HOW CAN A QUANTUM IMPROVEMENT IN PERSONAL AND

GROUP INFORMATION MANAGEMENT BE REALIZED?

Roger Tagg and Tamara Beames

School of Computer and Information Science, University of South Australia, Mawson Lakes SA 5095, Australia

Keywords: Personal Information Management, Collaborative Work Support, Information Overload.

Abstract: A number of authors have pointed out that the IT technology used by individuals and groups for general

information work support has not advanced very far in the last decade. At the same time the level of

information and cognitive overload on individuals has continued to rise. Research has been done in several

areas (e.g. text mining and categorization, email threads etc) but a significant improvement in everyday

tools has yet to be seen. This paper addresses the current problems, describes a range of issues that need to

be addressed, and discusses how it might be possible in the future to advance to a new level of IT support.

1 INTRODUCTION

(Whittaker, Bellotti and Gwizdka 2006) have

highlighted a current situation where some software

tools, which are primarily used as email clients, are

being pressed into service - somewhat beyond what

was originally envisioned – as full-function Personal

Information Management (PIM) systems. Although

products such as Microsoft Outlook do include

functions for Task and Calendar management, these

functions have yet to be as widely adopted as email.

Many users use separate tools for PIM and group

information management (GIM) functions such as

to-do lists, project management, workflow; they also

use separate software for both domain-specific and

general office applications. There appear to be few

intelligent bridges between these tools.

An often-expressed opinion is that PIM support

has not changed that radically in the last ten years.

Despite a number of predictions and announcements,

the impression is that PIM is still a piecemeal

practice. The reasons for this are unclear. One

reason could be that the state of the art in

recognizing context in free text is still too immature.

Another is that the big commercial vendors (e.g.

Microsoft, IBM/Lotus) are not prepared to risk a

failure in this market.

During the same time, however, the workload

pressure on individuals has continued to rise.

Globalization and greater competition for business

requires increases in both the rate and quality of

decision making. The pace of business change forces

more frequent re-learning of processes and tacit

knowledge. The increasingly litigious business

atmosphere forces organizations to not only have

procedures in place, but to have proof that they are

being followed. There is greater emphasis on

regulatory compliance, quality assurance, and

governance. These trends have often led to a culture

of “backside covering” –against being blamed for

when things go wrong. Economic trends are also

forcing organizations to operate with tighter budgets,

smaller headcounts – without any extra resources to

implement efficiency improvements. For many

users, these increasing demands and expectations

seem to have outpaced any benefits offered by

introducing IT.

This paper aims to review some of the most

critical shortcomings in current practice, and to

discuss a range of issues, both technical and

motivational, involved in overcoming these

shortcomings.. Unless these issues are addressed, it

seems unlikely that there will be a significant

enough advantage to justify users and their

organisations changing to a new PIM environment.

Section 2 of this paper describes three of the

most serious shortcomings of current PIM support

environments. Section 3 analyses a number of

dimensions that require addressing in any successful

revolution in practice. Section 4 suggests some

general conclusions, including how a quantum

improvement might be achieved.

136

Tagg R. and Beames T. (2009).

HOW CAN A QUANTUM IMPROVEMENT IN PERSONAL AND GROUP INFORMATION MANAGEMENT BE REALIZED?.

In Proceedings of the 11th International Conference on Enterprise Information Systems - Human-Computer Interaction, pages 136-141

DOI: 10.5220/0002155101360141

 SciTePress

2 CRITICAL SHORTCOMINGS

WITH CURRENT PRACTICE

The scope of Personal Information Management, as

described by Whittaker and others, is quite wide.

Messaging is not just email, but also “chat”, SMS,

fax – and even phone messages, letters and scribbled

notes. Task management is not just self-reminders

and emailed tasks, it can include tasks from a

workflow, project management or ERP system, or

“actions” recorded in the minutes of meetings,

which might be stored in shared folders. Personal

file management includes many types of files

generated by many different application programs;

some of which belong to a group in which the user

participates. Contact management needs to cover not

only people who have email addresses; some

contacts are shared among members of a group and

the prime responsibility for maintaining them could

be in, for example, a customer relations management

(CRM) system.

There are also some PIM functions not addressed

in the 2006 Whittaker paper. One is the less formal

world of “information scraps” (Bernstein et al 2001).

Another is “event notification”, where a user may

want to be informed when certain conditions arise

(Rifkin and Khare 1998). Also missing from PIM

support, not for want of effort, are tools that can

recognise meaning in text, whether for identifying

context or implied tasks.

