

(Store receiver

details)

(When sending to

such a receiver, get

a prompt to mark

up the message)

(Manage delegated

tasks)

(Check the

receiver’s response

(manually or

automatically)

To support this, a data structure, as shown in the

entity-relationship diagram in Figure 1, would need

to be held by both senders and receivers.

Figure 1: Data Structure to support a “Privileged Senders”

add-on.

Receivers would, of course, also need to maintain

their personal ontology of work categories, some or

all of which would be publicized to senders. A

sender-side prompting function would access the

receiver’s ontology, using either a web service or a

local copy.

3.3 Compliance Management

This is an extension to “privileged senders”,

intended to support users who receive so many task

requests that they find it hard to remember where

they have got to on each task. At the simplest level,

a user might wish to check whether the requester

regards the task as done or not. The choreography

for this is shown in Table 3 below.

Table 3: Suggested choreography for a “Compliance

Management” add-on.

Receiver (Message) Sender

(Install “compli-

ance management”

add-on)

(Store sender

details)

(Receiver’s PIM

agent recognizes

the task)

(Review one’s to-

do list)

<---Advertise

offering of a

compliance

service

Optionally

advance------->

<---Send a task

with option to

join the

compliance

service

Optionally

fly -------->

Check my

compliance----->

<----Compliance

status

(Install “compli-

ance management”

add-on)

(Add compliance

info to receiver

details)

(If it is a new task

type or series, set it

up – otherwise start

the task instance)

(Check receiver’s

compliance status

by web service)

The data to support Compliance Management would

have the same structure as in Figure 1, but would

have additional fields. However a major addition

would be needed on the sender side, namely the

“Compliance Register”. This would be a web-

accessible database that would respond to “Check

my Compliance” requests, for which a user login

and task id would be required.

It is worth noting that tasks are rarely stand-

alone, but form part of multi-stage business

processes. There is therefore a likely need for more

integration between PIM systems used for task

management and communication of tasks, and

Workflow Management Systems.

3.4 Visualization Improvements

The “inbox” metaphor is well ingrained for both

email and tasks, but may well not be optimal. One

commonly proposed alternative is a spatial layout,

Task Type

or Series

Sender (or

Receiver)

Details

Task

Instance

Task Instance

Event

ENCOURAGING A CULTURE CHANGE IN TASK MANAGEMENT WITHIN PIM TOOLS

139

with task clusters dotted around the desktop.

In the “Task Type Thumbnails” (Bardram et al

2006); and “Scalable Fabric” (Czerwinski 2006)

metaphors, task groups are represented by

thumbnails which expand when double clicked, the

other task groups being downsized and moved to the

periphery of the screen. Another example is the

“paper piles” metaphor suggested by (Lam 2005),

illustrated in Figure 2.

Figure 2: Paper Piles interface as proposed in Lam’s

(2005) MailStacker prototype.

Task Vista (Bellotti et al 2007) uses a paper

notebook metaphor with tiles for task groups, and

links to documents and people. Lotus Connections

(IBM 2009) uses an “office telephone metaphor”

with buttons for frequent task groups.

Several prototypes have attempted to represent

visually the flow of tasks within a larger process,

e.g. by showing task threads (Bellotti et al 2007,

Kerr and Willcox 2004) or by mapping emailed

tasks onto a horizontal time scale, e.g. TaskView

(Gwizdka 2002).

A number of systems still use inbox-style task

lists but with add-ons such as pop-up windows,

coloured flags, etc. Examples include coloured icons

(Bellotti et al 2007); and intuitive icons for task

groups (Ferreira et al 2005).

Task Tracer (Stumpf et al 2006) offers prediction

of which task area an email belongs to, and a Task

Explorer that shows all the documents that relate to

that task. TaskMinder (Landry et al 2002) uses

varying font size text to represent different priorities

in a clickable summary list.

A number of interesting visualization models

have been proposed for ancillary perspectives.

