Content uploaded by Thomas Olsson
Author content
All content in this area was uploaded by Thomas Olsson on Aug 28, 2014
Content may be subject to copyright.
User-Centered Design of a Mobile Application for Sharing
Life Memories
Thomas Olsson1, Marika Lehtonen1, Dana Pavel2, Kaisa Väänänen-Vainio-Mattila1
1 Institute of Human-Centered Technology,
Tampere University of Technology
P.O Box 599, 33101 Tampere, Finland
{thomas.olsson, marika.lehtonen,
kaisa.vaananen-vainio-mattila}@tut.fi
2 Nokia Research Center, Helsinki
Itämerenkatu 11-13
00120 Helsinki, Finland
dana.pavel@nokia.com
ABSTRACT
People have an inherent need to capture and collect life memories
such as moments with children or special events with friends.
Capturing life memories is either spontaneous or planned.
Memories are stored and frequently shared with other people.
New internet services allow online sharing of photographs but
they bypass the mobile aspect in capturing and sharing
multimedia. In this paper, we present a concept mobile
application for capturing and sharing memories while mobile and
the user research studies that supported its design. Our aim was to
design a prototype application that would enable mobile users to
capture and share precious moments. Various categorizations of
user behaviors related to personal media management emerged
out of our user needs studies. We then followed a user-centered
design process according to Contextual Design. The user
evaluation of the resulted paper prototype showed that users
appreciate the event based, instant sharing within particular
groups of people.
Categories and Subject Descriptors
H.3.5 Online Information Services: Data sharing. H.5.2 User
Interfaces: Prototyping, User-Centered Design. H5.m.
Information interfaces and presentation (e.g., HCI):
Miscellaneous.
General Terms
Human Factors, Design
Keywords
Life memories, mobile application, personal content management,
Contextual Design, user-centered design
1. INTRODUCTION
People have an inherent need to capture and collect memories
throughout their life span. Life memories can be seen as both
people’s internal memories, but also as physical and digital
artifacts relating to meaningful events. Memory artifacts have
traditionally been captured by taking photographs and storing
them into photo albums, perhaps with some textual information
attached. In addition, letters and mementos from special events
such as family gatherings, concerts, meetings can be stored for
later viewing. Personal life memories are traditionally shared to a
limited extent, mostly by viewing the artifacts together or by
sending them to friends via email or MMS. More recently,
internet services, such as Flickr or Picasa, have enabled people to
share photographs online. These and other similar applications
enable new forms of user-generated content management.
Mobile technology is suitable mostly for capturing the media –
useful mobile sharing tools have not emerged at the same pace.
However, mobile phone is a versatile platform in personal content
creation and management since capturing, managing and sharing
the data can be done instantly and using only one device. In
addition, mobile phones offer natural opportunities for collecting
instant digital pictures and videos because of their immediate
availability to users. Easy access to data on the phone or around it
enables easy metadata annotation of captured content, which can
greatly improve the management of the personal content [2].
Our research work is part of the SharMe exploration activities
from Nokia Research Center. The focus of our research was on
finding solutions for how mobile phones could enable people to
easier capture and share precious memories so that they can focus
more on living the moment than caring about technical problems.
We started out from Vannevar Bush’s vision of continuous
capturing [3] and tried to identify – through interviews,
observations and interactions with users – what kind of situations,
interaction and application paradigms people would find
acceptable and desirable to be captured.
As extensive previous work shows (e.g., [4][5]), any type of
media capturing with mobile phones that involves other people’s
presence or future sharing is extremely sensitive. In our work, we
conducted in-depth user needs research with different user groups.
Existing applications involving mobile phone-based capturing
have been evaluated. Usage of existing applications has been
observed. All these eventually lead to the creation of a concept
prototype that has been further designed and tested with users.
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee. The design of the concept utilized the Contextual Design method
[1] to ensure the user-centeredness of the developed concept
Mobility’07, September 10 -12, 2007, Singapore, Singapore.
Copyright 2007 ACM 1-58113-000-0/00/0004…$5.00.
prototype. The challenge was in designing a visionary concept
prototype when no existing, similar task processes for mobile
memory sharing could be observed. Instead, existing sub-
processes of photograph capture and sharing were used as a basis
for Contextual Design.
Section 2 describes related work on personal content management
on mobile devices. Following that, Section 3 describes the user-
centered design process. Section 4 presents a selection of user
needs research findings. The resulting concept prototype and user
evaluation results are presented in Section 5. Section 6 discusses
the findings from the prototype acceptability and design process
perspectives. Finally, Section 7 presents a summary and
conclusions regarding the research and design work.
2. RELATED WORK
This research involves several intriguing aspects to be taken into
account, for example mobile capturing and sharing, metadata
annotations, context awareness and continuous capturing.