A more general criticism is that today’s email-

based tools are not sufficiently activity-oriented;

they give precedence to what someone else wants us

to read, than to what we really need to do. This is

being addressed by the European DELOS project

(Catarci et al 2007) and TV-ACTA (Bellotti et al

2006).

Three particular symptoms of these

shortcomings are discussed below.

2.1 Information and Cognitive

Overload

Information overload has been widely discussed in

the literature, ranging from (Whittaker and Sidner

1996) to, more recently and comprehensively, by

(Klausegger et al 2007). In over 40 years in the

workplace, the first named author has noticed a

relentlessly creeping number of hours one is

expected to work per week, and a greater proportion

of working time being spent on non-core work and

attending to ever-expanding volumes of email.

(Kirsh, 2000) discusses “cognitive overload” as

the problem of frequently switching focus from one

job to another. Fragmentation, here defined as not

getting a clear run to get things done without

interruption, has also been recognized by

(Czerwinski, 2006 and Tungare et al, 2006).

The now ubiquitous availability of email itself

has sometimes been blamed for information

overload. Email misuse, such as spam, excessive

bulk emailing,, and use for conversations that would

be better done by chat, phone or face-to-face –

causes the number of messages to explode. In an

attempt to counteract this, practices such as

individuals declaring email bankruptcy (Fitzgerald

2004), and of companies setting email-free days

(Booher 2007), have arisen.

2.2 Difficulty of Bridging to Other

Software and User Interfaces

If a task arrives by email, a user often needs to

correlate some data attached to the task with data in

other systems, such as a shared database or

application system. An example might be that

someone asks via email for the status of five

students. The data is perhaps in a spreadsheet, a

database or the university student records

application.

Correlating data about the same “things” is

rarely a trivial matter in today’s computers. One has

to either use a separate tool to query the spreadsheet

or database, or one has to work through an

application, often one that only caters for manual

interaction through a web form. Wrapping such

applications as web services may ease this latter

problem. However these solutions tend to be specific

to the particular process the user is carrying out.

2.3 Squeezing the Time Remaining to

do the “Real” Work

The problems above, together with the increase in

non-core activities mentioned in the introduction,

mean that users have less time to do what they are

really employed to do, which is presumably to

contribute his or her intelligence, knowledge and

professional skills to further the organization’s

business objectives. The result can be failure or poor

quality in what the group or organization is trying to

achieve.

Many busy workers with administrative

responsibilities are primarily driven by what is in

their email inbox. A colleague recently was heard to

say – “if it isn’t in an email, I won’t get round to

doing it”. Using such a reactive work approach

based on one’s email inbox is less than ideal, since

HOW CAN A QUANTUM IMPROVEMENT IN PERSONAL AND GROUP INFORMATION MANAGEMENT BE

REALIZED?

137

the inbox gives a rather limited and unbalanced view

of the what are the important tasks within the user’s

current projects and duties.

3 THE DIMENSIONS OF THE

CHALLENGE

A mind map, drawn in FreeMind (FreeMind 2008),

is presented as Figure 1. The following sub-sections

address the major branches of this mind map,

proceeding clockwise from the top right.

3.1 Integrating Heterogeneous

Information Sources

A typical user receives information from many

sources and in different forms. Messages may arrive

in emails, IM chat sessions, SMS messages on a

mobile phone, paper or audio (e.g. face-to-face

communication, phone calls, recorded messages).

Likewise, relevant information could be stored in

shared folders (e.g. Sharepoint, Lotus), project

management systems, ERP systems or custom

databases.

Further systems, such as event handlers,

workflow systems and RSS, use rules applied to

incoming changes to stored information to send

notifications to the appropriate human participants.

The general problem today is that there is very

limited integration of this information, which throws

a heavy load on the human user. Data is stored in

different structures and in different formats – a

classic case is the .pst files used in MS Outlook. It is

possible to resolve this, with something like the

proposed MS Windows file store, (Grimes 2002).

However this has still not been released, – and

maybe never will.