Dionypos Task Recognizer (Rath et al 2008) uses an

“electric light” icon with radiating spokes indicating

what documents have been accessed, in order to give

clues to recognizing tasks. Collaboration Maps

(Hawryszkiewycz 2008) and Participation Maps

(Constantine 2006) address the need to show a

model of human as well as computer-related

interactions.

4 RELATED WORK

While a number of research teams have been

working on PIM for around a decade, it appears that

little of this good work has so far found its way into

mainstream commercial PIM tools. Most projects in

the table below include some degree of recognition

of tasks, use of ontologies and integration of the

multiple PIM functions. Most have also integrated

their tools into email clients, which are regarded as a

natural conduit for tasks (Whittaker et al 2006).

However many of these projects have now been

terminated.

Table 4: Other recent PIM research prototypes.

Research Group Prototype Reference

PARC TV-ACTA Bellotti et al 2007

Microsoft

Research

SmartMail

Corston-Oliver et

al 2004

IBM (Lotus) Connections IBM 2009

IBM Remail

Kerr & Wilcox

2004

IBM Activity Explorer

Muller & Geyer

2004

IBM Scou

Sow et al 2006

Open Software Chandle

Chandler 2008

EU / Uni of

Kaiserslau

ern

Nepomuk

Sauermann et al

2008

EU consortium DELOS/TIM Catarci et al 2007

Carnegie-Mellon

SEI / DARPA

RADAR Freed et al 2008

Pi Corp

Smart Desktop (was

Task Tracer)

Naone 2007

Apart from these there is a vast number of email

add-ons being developed outside of research teams,

both commercially and as freeware/open source.

Examples include Google Gmail Tasks, Xobni,

ClearContext and Nelson Email Organizer,

targetting especially Microsoft Outlook and the open

source client Mozilla Thunderbird.

5 ADOPTION CHALLENGES

A major challenge in introducing any PIM

improvement is not just to make a good tool; it is to

motivate users, both individually and in groups, to

actually use it.

5.1 For Individuals

Rogers (1983) describes a number of models

ICEIS 2010 - 12th International Conference on Enterprise Information Systems

140

affecting the adoption of any innovation. One

describes the adoption stages in which a user

proceeds through, from first becoming aware of the

innovation (“knowledge” stage) to using and

continually evaluating it (“confirmation” stage).

Another model describes the features which may

affect the rate of the adoption of an innovation,

including the “relative advantage” a user will gain

from its use and the ability to trial it before full

adoption (“trialability”). Categories of user types

and their characteristics relating to their rate of

adoption of an innovation are also given.

Another prominent model looking into the user

decisions to adopt a new technology is the Unified

Theory of Acceptance and Use of Technology

(UTAUT) by Venkatesh et al. (2003). This model

consolidates eight previous models to describe four

core determinants and their moderating factors for a

user’s intention to use a technology and their

subsequent usage behaviour.

These, along with other models, provide

information about many factors which are expected

to affect a user’s motivation and decision to trial and

adopt an innovation or new technology, such as PIM

tools, add-ons - or functionality that already exists

but is not currently used - within their existing tools.

5.2 For Working Groups

The work by Rogers (1983) also applies to groups of

users; it describes how the knowledge and usage of

an innovation is “diffused” throughout the group, or

a larger network. Elements affecting the diffusion of

an innovation are the characteristics of the

innovation (e.g. its perceived benefits), the means

through which it is communicated (e.g. mass-media

marketing versus. word-of-mouth), the social system

in which it is being diffused, and time (any new

product requires time to diffuse throughout a group).

Looking into the motivating factors of the

adoption of a new tool within a group or network

setting is particularly important when the tool has

group functionality, such as messaging or group task

management. Often these tools become more useful

as more users adopt (as with e.g. Facebook,

LinkedIn), and so the users which have adopted at

any particular time are interesting as a group, and

not just as individuals.

Much depends on the intensity of a group’s work.