Previous studies have shown that users’ needs for capturing,
managing and sharing memories are universal but the users’
habits seem very diverse. The photos and other media are
captured for both personal use and for sharing them with others.
Users share their recordings to maintain the social relationships,
for self-expression and to construct personal and group memory.
Mobile sharing is usually instant and for very small group of
recipients. Sharing on PC and using web sharing tools is more
public. PC sharing is addressed to smaller groups by using e-mail
or peer-to-peer networks. [6][7]
The idea of continuously recording one’s life memories has been
around for quite some time. The initial vision was developed by
Vannevar Bush [1]. Notable research activities have been going
on since then in various groups, e.g. [7][9]. While the topic still
generates various controversies either due to gadgets that need to
be worn or due to ethical and legal issues involved, recording the
most precious, but usually missed moments, interests users [4][5].
While continuous capturing has certain advantages, allowing
people to focus on experiencing the moment with little or no
interaction with the capturing device, the opinions are split when
it comes to losing privacy and control over what and when is
recorded. While certain individuals are more willing to give up
this control [10], most of the people are not, as also shown in our
studies.
Personal media annotation has recently been extended to mobile
phones [2], due to their capacity to collect various internal and
external data. The amount and type of metadata that can be
attached to digital media is very diverse, including context data
[11][12], user-set tags, descriptions of the media content, etc. The
annotations can be either automatic or user input. Recent mobile
phone applications and platforms (e.g. ContextPhone, Meaning
and Jaiku) have been taking advantage of the availability of this
data in phones. One of the most important context information,
location, is already being used both in prototype systems [13][14]
and in commercial phone-based navigation software (e.g. Nokia’s
Smart2go).
Context awareness has for long been a rising topic in mobile
computing research [11][12]. The possibilities of mobile device to
automatically capture various context data have affected on
emerging of context aware mobile applications, such as Jaiku and
Context watcher. In personal content management, adding context
information to recorded media makes it easier for the user to
remember the real life event more evocatively [15]. It can also
enable easier management and search of personal content.
Experimental mobile applications such as Mobile Media Metadata
(MMM2) [16] and Meaning have enabled new forms of mobile
instant sharing. MMM2 introduces a method for predicting the
recipients of a to-be-shared photo by identifying the near-by
Bluetooth devices at the time of photo capture. Meaning supports
using an external GPS device for locating. The location
information can be added to the captured photo as raw location
coordinates or as semantic location tags (e.g. home, work,
downtown). Using a web-based portal to browse and post process
the uploaded data utilizes the benefits of the PC user interface and
provides good accessibility to the memories.
In the field of web-based personal content management
applications the growth has been remarkable for the last few
years. Web2.0 applications, such as Flickr, Picasa and YouTube,
have offered services for organizing and sharing personal media
files. In the research presented in this paper, the aim was to
develop – based on the user needs – a set of versatile features for
mobile memory capture and sharing.
3. USER-CENTERED DESIGN PROCESS
As the memory sharing is a very sensitive area of user activity,
we first conducted an in-depth user needs research study. Our
user-centered design (UCD) and research process then involved a
slightly modified version of Contextual Design method (CD) [1].
The main challenge in applying CD was the fact that mobile
memory management and sharing contains tasks that do not exist
in mobile applications currently. Thus, related sub-tasks of mobile
photography were taken as a basis for design.
3.1 Study Goals
The main objective of the research was to bring the end-users’
perspective to the development of the SharMe concept of life
memory management. A wide scale of user-centered design and
research methods was used during the project. Users were
involved in all phases of the design: user needs research,
contextual inquiries, paper prototype tests in two rounds and
acceptability tests for the designed prototype system.
Contextual inquiries aimed at revealing the selected target users’
current behaviour with mobile photography. Contextual inquiries
would result models of the current behaviour, comprehension
about the division of tasks between mobile and PC usage and
users’ opinions (benefits and flaws) about the currently used
capturing and sharing applications. A list of both universal and
case-specific design drivers was to be generated by interpreting
the inquiries’ results. These results would then be used to design a
skeleton of the concept prototype. The user needs results should
also reveal issues that could be utilised in the creation of further
memory management applications.
The most concrete goal was to develop a prototype of the SharMe
concept according to the user studies. Paper prototype testing was
aimed at iterating the first versions of the prototype and its UI
towards a more usable direction. Additionally, paper prototyping
was used to gather users’ opinions regarding the acceptability of
the current prototype. Finally, an important goal was to evaluate
the resulting concept prototype with the target users.
3.2 User Needs Research
The aim of the user needs research was to tackle wide range of
questions dealing with user values, expectations and emerging
needs in relation to the overall SharMe concept and in specific to
the processes of capturing, retrieving and sharing life memories.