3.2 Recognizing Context

As information arrives for an individual or group,

someone has to resolve the question “what is this all

about”. Traditionally this has been done either by the

recipient, or by a trained assistant. Context

recognition is particularly important in the case of

email, where inboxes contain a mixture of items on

Figure 1: Mind Map of the Issues to be Addressed for PIM/GIM Improvement.

server stora

managing

the change

forcing

senders

to fill

forms

marku

RSS

ERP s

stems

workflow s

stems

project managemen

stems

Sharepoint, Lotus

shared folders

custom databases

strin

matchin

natural language

anal

sis

hone

face to face

conce

t structure

textual clues

messaging

stems

shared folders

and databases

analyzing

individual

text documents

recognizing threads

and inter-document

links

work management

ractices

resentation st

les

applying external

task t

deadline and othe

date reco

nition

applying

taskiness clues

recognizing

contex

messa

e senders

erson names

organizations,

ects,

roducts

applying

external

ontolo

text minin

name

harvestin

integrating

different

information

sources

web forms

email forms

universal

rou

ecific

messa

e filin

attachment

handling

technical

switch ove

motivation to

switch ove

data sta

nchronization

collaboration

aspects

bridging to

their UIs

emulating their

look and feel

keeping

the system

working

effectivel

changing

etiquette

manual

com

uter assisted

with keyboard

shortcuts

task XML

code words

successfull

making the

change

systems

software,

OS, etc

user hardware

popular existing

PIM tools

storing and

retrievin

fitting with

compute

hardware and

software

environments

disconnectedness

differing UI and

presentation

capabilities

geographi

mobility and

disconnect-

edness

Lifting

effectiveness

of PIM

and GIM

by a

quantum

leap

publishing to the

rou

dele

ation

synchronous

collaboration

ooled work

ermissions

emails

SMS

er scanned

audio

instant messa

defining the

context structure

fitting with

user

preferences

in handling

the work mix

recognizing

need for

action

time needed and

available

rou

information available

(

or not

)

rou

volunteerin

culture

batching similar tasks

project etc grouping

deadline-based prioritization

iles

to-do lis

algorithms

process mining

monitor user

actions, prompt

for chan

computer

assisted learning

manual

res

onsibilit

sub-laptops,

Blackberries

smart phones,

iPhones

workstations with

large or multiple

screens

ICEIS 2009 - International Conference on Enterprise Information Systems

138

widely varying subjects. We identify three aspects to

the context recognition problem.

First there is the task of identifying what

possible contexts there are, both for an individual

and for the groups he or she participates in. These

contexts could form an ontology structure, which

could be adopted from outside, created manually, or

generated from analysis of documents previously

received. In addition, an associated set of textual

“clues” (including names, is needed to enable the

recognition process to deduce that a document is

relevant to a context.

The second aspect is that of analyzing

documents of incoming information, and

determining what context or contexts apply. This

requires matching of strings and possibly inference

using relationships in the context ontology.

A third aspect is the recognition that one

document is related to another (typically earlier)

document. A simple example is recognizing threads

in email conversations (Bellotti and Thornton

2006). There are also other relationships between

messages and documents, such as “attached to”,

“new version of”, “comment on” etc.

3.3 Recognizing when Action is

Required

Current messaging-based PIM systems are often

imprecise about what action a recipient is being

asked to carry out.

Currently there seems to be a low uptake – and

lack of standardization – of task management

facilities in PIM and groupware systems. Not many

users actually use the Outlook task facility, which

often requires a sender to attach a task to an email. A

task list, encompassing tasks arising from all

sources, might be a better primary view through

which the user is prompted, with lists of emails as a

secondary option displayed if required.

Converting free-form emails to “to-do” task

items does not appear to be a simple process (Tagg

et al 2009) 2008). Message senders use many

circumlocutions to request another human to do

something. Some level of formality, e.g. marking up

in a “Task XML”, or use of agreed code strings (e.g.

<deadline>”) might help.

Meetings and appointments are special cases of

things we need to take action on, namely to attend

them. Adoption is possibly more widespread than

for tasks, but depends on there being a critical mass

of competent users as well as software that supports

a calendar exchange standard.

3.4 Catering for Users’ Work Handling

Preferences

While many users tend by default to be driven by

their email inbox, this may well not be an effective

approach to handling one’s workload.

Alternative approaches to ordering a to-do list

include deadline-based prioritization; batching of

similar tasks; grouping by project, customer or

contract; grouping by the time needed to perform

(e.g. all tasks that I can do in less than 5 minutes);

grouping by whether the information needed is

readily available or whether it has to be rooted out;

and pooling of work that can be picked up by any

member of a team.

Lists and folder structures (in the style of email

inboxes) may not anyhow be a good metaphor. Piles

(Lam 2005) have been proposed as an alternative.

Dossiers or piles could show icons for associated

documents as well as for the messages, and indexes

of contents could be included.

Portals allow a user to tailor a combination of

visual components that present the information and

tools that the user needs to be regularly aware of.

There are developments for mobile devices, e.g.

ePortal (CRM 2009), but they are not yet widespread

for general PIM. More development is needed,

particularly in standard paradigms for creating,

adding and removing components that are to appear

on an individual’s portal.