There is a big difference between e.g. on the one

hand, brainstorming meetings and on the other,

committees that meet occasionally. In the latter case,

many of the tasks are done in the space between

meetings; this leads to the importance of recognizing

the workflow connection.

5.3 For Organisations

Organisations should be motivated by activities

which can save costs and increase profits, such as

raising administrative efficiency, reducing stress on

employees and improving service to customers.

Successfully deploying improvements to PIM tools,

and training users accordingly, would be in an

organisation’s interest.

However, it seems that organisations are not yet

completely in this mindset. Cain (2006) notes how

typical knowledge workers, and even the most

highly paid workers, are spending up to a quarter of

their work week in an email client (used heavily for

their PIM activities), without having been trained to

use it effectively. It is perhaps symptomatic that

email clients are often used as a user’s primary PIM

tool, without any better solution or improvements

being considered.

5.4 For Software Vendors

For software vendors, it appears that the risk of a

“false dawn” – with consequent embarrassment to a

vendor’s brand – may be a reason for reluctance to

offer improved PIM tools commercially. Vendors

are also aware that analyzing natural language text

to determine precise meaning is not yet a mature

technology. It is likely that open source vendors,

whose motivation differs from commercial vendors,

may come to play a larger part in the PIM market.

6 CONCLUSIONS AND FUTURE

WORK

We believe that dissatisfaction with low effective-

ness information work will soon trigger higher

priority attention to PIM. However, for there to be

any significant change for the better in PIM, several

things need to come together. These are:

1. Better integration of task management with

messaging and business processes;

2. New, imaginative user interfaces that can attract

users to make a change;

3. Entrepreneurial moves by software suppliers

(whether established, new or “open source”) to

offer add-ons or new “disruptive” products;

4. A culture change among users towards giving

more explicit clues to those to whom they send

messages.

ENCOURAGING A CULTURE CHANGE IN TASK MANAGEMENT WITHIN PIM TOOLS

141

Improved algorithms for natural language analysis

and automated learning of users’ work categories

will also contribute, but may become less critical if

user culture can evolve.

We note, however, that evaluation of all PIM

prototypes suffers from the handicap that to get any

realistic assessment, a large group of related

knowledge workers needs to trial the prototype in as

near as possible real-world conditions.

Like many others including DELOS, TV-ACTA

and Nepomuk, our own project has come to an end.

The plans of Microsoft and IBM-Lotus are not clear,

but the RADAR project appears, with DARPA

support and a large team, to be making progress in

many of the directions discussed in the paper. We

hope that their improvements will become widely

available before too long.

REFERENCES

Bardram, J., Bunde-Pedersen, J., Soegaard, M., 2006.

Support for Activity-Based Computing in a Personal

Computer OS. In ACM Conference on Computer-

Human Interaction (CHI), Montreal, Canada.

Bellotti, V., Thornton, J., Chin, A., Schiano, D., Good, N.,

2007. TV-ACTA: Embedding an Activity-Centered

Interface for Task Management in Email. In 4

Conference on Email and Anti-Spam, Mountain View,CA.

Cain, M. 2006. Who Needs Training on E-mail? Gartner

Inc Research note G00141290.

Catarci, S., Dix, A., Katifori, A., Lepouras, G., Poggi, A.,

2007. Task-centered Information Management. In

DELOS Conf. on Digital Libraries, Pisa, Italy.

Chandler, 2008. The Chandler Project. http://

chandlerproject.org/ (accessed 25-11-2008).

Constantine, L., 2006. Activity Modeling. Technical

Paper, Laboratory for Usage-centered Software

Engineering, University of Madeira, Portugal.

Corston-Oliver, S., Ringger, E., Gamon, M., Campbell, R.,

2004. Task-focused Summarization of Email. In ACL

2004 Workshop ‘Text Summarization Branches Out’,

Barcelona, Spain.

Czerwinski, M., 2006. From Scatterbrained to Focused: UI

Support for Today’s Crazed Information Worker. In

SIGIR Workshop on Personal Information

Management, Seattle, WA.