Interviews and scenarios were chosen as the research methods to
elicit insightful discussion with the potential target users.
Specific challenges in planning the interviews dealt with the
decisions between potential target groups and themes of the
interviews. Difficulties were due to wide-ranging SharMe concept
and unclearness of the target group of such a novel concept. Also,
planning the usage scenarios to be used in the interviews was
demanding, since there is no current system which could have
been used as a starting point in the scenarios.
The findings of the interviews and scenarios were used as a
background for user-centered design. Interviews were carried out
in two consecutive rounds with partly different sets of interview
questions and variety of target groups. Altogether 54 persons
participated in the interviews.
In the first interview round, qualitative thematic interview was
selected as a suitable research method. Two of the nine interviews
were focus group interviews and the rest were individual or
couple interviews. To get an overall viewpoint of potential users’
opinions to life memory management, participants were selected
to represent three user groups: 1) Young families with babies or
young children, 2) Young people familiar with on-line
communities and 3) People living away from their relatives (in
another country).
In the second interview round the methodology included
evaluation of the usage scenarios. Based on the findings from the
first round, 11 scenarios were written to describe different types
of use situations and aspects of SharMe (see Figure 1 for an
example). There were no pictures or other illustrations in these
textual scenarios. Second round was composed of five focus
group interviews with the following user groups: 1) the parents, 2)
the elderly, 3) the travelers, 4) the enthusiasts (car tuners) and 5)
the athletes. Overall, this broad selection of potential user groups
gave a useful overview of different needs and expectations
towards SharMe concept.
Vera has just settled comfortably on the hotel room bed in Barcelona and
browses with her mobile device through all memories saved during
the day. Especially the visit to the church “La Sagrada Familia” was really
exciting and there is a lot of good material on that. Also, the
automatically saved official presentations of the city seem to be useful.
It’s easier to tell about the holiday for your friends when there are so much
valuable details saved in the system. Vera decides to authorize sharing
these memories only for her friends, although there is nothing secret
here. Perhaps she is just concerned about anonymous troublemakers
who want to play jokes on the internet. She notices that Erika is already
adding some comments on the files. [...]
Figure 1. Excerpt of an example usage scenario used in user
needs research (central topics bolded for emphasis)
3.3 Contextual Design
After the user needs study the Contextual Design phase followed.
The target group of research at this stage was chosen to be
youngish people who already use some mobile devices and
applications for personal content management. This user group
was chosen as the target group for the design phase, because they
were assumed to be able to give insightful comments about the
novel ideas of SharMe. They were also assumed to be the first
real consumer groups of such service. Altogether 10 contextual
inquiries were carried out. Eight people of this group were male,
two female, and the ages varied from 25 to 35. The study had two
sub user groups: Flickr users and users of modern camera phones.
As there are no corresponding applications to the SharMe
concept, profound interpretation was needed for the inquiry
results. The inquiries took place at the user’s normal usage
environment in order to maximize the naturalness of the usage
situation.
The contextual inquiries focused on finding out the usage patterns
and sequences of SharMe concept. The specific focus was on
revealing when, where, with whom, in what kind of situations and
how users capture or share recordings. The contexts of use and
usage of contextual information were other important focus areas.
While the focus was on mobile use, the study also aimed at
finding what features that are currently in the PC UI would be
needed in the mobile UI. Furthermore, the wide range of functions
in personal content management related to creating, browsing and
sharing the memory objects was observed and discussed with the
study participants.
After inquiries the interpreted data was reclassified on an affinity
wall and consolidated sequence and interaction models were
drawn. However, the cohesion between inquiries and user needs
interviews had to be confirmed. An integrated affinity wall was
built to show the consolidated research material, not just the
contextual inquiries. The notes were broadly at same level of
abstraction and they represented user needs, habits or opinions of
the participants. The resulted integrated affinity was used to
create design drivers to steer the prototype design.
The resulted affinity wall along with the interpreted models let us
form design drivers to steer the prototype design. Totally 17
drivers with various focus areas were written. The drivers served
as an idea source when designing the prototype outline and as a
check list during the design iteration of the prototype. Figure 2
shows examples of the design drivers.
1) Provide features for creating collages from several users’
recordings.
2) Provide easy ways to organize and browse the material
according to time, event and place.
3) Let user control what kind of metadata is added and shared.
4) Provide features for browsing the most recent captures.