3.5 Fitting with Hardware and

Software Environments

Any solution cannot avoid having to work within

existing operating systems and other systems

software, usually including databases and

communications. It may also have to work, at least

temporarily, with existing PIM software and

groupware, either by emulating their interfaces or by

bridging to the products themselves.

Because of user mobility, any software also has

to be implementable on a range of portable

computing devices ranging from laptops to mobile

phones. (Grundy et al 2002) proposed a form of

thin-client groupware oriented to this situation.

One particular issue, already mentioned in 3.1, is

to the need for a much more integrated file system,

covering – in a consistent way – the tagging, storing

and retrieval of messages, attachments, contacts,

documents, databases and simple files. Google

Desktop and other similar search tools have offered

a search facility that caters for files as different as

.doc and .pst, but generality to cover networked

HOW CAN A QUANTUM IMPROVEMENT IN PERSONAL AND GROUP INFORMATION MANAGEMENT BE

REALIZED?

139

storage locations and plug-in devices are not as

advanced yet.

3.6 Allowing for Geographic Mobility

and Disconnection

Information workers today are becoming

increasingly unlikely to stay tied to a single desk,

and their networking is often wireless. They use

different computing devices at different times and

places, so they cannot use tools where working data

is stored on a particular client device, unless they

carry that device everywhere. Storage of information

on a server therefore seems unavoidable, both for

back up and for overflow. Synchronizing after a

period of disconnection is also an issue.

Automatic indexing of documents (as in e.g.

Google Desktop) needs to be extended to remote

storage devices, and synchronization between the

different computers becomes critical.

3.7 Successfully Making the Change

However good a new PIM/GIM solution may be, the

question will arise: “what will it take for a critical

mass of users to change to it from their current

toolkits?” One can usefully look at quantum leaps in

the past, such as spreadsheets, databases, B2C

electronic commerce etc. One motivation would be a

clearly demonstrable advantage to users, or to bring

some new or previously minor niche functionality to

the forefront.

However the power of fashions and of “follow

my leader” should not be underestimated. If a large

organization takes on a new system, or an

outstanding public speaker or computer journalist

preaches the cause, enthusiasm can become

widespread. This power can also be seen from an

individual user point of view when there are tools

that require all collaborating users to use them (e.g.

Facebook).

It is possible that as well as a major software

improvement, a change in human-to-human etiquette

will be needed. Current culture is to send

information to other users without any contextual

clues for the receiver. The acceptability of totally

free-form emails may have to be reduced for the

sake of reducing overload. Use of pre-formatted

email forms (as in e.g. an Outlook Forms library),

could be prescribed within a group or organization.

Senders not following guidelines could be warned

that their messages would receive lower priority.

Powerful organizations, or government, can

already enforce their customers to use “Contact Us”

web pages instead of email, which shows a distinct

possibility for form-based emails to be accepted in

practice. However, in a survey carried out in 2006

by members of our research group, one respondent

declared that users he knew would never accept

forcing small groups or individuals to adopt this

approach. Perhaps if users did become accustomed

to a form-based system, with space for at least some

free-form text to allow for language expression

(Guerin 2004), acceptance would be higher.

3.8 Keeping the System Working

Effectively

As has been shown with spam filtering, patterns of

communication evolve. Peoples’ jobs, and the

balance of an organization’s business, can quickly

change. If humans are stressed, they may not have

the time to revise their ontologies, especially at

moments when they most need to do so.

Automatic learning systems, driven by artificial

intelligence algorithms, seem desirable, but users

need to have confidence in them. It would not be

acceptable to users if the category structure was

changed too often or too drastically. It might be

better to use an agent that can monitor the efficiency

of the context recognition tools over time and warn

the user of the need to make manual changes.

Process mining – the technique of detecting patterns

in related operations, messages and documents,

could be a useful way of providing such feedback.

3.9 Handling the Collaboration Aspects

Collaborative work brings up a number of needs,

including synchronous collaboration, pooled work

(see 3.3), publishing information to the group,

delegation and permissions to access data.

It is assumed that one user may be involved in

many groups, each of which may have their own

context structure and preferred practices. So there is

a need for any new system to be able to

simultaneously link to more than one group work

support system.

4 CONCLUSIONS

There have been a number of studies attempting to

find improvements in PIM and GIM, e.g. (PIM

2008). The reality seems to be that a piecewise

approach will not be enough to trigger the quantum

leap needed, especially when considering the vast

number and range of custom-made PIM tools and

ICEIS 2009 - International Conference on Enterprise Information Systems

140

add-ons (e.g. to Outlook) already in existence. At

least 3 or 4 of the issues raised in section 3 above

would seem to be required, in order for things to

start happening, and even then one would need to

consider how these solutions would be adopted and

migrated to in practice.

