SlideShare une entreprise Scribd logo
1  sur  6
Télécharger pour lire hors ligne
Social Media Metrics for the Cultural Heritage Sector
                                                                                                                                                                                Developing a Prototype

                                                                                                                                                                                                            Thijs Waardenburg

              Media Technology group, LIACS, Leiden University, the Netherlands
    Crossmedia Business research group, University of Applied Sciences Utrecht, the Netherlands
                                [thijs.waardenburg@hu.nl]



                                           Abstract: The online presence of organizations is long gone from being just a
                                           web page. Social media have enabled easy and inexpensive interaction between
                                           millions of individuals and communities. This has not gone unnoticed by cul-
                                           tural heritage institutes. The question is what all these social media activities
                                           bring. Even if an institute knows what it tries to achieve online, the metrics of-
                                           ten consist of confusing accumulation of statistics, across several systems and
                                           reveal little about online user behaviour, engagement and satisfaction. In the re-
                                           search project Museum Compass a prototype of a social media monitor is de-
                                           veloped, which will contain data of current and historic online activities on Fa-
                                           cebook, Twitter, YouTube, Foursquare and Flickr of all registered Dutch muse-
                                           ums. The first version of this monitor has been developed, and we believe that
                                           this is a good moment to discuss – mostly in a practical sense – our general ap-
                                           proach and preliminary results.

                                           Keywords: social media, analytics, metrics, cultural heritage, museums


1                                          Introduction
Social media have enabled easy, inexpensive interaction between millions of individ-
uals and communities. This has not gone unnoticed by archives, museums, libraries
and cultural heritage institutes and is visible in a broad array of initiatives and exper-
iments with social media, in and outside these institutes (Hekman and Van Vliet,
2011). The question is what all these social media activities bring. To answer that
question, one has to start with measuring these activities. However, current metrics
solutions often consist of confusing accumulation of statistics, across several systems,
that reveal “little about online user behaviour, engagement and satisfaction” (Finnis,
2011).
    The Crossmedia Business research group of the University of Applied Sciences
Utrecht runs a project called “Museum Compass”. Within this project, several ser-
vices and tools are being developed to support Dutch museum professionals to use
crossmedia solutions more effectively and efficiently. One of the tools is a prototype
of a monitor for social media activities of Dutch registered museums1. The main goal
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
1
                 The Netherlands Museum Register, http://www.museumregisternederland.nl/, checked 11-04-2012.
of this monitor is to offer museum professionals better insight in the effects of social
media usage. It is being developed in the context of the Museum Compass project,
and is therefore focused on museums, or more generally the ‘cultural heritage sector’.
However, it is developed in such a way that it can also be used for other sectors. In
that sense, the word ‘museums’ in this paper can also be substituted with, for in-
stance, ‘festivals’ or ‘retailers’.
    The first version of the monitor has been developed, and we believe that this is
a good moment to discuss – mostly in a practical sense – our general approach and the
choices we made for the development of this prototype. In this paper, the actual data
that is being collected with the monitor is not discussed. The organization of this pa-
per is as follows. Section two is focused on related work, in section three the ap-
proach and choices that we made are described. The conclusions and discussion are
described in the remainder of this paper.


2                                          Related Work
Many articles, books, papers, etc. are written about social media, and various defini-
tions are proposed. For instance, Kaplan and Haenlein (2009) define social media as
“a group of Internet-based applications that build on the ideological and technological
foundations of Web 2.0, and that allow the creation and exchange of User Generated
Content”. Brussee and Hekman (2010) propose a more philosophical definition. Ac-
cording to them social media are “highly accessible media”. They say that social me-
dia are not characterized by the role (digital) technology takes an sich, but by the
accessibility of the media supply chain to the general public. However, most defini-
tions agree that current social media are digital by nature.
    The digital nature allows automatically collecting and viewing data of social me-
dia activities. Commonly used terms for this are ‘social monitoring’, ‘social media
analytics’, and ‘social media metrics’. Many commercial of-the-shelf solutions can be
found for this. For instance, an overview of more than 220 solutions can be found at
the website wiki.kenburbary.com2. These solutions range from relatively simple and
free of charge, to extensive and costly. Hofer-Shall (2011) recognizes three catego-
ries:
    (1) “Social Dashboards: web-based tools that focus primarily on managing and
            analysing social media data”.
    (2) “Multichannel Analytics Providers: analytics infrastructures that mine social
            media data along with other structured and unstructured data sources”.
    (3) “Listening Service Partners: vendors with proprietary social analytics tools
            and professional consulting teams that write custom research reports based
            on social media data.”
    Most of the solutions of the first category are ‘one-size-fits-all’ tools, and they
collect data of one, or a limited number of platforms. The solutions of the second
category, as Hofer-Shall describes, mine both structured and unstructured data, mean-
ing that besides quantitative data, also qualitative data and data from traditional media
is collected and analysed. According to Kaplan (2009) this is important, because
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
2
                 http://wiki.kenburbary.com/, checked 11-04-2012.	
  