Figure 2. Examples of design drivers created based on the
affinity wall
3.4 Paper Prototype Testing and Prototype
Acceptability
Based on the Contextual Design results, a paper prototype of the
central functions of the SharMe service was created. The
prototype is presented in Section 5, Resulting Prototype. Paper
prototype testing was carried out in two rounds, the first round
with seven participants and the second round with five. The first
round of prototyping aimed at getting user’s opinions on the
general idea of the system, and specific parts of the user interface
design. This revealed the biggest flaws and produced
improvements on the prototype user interface.
The design process was very iterative: after every test the most
obvious flaws in the UI were identified and the related screens
were redesigned. However, altering any fundamental part of the
prototype was changed only based on input from several users.
When designing a new prototype of a futuristic concept it proved
to be useful to perform paper prototype tests that allowed frequent
iteration. Most of the unwanted or unclear design solutions were
discovered at an early stage and the main functionalities and
screen layout were redesigned immediately.
The second round of paper prototype tests was done with much
more detailed mock-up screen and menu images. The mock-up
images were represented with example use cases. In this round the
focus was also on preliminary user acceptance of both the
prototype in general and certain parts of it. In this phase, some
new and interesting design features were still invented. Users
offered us valuable insight on what they regard as troublesome in
current mobile user interfaces. These facts proved the significance
of paper prototype testing in our study.
4. USER NEEDS FOR LIFE MEMORY
MANAGEMENT
Extensive user needs studies were carried out in order to find out
users’ expectations and needs in relation to SharMe system.
Attention was also paid to those factors which might act as
obstacles for usage of such system.
4.1 Needs for Saving Memories
Users have generally positive outlook of the overall concept of
SharMe. It is important for the users that SharMe can be used
with a mobile device, while in certain situations other devices
would be used as well when, for instance, larger screen is
considered necessary. It is definite that users have a need for
versatile memory saving devices, the kind that will make it easy
to collect life mementos, manage them and share them with
others.
Based on the interviews, it was found out that needs for saving
can be divided in three separate categories on the basis of the type
of collected memory data (see Table 1). In addition, each category
represents either long term or short term memory. This division
makes visible the users’ needs for saving memories either for
good or at least for a long period of time or for shorter time
corresponding to such memory tools as calendar, note book or
PDA.
Table 1. General needs for saving memory items
Types of memories users
want to save Examples of memories
1. Happy moments
(related to long term
memory)
Family and friends, nice events
(birthdays, weddings etc.), related
context information etc.
2. Life history (long term
memory) Important events of personal life,
travel mementos, development
steps of children, related context
information etc.
3. Details to facilitate
daily errands (short term
memory)
Calendar information, checklists,
medical history etc.
To sum up, users are interested in saving data which captures
moments of some significance. However, needs for saving does
not rise only from users’ personal interest. For instance, elderly
people could see the value of the savings from the perspective of
future generations. Also parents wanted to save memories of their
children to be shared with them when they grow up. Thus, needs
for saving are often interconnected with the idea of sharing and
persistence of the memory data.
Users also emphasized that they would like to capture life
moments which usually pass by so quickly that nobody is able to
save them.
4.2 Levels of Sharing Memories
The idea of sharing memory data with others is very appealing for
users who have realized that currently the technology in use
restricts sharing in many occasions. For instance, large amounts
of data cannot be sent via e-mail. Moreover, using web-based
platforms for sharing is not always easy or suitable for given
situation. Yet, users are interested in sharing visual and audio
presentations of their memories. Simultaneously, communication
culture is becoming increasingly visual by its nature, which
emphasizes the importance of making it easier to share visual
materials.
Even though sharing of memories was considered to be very
positive by participants of the interviews, this does not mean that
everybody wants to share all their memory data with anybody.
Instead users want to be in control over granting the rights for
viewing or otherwise utilize their saved materials. Thus, granting
the rights for others must be carried into effect very easily.
Table 2 presents an interpretation of the categories of sharing
based on the interviews. As seen from the table, the idea of
sharing divides the users. While some are ready to share all or at
least a lot of their memory data, others create memory materials
mainly for their personal use. It seems, however, that majority of
users are in between of these two extremes. But without statistical
study this cannot be stated with certainty. Nevertheless, based on
our qualitative interviews and analysis, it is obvious that the
context of sharing and content of the memory data are the
defining factors, when users consider with whom to share their
data. Generally speaking users are more interested in sharing
happy memories, while mementos of negative life events are more
often considered to be private.
When sharing their memory items users are often interested in
having comments from others. They want to hear other people’s
opinions on their saved materials or to have other participants to
add their own perception of the event. This collectiveness in
commenting others data shows us that memories are often created
together. It can be done through discussion on past events, but
nowadays reminiscing can also be carried out virtually, via an
internet or GSM application.
Table 2. Three levels of sharing memories
Willingness
to share Perception of
memories Explanation
of ways of
action
Motivation for
sharing/ not
sharing
1) No
sharing at
all
“Private
memories”
Memories are
private. Others
are probably
not even
interested in
them. Sharing
them might
give others an
opportunity to
harm me.