Ferreira, J., Barr, P., Noble, J., 2005. The Semiotics of

User Interface Redesign. In 6th Australasian User

Interface Conference, Newcastle, Australia.

Freed, M., Carbonell, J., Gordon, G., Hayes, J., Myers, B.,

Siewiorek, D., Smith, S., Steinfeld, A., Tomasic, A.,

2008. RADAR: A Personal Assistant that Learns to

Reduce Email Overload. In AAAI Workshop on

Enhanced Messaging, Chicago, IL.

Gwizdka, J., 2002. TaskView: Design and Evaluation of a

Task-based Email Interface. In IBM CAS Conference

on Collaborative Research.

Hawryszkiewycz, I., 2008. Modeling Increasingly

Complex Socio-Technical Environments. In ICEIS

Conference, Barcelona, Spain.

IBM, 2009. Lotus Connections. http://www-01.ibm.com/

software/ lotus/ products/ connections/ activities.html

(accessed 30-12-2009).

Kerr, B., Wilcox, E., 2004. Designing Remail:

Reinventing the Email Client through Innovation and

Integration. In ACM Conference on Computer-Human

Interaction, Vienna, Austria.

Lam, B., 2005. MailStacker: Applying a Pile Metaphor for

Email Management. Honours Thesis, School of ITEE,

University of Queensland, Australia.

Landry, B., Nair, R., Pousman, Z., Tungare, M., 2002.

TaskMinder: A Context- and User-Aware To-do List,

http://www.rahulnair.net/files/

taskminder_paper.pdf (accessed 30-12-2009).

Muller, M., Geyer, W., Brownholtz, B., Wilcox, E.,

Millen, D., 2004. One-Hundred Days in an Activity-

Centric Collaboration Environment based on Shared

Objects. In ACM Conference on Computer-Human

Interaction CHI, Vienna, Austria.

Naone, E., 2007. Software That Organizes Intelligently.

MIT, http://www.technologyreview.com/

communications/19788/?a=f (accessed 12-04-2010).

Rath, A., Weber, N., Kröll, M., Granitzer, M., Dietzel, O.,

Lindstaedt, S., 2008. Context-Aware Knowledge

Services, In CHI Workshop on Personal Information

Management, Florence, Italy.

Rogers, E. (1983) Diffusion of innovations

, 3rd edn, The

Free Press, New York, USA.

Sauermann, L., Grimmes, G., Roth-Berghofer, T., 2008.

The Semantic Desktop as a Foundation for PIM

research. In CHI 2008 Workshop on Personal

Information Management, Florence, Italy.

Sow, D., Davis, J., Ebling, M., Misra, A., Bergman, L.,

2006. Uncovering the To-dos Hidden in your Inbox.

IBM Systems Journal, 45(4) pp 739-758.

Stumpf, S., Herlocker, J., 2006. TaskTracer: Enhancing

Personal Information Management through Machine

Learning, In SIGIR Workshop on Personal

Information Management, Seattle, WA.

Tagg, R. and Mahalingam, N. 2005. Improving Customer

Service and Reducing Administrative Overload with

Sender-assisted Message Categorization: Working

paper.

Tagg, R. and Beames, T., 2009 How can a Quantum

Improvement in Personal and Group Information

Management be Realized? In ICEIS Conference,

Milan, Italy.

Tagg, R., Gandhi P. and Srinivasan Kumaar, R.,2009.

Recognizing Work Priorities and Taskiness in

Incoming Messages, In European Conference on

Information Systems, Verona, Italy.

Venkatesh, V., Morris, M., Davis, G., Davis, F., 2003.

User Acceptance of IT: Toward a Unified View, MIS

Quarterly 27(3) pp. 425-478.

Whittaker, S., Bellotti, V., Gwizdka, J., 2006. Email in

Personal Information Management, Communications

of the ACM 49(1).

ICEIS 2010 - 12th International Conference on Enterprise Information Systems

142