This is an area where a multidisciplinary

approach seems essential. No single researcher, or

academic research group, is likely to have the time

or resources to build a complete enough toolbox. A

well managed consortium, with involvement from

commercial interests, might do better.

Otherwise we will all have to wait for Microsoft

or IBM to come up with and implement an answer.

The need for such improvements has been discussed

for at least 5 years. We can wonder why we have

seen so few signs of a solution so far. Possible

causes are that a) other issues have been more

pressing (e.g. spam and information security); b) that

the state of the art in recognising context and tasks is

not yet good enough to give reliable results; and c)

introducing such a wide range of interconnected

software is so risky that a commercial organization

cannot afford a spectacular failure.

So, the opportunity might be there for an

entrepreneur (remember the original Bill Gates?) - or

an academic consortium less hamstrung by the

constraints of KPIs and research quality

assessments.

REFERENCES

Bellotti V. and Thornton J. (2006), Managing Activities

with TV-Acta: TaskVista and Activity-Centered Task

Assistant. In Proceedings of the Second SIGIR

Workshop on Personal Information Management

(PIM).

Bernstein, M., van Kleek, M., Karger, D. and Schraefel,

MC., (2001), Information Scraps: How and Why

Information Eludes our Personal Information

Management Tools, ACM Transactions on Office

Information Systems.

Booher, D. (2007), E-mail: Could You Live Without It?

Booher Banter, http://www.booher.com/

wordpress/?p=282.

Catarci, T., Dix, A., Katifori, A., Lepouras G. And Poggi

A. (2007), Task-centered Information Management, In

DELOS Conference on Digital Libraries., Pisa, Italy

CRM (2009), Ebridge Software Eportal Now Compatible

with Blackberry Storm, Blackberry Bold and Apple

iPhone, 12 January, http://crmhelpdesksoftware.com/

index.php?s=erp

Czerwinski, M. (2006), From Scatterbrained to Focused:

UI Support for Today’s Crazed Information Worker.

In SIGIR Workshop on Personal Information

Management, Seattle.

Fitzgerald, M. (2004), Call It the Dead E-Mail Office,

Wired magazine, 7 June.

FreeMind (2008), Free Mind Mapping Software, Wiki,

http://freemind.sourceforge.net/wiki/index.php/Main_

Page

Grimes, R. (2002) Revolutionary File Storage System Lets

Users Search and Manage Files Based on Content,

http://msdn.microsoft.com/en-

us/magazine/cc164028.aspx

Guerin, B. (2004), Handbook for Analyzing the Social

Strategies of Everyday Life, Context Press, USA,

Chapter 5.

Grundy, J., Wang, X. and Hosking, J. (2002), Building

Multi-device, Component-based, Thin-client

Groupware,, In Proc Australasian User Interfaces

Conference, Melbourne, Australia,

http://www.cs.auckland.ac.nz/~johng/papers/auic2002.

pdf

Kirsh, D. (2000), A Few Thoughts on Cognitive Overload,

Intellectica.

Klausegger, C., Sinkovics, R. and Zou, H. (2007),

Information overload: a Cross-national Investigation

of Influence Factors and Effects, Emerald Insight,

www.emeraldinsight.com/0263-4503.htm

Lam, B. (2005), MailStacker: Applying a Pile Metaphor

for Email Management, Honours Thesis, School of

Information Technology and Electrical Engineering,

University of Queensland.

PIM (2008), CHI Conference Workshop on Personal

Information Management, Florence, Italy

Rifkin A. and Khare, R. (1998) The Evolution of Internet-

Scale Event Notification Services: Past, Present, and

Future, http://www.ics.uci.edu/~rohit/wacc

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

H. (2009) Recognizing Work Priorities and Taskiness

in Incoming Messages by using Personal Ontologies

with Lexical Clues, submitted for publication)

Tungare, M., Pyla, P., Sampat, M. and Perez-Quiñones,

M. (2006), Defragmenting Information using the

Syncables Framework. In SIGIR Workshop on

Personal Information Management, Seattle.

Whittaker S. and Sidner C. (1996), Email Overload:

Exploring Personal Information Management of

Email. In ACM Conference on Computer-Human

Interaction.

Whittaker, S., Bellotti, V. and Gwizdka, J. (2006), Email

in Personal Information Management,

Communications of the ACM (49) 1.

HOW CAN A QUANTUM IMPROVEMENT IN PERSONAL AND GROUP INFORMATION MANAGEMENT BE

REALIZED?

141