I save
memories
mainly for
personal use.
Saved items
provide means
to recollect,
contemplate
and develop
myself.
I enjoy
recollecting
happy moments,
but I can also
relieve
unpleasant
moments of my
life through
memory data.
2) Sharing
only with
my family,
closest
friends or
people
presenting
the event
“Sharing
creates
intimacy”
I share my
saved
memories with
the people I
can trust. Other
people might
do me harm, if
they had
access to my
memories.
Sharing
strengthens our
relationships
with each
other. It creates
feeling of
community
and belonging.
I like to make
others happy by
showing them
pleasant things. I
enjoy sharing
positive
memories.
3) Sharing
with anyone
“Creating
new
communities
and shared
memories”
I have nothing
to hide. Free sharing of
all the savings
is a must in
Internet age. It
is a part of new
culture.
I like to get
comments from
others about my
saved memories
and myself. I
enjoy getting
attention.
4.3 Timing of Sharing
One of the interesting themes is the timing of sharing. When
sharing is carried out via mobile phone, e-mail or other internet
applications, it does not always happen in real time, which means
that receiver does not necessary view materials immediately after
having received them. However, users are quite interested in
instant sharing in order to be able to give instantly additional
information on the data, if necessary. They would also be eager to
have receiver’s comments on the data right away after the
mementos have been sent. Obviously there are situations when
real time sharing is not required or even possible, but when it is, it
creates a feeling of intimacy and closeness between users of the
system. Of the different user groups especially travelers and
people living away from their relatives (in different countries)
found real time sharing adding something new to their way of
communicating with the people close to them.
4.4 Event-based Management
Both interview rounds pointed out that users want to record
especially happy moments, trips, parties and other interesting
events. In social gatherings the conversation easily drifts to
reminiscing the past together experienced events. In other words,
users reminisce about the memories as entities. This is
particularly emphasized in long-term memories.
As studying the users’ habits in the contextual inquiries it came
out, that users tend to organize their media files according to real
life events’ occurrence time. Users often capture photos in bursts.
The ordinary everyday life is little recorded but when the users
are involved in something interesting, abundant recordings are
taken. Therefore, it is natural to organize the recordings in entities
according to real life events. One burst of recordings can easily be
identified as one event.
An interesting event is usually something done with a group of
other people (e.g. hobbies, family vacations or public events). The
recordings are meant to be shared with that group or even
publicly. Inquiries showed that groups with close relationships,
such as family or a group of friends, have need for gathering
collective events. The group usually participates together in an
event, and the recordings from different devices are then wished
to be accessible for the whole group. Secondly, if only few
members participate in an event, the rest of the group wants to
view the recordings.
The group members’ personal recordings from the same event
might differ a lot from each other. Thus, the collective memory
becomes a dynamic collage of different private memories. Even
small pieces of personal recordings shared to others might change
others’ mental impression of the event. Users manifested also that
it is interesting to keep both the private and the collective version
of the event. This way it is possible to reminisce the event from
only personal point of view or by using all the collective data.
This makes the concept of collective memory so intriguing.
5. RESULTING PROTOTYPE
The results of the UCD process yielded a prototype of the SharMe
concept. The prototype’s outline is first described briefly. Then,
we describe the two most interesting features in the prototype,
which are also thought to be the most novel.
5.1 Prototype Outline
The resulted prototype is an application for capturing memories,
sharing them, and automatically organizing them with both
mobile phone and PC. The designed part of the system is to be
used with a mobile phone. The designed prototype could be
thought as a replacement of modern mobile phone’s camera and
gallery softwares. Figure 3 describes the prototype’s main
functionalities as a whole system.
Figure 3. The prototype outline. The prototype focused on the
Mobile Phone user interface.
As in most related systems, the media content is stored in the
Internet, accessible via both mobile and PC UIs of the prototype.
Collective use of the media content is thus emphasized. User can
easily share content by forming events of the media and sharing
the event to certain user or groups of users. Event creation and its
effects on application’s other functionalities is more exactly
described in Section 5.2. Capturing the media includes also other
features, like context data added to the memories. Attaching other
than mere visual or audio data makes the memory more alive.
Moreover, users may add their own tags to the memory items and
annotate them in order to enrich the shared memory. Browsing the
recordings is more versatile: user can browse the media by
different browsing rules and view only part of the viewable media
by filtering the content. Adding various types of metadata enables
versatile search features. In the contextual inquiries users
emphasized the need for quickly viewing the most recent
captures. This was taken into account by offering a quick route to
the gallery of the most recent captures.
5.2 Collective Events for Instant Sharing
The results of the user needs research and contextual inquiries
lead us into creating an event-based capturing feature. The event
is thought to contain all the material relevant to the memory of the
corresponding real-life event. As the modern mobile phones are
capable of capturing diverse metadata and PC platform is
excellent for creating versatile data types, creating events to
integrate all the data is rather straightforward. This idea of event-
based content management is a fundamental thought in the design.
Using it affects rather much of the functionality of the rest of the
prototype.
As the event can be thought as a certain period of the user’s time
that has common characteristic, various settings can be given to
the event. A user can create an event before the actual capturing
event takes place, during it or even afterwards. The recordings
captured during the duration of the event are automatically
organized under the event. Several options to define the duration
of the event are provided in order to support flexible management
of the events (see Figure 4).
Figure 4. Left: the event creation wizard, Middle: the
submenu under Use preset settings, Right: the submenus
under Event starts, Event stops and Invite users.
Creating events makes it easier to share captured media to
different groups. User can form a group of users to be invited to
the event. Every invited user is allowed to view others’ recordings
and share their own. Moreover, different users’ memories from
the same real-life event can be merged as collective events.
During the event, sharing is either automatic by sharing all the
recorded material or manual, in which case the user chooses the
sharable recordings any time after capturing. As the recorded
memories are usually for collective use, group control,
commenting and discussing features are provided. This helps
collecting all the event data into one collective memory and
annotating it. This feature was designed for close groups that have
mutual trust. Thus, only 2 different access right levels are needed
(the event’s creator and the invited users). Tools for public
sharing exist already (e.g. Flickr) and they can be utilized in the
prototype’s PC user interface.
Event-based automatic organizing makes it easier to browse
media in natural entities. It is vital to support browsing the media
from different viewing levels, because browsing can be done
either for close viewing the recordings or quick searching for a
certain recording. The filtering options ease finding desired
recordings. Sorting options provide alternatives how to sort the
content (e.g. time, name). Furthermore, according to contextual
inquiries organizing content based on user set tags, time or place
is most natural for users. In the prototype the user may add own
tags to single recordings or to all the recordings in the event.
Event tags added after the moment of capturing can however be
added to previous recordings in the event, too.
Event based thinking in the paper prototype tests elicited an idea
of sub events. Users agreed that one sub level under each event
would be sufficient. For example, a weeklong vacation could
include several lesser events that should be located under the
vacation event and, still, form an event of its own with individual
settings. The sub events’ size would be more suitable for one-time
sharing, because users do not want to view too much at one time.
Finally, the feature of sub events was designed to be based on
user set continuous tags.
5.3 Sharing through Continuous Capturing
Continuous capturing was one of the features explored in our
concept. The aim was to design a system that can minimize the
lost important moments and that allows users to focus on
experiencing the moment rather than dealing with a device. This
drove us into designing features that will allow for user-controlled
continuous recording.
The capturing solution presented in Section 5.3.1 was designed
first. Iterative paper prototype tests, however, brought out very
diverse aspects on how the continuous capturing could be
implemented. The solution of continuous context data (5.3.2) is a
part of the resulting prototype and was tested with users in the
second user evaluation round (see Section 6).
The media captured continuously is video, audio or merely
context data. Truly continuous video recording is limited by the
fact that the camera is not wearable or convenient to be aimed at
the target all the time [4], but this shortcoming could easily be
addressed by using a wearable camera.
5.3.1 User Initiated Buffering
Here, buffering is a capture mode where the user can set the
device to record video or audio into a buffer of selected length.
When something interesting occurs the user may save the media
in the buffer – or a part of it – and continue the recording in video
mode after that. Figure 5 demonstrates the buffering user
interface.
Figure 5. Left: the user initiated buffering in stopped mode.
Right: buffering running and the quick menu options
The length of the buffer defines how much of the past can be
saved permanently. If the buffered material is not saved, it will be
overwritten as the buffering goes along. This solution is most
suitable for predetermined situations where only part of the
recording is important, but unforeseen (e.g. a sports event). Users
thought this model to be the lightest and most ethical, but at the
same time inefficient for saving the usually missed moments. The
system would greatly benefit from having hard buttons on the
device that would allow the user to easily save, start and stop any
recording.
5.3.2 Continuous Context Data
One way to record data continuously and more ethically is to
record mere context data without any audio or visual media.
In the prototype, recording context data (e.g. location, weather,
near-by devices) during the event can be either truly continuous
throughout the whole event or the context data can be attached to
a single memory item at the capturing moment. This continuous
and semi-continuous metadata can be presented in graphs over
time or other visualization methods, and may serve as search
keys. The included context data types were chosen after paper
prototype tests. To keep the sharing of recordings simple the
objective was to choose such context data types that users are
willing not just to capture but to share, too.
5.3.3 Sharing of Continuous Data
From the sharing perspective, continuous capturing is promising,
since the missed moments are usually the basis of the precious
memories that would have been shared. The continuously
captured context data enriches the life memories for collective
use. Users can then further annotate the memory item in order to
make it as vivid and personal as desired.
6. PROTOTYPE ACCEPTABILITY
The whole SharMe concept design process was based on user
needs research and prototype iteration with users. The second
round of paper prototype tests were executed with well specified
mock-up pictures. Thus we were also able to gather acceptability
feedback about the prototype. Acceptability testing aimed at
gathering feedback for the overall concept and a few selected
features: event-based sharing and continuous capturing. The tests
were carried out in a laboratory. Five users of the chosen target
group (young people who are experienced with technology) tested
the prototype by carrying out three prepared test tasks. We asked
the test users to evaluate the usefulness of each feature while
testing it.
Generally, the prototype received good acceptance feedback from
all the five test users. Users liked the idea of creating the event
and setting parameters for it already before the actual event.
Preset settings for events and group inviting features were highly
appreciated. Instant sharing to the group of invited users suited
the users’ usage habits well. Users thought that the event settings
should be changeable anytime because the time and invited group
of the event might change even during it. Moreover, creating the
events afterwards should be supported. Certain features, such as
reorganizing events, managing the overlapping and tagging the
events’ recordings were regarded as PC UI’s features.
The idea of capturing continuous metadata during the event
received a warm welcome. However, the comments on different
metadata types were conflicting. Not surprisingly, location and
near-by resources (devices, services) interested most. Capturing
user’s bio-signals interested a few of the users but few needs for
continuous capturing or personal use cases were identified.
Weather data, user or device activity history and news feeds
received only little interest. Users manifested the need for testing
the data types in a functional prototype before being able to
express the final opinions. They emphasized the role of semantic
meanings when representing any of the captured metadata.
Continuous capturing as presented in Section 5.3.1 polarized
users. Most users doubted the need for truly continuous capturing,
and liked the idea of user-initiated buffering. On the other hand,
some users expressed the desire for truly continuous capturing.
Some of the test users presented innovative ideas, for example,
they brought in ideas inspired by the buffering features in modern
digital TV set top boxes and bookmarking of streamed data.
Bookmarking the continuous data was liked as it enables easy
finding of the interesting part in a long recording and sharing data
in real time. The specific implementation has to be well designed,
since truly continuous capturing raised suspicion and ethical
questions within both the user needs research and contextual
inquiries.
Generally, users thought that the event-based capturing and
sharing would raise their use of camera phone and sharing of the
recordings. They stated that many awkward phases are improved
in this prototype, for example, manual sharing and organizing the
recordings. The idea of capturing the usually missed moments is
appealing, but the implementation needs further design. Users
stated that the control in both capturing and sharing has to stay in
users’ hands. Since our study group included also very
technology-oriented users, the current technology’s restrictions in
prototype’s feasibility slightly diluted the reception. Knowing the
current technology limited some users’ future-geared thinking.
The biggest suspicions were related to the cost of using the
service and how the current technology would serve the
implementation.
7. CONCLUSIONS AND FUTURE WORK
The research revealed interesting aspects about capturing and
sharing life memories. Users want to share different kinds of
memories but the extent of sharing – with whom and what data –
varies between users and situations. One precondition on sharing
a memory is its interestingness. Hence, a system that allows users
to capture the most interesting moments that usually are missed,
fascinates users. People tend to capture memories in different
events. As the events are usually experienced with a group of
close people, it is natural to base the sharing of memories around
events and the related group. Test users agreed on the basic
ideology of the developed SharMe concept.
Continuous capturing turned to be a central feature in this study.
However, it requires further design to ensure its suitability with
users’ needs. Continuous capturing generates new problems with
the amount of recorded material. Proper collages and browsing
features become essential when using the vast amount of recorded
data. Truly continuous capturing (“24/7”) is challenging to
implement but it would offer users extensive possibilities to
capture the important life events.
Context data and its usage as metadata or triggers for adaptable
user interfaces need further extensive studies. The visualization
paradigms and use cases for each context data type should be
developed. A proper context-aware system with ability to adapt
itself to the current conditions and easy ways to mark the
interesting parts of the stream are desired. Automatic user- or
environment-originated triggers, such as high pulse rate or high
noise, should be studied.
As the total concept of SharMe is rather extensive, further
research in several fields is required. This study focused on
capturing and sharing the memories. Regarding sharing, we have
planned a future research for further studying the concept of
collective memory and its utilization possibilities. To be able to
study and evaluate the use in a more realistic setting, a
(semi)functional version of the prototype is required. A functional
prototype that users are able to interact with in real contexts of
use can bring out further problems in the usage.
Upcoming acceptability tests will be more interactive with a real
user interface (on a mobile phone), and with a pairs or group tests.
Prepared use cases and scenarios will be created to show how the
usage could proceed in certain situations. Cases will include, for
example, capturing the recordings and metadata using events, and
sharing and browsing of memories with the mobile device.
Assigning test users with different roles within the test group
makes it possible to observe the collective usage of memories.
It is challenging to apply user-centered design process when
designing applications with no existing comparison systems. In
the contextual inquiries we had to choose a focus from existing
sub-task of mobile photography. Even though this worked out
well – we got very relevant data from the inquiries – it would be
useful to combine user-centered design with other methods. One
interesting approach is “user innovation” [17] which could be
combined with the user-centered data gathering and analysis in
supporting designers’ innovativeness.
8. ACKNOWLEDGEMENTS
This article was part of the SharMe project led by Nokia Research
Center (NRC) in cooperation with Tampere University of
Technology, Institute of Human-Centered Technology (IHTE).
We want to express our gratitude for NRC’s Salla Myllylä,
Markku Laitkorpi, Pertti Huuskonen, and IHTE’s Minna Wäljas
and Hannu Soronen. Moreover, we wish to thank all our user
representatives in the research and user-centered design process.
9. REFERENCES
[1] Beyer, H., Holtzblatt, K. 1997. Contextual Design: Defining
Customer-Centered Systems. Morgan Kaufmann Publishers
Inc. San Francisco, CA, USA.
[2] Davis, M. et al. 2004. From Context to Content: Leveraging
Context to Infer Media Metadata. In proceedings of 12th
Annual ACM International Conference on Multimedia.
[3] Vannevar B. 1945. As We May Think. The Atlantic
Monthly, July 1945.
[4] Gemmel, J. et al. 2004. Passive Capture and Ensuing Issues
for a Personal Lifetime Store. 2004. CARPE ’04. New York.
[5] Hayes, G. et al. 2004. The Personal Audio Loop: Designing
a Ubiquitous Audio-Based Memory Aid. MobileHCI 2004.
Glasgow, Scotland.
[6] Van House et al. 2005. The Uses of Personal Networked
Digital Imaging: An Empirical Study of Cameraphone
Photos and Sharing, CHI ‘05. Portland, Oregon, USA.
[7] Okabe, D., Ito, M. Everyday Contexts of Camera Phone Use:
Steps Towards Technosocial Ethnographic Frameworks. In
Höflich, Joachim & Hartmann, Maren Eds. Mobile
Communication in Everyday Life: An Ethnographic View.
Berlin: Frank & Timme, 79-102.
[8] Gemmell, J. et al. 2002. MyLifeBits: Fulfilling the Memex
Vision. ACM Multimedia, pp. 235-238. ACM Multimedia
’02. Juan Les Pins, France.
[9] Hewlett Packard: Photographic memories. http://www.hpl.
hp.com/news/2004/jan-mar/casualcapture.html. Viewed:
1/2007.
[10] Millican, T. iam: Experiences with Persistent Video
Recording. 2005. Publishing and Sharing, CARPE’05,
Singapore.
[11] Dey, A., Abowd, G., Wood, A. 1999. CyberDesk: A
Framework for Providing Self-Integrating Context-Aware
Services. Knowledge-Based Systems, 11. pp. 3-13.
[12] Schilit, B., Adams, N. Want, R. 1994. Context-Aware
Computing Applications. 1st International Workshop on
Mobile Computing Systems and Applications. pp. 85-90.
[13] Kaasinen, E. 2003. User Needs for Mobile-Aware Location
Services, J of Personal and Ubiquitous Computing (6). pp.
70-79. Springer-Verlag, London.
[14] Häkkilä, J. 2006. Usability with Context-aware Mobile
Applications – Case Studies and Design Guidelines. Doctoral
Dissertation. Acta Universitatis Ouluensis, Finland.
[15] Sarvas, R. 2006. Designing User-Centric Metadata for
Digital Snapshot Photography. Doctoral Dissertation.
Helsinki Institute of Information Technology, Finland.
[16] Davis, M. et al. 2005. MMM2: Mobile Media Metadata for
Media Sharing. Conference on Human Factors in Computing
Systems, CHI ’05. Portland, Oregon, USA.
[17] Kanstrup, A.M. & Christiansen, E. 2006. Selecting and
Evoking Innovators: Combining Democracy and Creativity.
Proceedings of the fourth Nordic Conference on Human-
Computer Interaction (NordiCHI), Oslo, Norway.