“what is true for different types of Social Media also holds for the relationship be-
tween Social Media and traditional media: Integration is key!” Murdough (2009), on
the other hand, stresses that “the important rule is to focus on just a few metrics for
each objective so that program evaluation remains simple and one does not end up in
"analysis paralysis."”. The third category implies the need for customized research
reports. The fact that social media research reports are written for specific sectors
supports this observation. Two examples of these kind of reports are the Dutch social
media monitor for healthcare (“Social Media Monitor Zorg”)3 and the British Cul-
ture24 report, titled “Let’s Get Real: How to Evaluate Online Success?”4.


3                                          Approach and Choices
The Museum Compass research is similar to, and partly inspired by the Culture24
research. An important difference is that the Culture24 research is based on social
media data from a selection of museums, and at a given moment. The goal of the
monitor that we develop is to continually monitor social media activities of a ‘whole’
sector, and mine its history. As mentioned in the previous section, there are many
(commercial) ‘off-the-shelf’ solutions available to monitor social media. However,
we chose to develop our own monitor for the following reason: developing our own
monitor offers better opportunities to experiment, customize, and learn, in order to get
a better understanding of the subject. This is also recognized by, for instance, Bruns
and Liang (2012), as they explain that “for more sophisticated research programmes,
and for the tracking and study of larger-scale datasets over longer time periods, more
advanced and usually custom-made tools and methods are required.”
    As a research group we do not have the intention to develop production software.
The monitor that we develop will therefore be a prototype that, in the end, may serve
as a basis to develop production software (which is more reliable, maintainable, scal-
able, etc.).


3.1                                        Requirements
We identified three basic requirements for the monitor, which can also be considered
as development phases:
    (1) Measuring social media activities of museums (what are museums sending?).
    (2) Measuring impact of social media activities of museums (how does ‘the pub-
        lic’ respond on these activities?).
    (3) Serving as a social media benchmark, so that museums can compare them-
        selves to one another on this area (how do museums relate to one another,
        with regard to social media activities?).
    On the basis of popularity (‘maturity’) and type we chose the following five social
media platforms: Facebook (‘social networking’), Twitter (‘micro blogging’), Flickr
(‘photo sharing’), YouTube (‘video sharing’), and Foursquare (‘location sharing’).
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
3
                 http://www.socialmediamonitorzorg.nl/, only available in Dutch, checked 18-04-2012.
4
                 http://weareculture24.org.uk/projects/action-research/how-to-evaluate-success-online/, checked at 18-
                 04-2012.	
  
Regarding the handling of data of social media platforms, we identified the fol-
lowing requirements for our monitor:
    (1) Data needs to be collected and stored in a database (‘backend’).
    (2) Data needs to be interpreted and combined where possible.
    (3) Data needs to be presented (‘frontend’, in the form of a ‘dashboard’).
We started with collecting and storing publicly available data of social media activi-
ties of museums.


3.2                                        Data Collection
The above-mentioned platforms all offer an application programmers interface (API)
that enables basically anyone to develop an application that exchanges data with the
platform. Each API (i.e. platform) has its specific set of data-elements. Which data-
elements can be accessed depends on authentication levels. We recognize three gen-
eral levels of data access:
      (1) No authentication: ‘Access token’ and account-approval are not needed. On-
          ly basic account data can be received (e.g. account name, account-ID, profile
          image, etc.).
      (2) One-sided authentication. ‘Access token’ is needed, but account-approval is
          not needed. Publicly available / visible account-data can be received (e.g.
          messages, number of fans, etc.).
      (3) Two-sided authentication. Application needs to be registered at platform,
          ‘access token’ and account-approval are needed. Extensive account-data can
          be received (e.g. friend list, private messages, etc.).
     The number of data-elements can be fairly comprehensive. A Facebook ‘post’, for
instance, consists of 26 data-elements like ‘id’, ‘from’, ‘to’, ‘message’, ‘likes’, etc.
For practical reasons, we chose to collect a selection of data-elements. The data-
transfer rate varies per platform. For example, the Twitter API5 has a limit of 350 data
requests an hour (called ‘rate limit’). Flickr has a limit of 500 requests. However, we
established that data request limits are not always very clear, and for no obvious rea-
son can be turned down. Facebook does not offer a clear insight in request limits at
all. We anticipated on this by carefully planning our ‘cronjobs’ for data-requests. An
easier, but much costlier solution for this is to buy the data6, which has the advantage
that one actually receives all historical data. For instance, it turns out that the free API
of Twitter ‘only’ provides 3200 historical tweets per account7. Due to the financial
consequence, buying data was not an option for us at this moment.
     To actually collect and structure the data we used the scripting language PHP in
combination with the database software of MySQL. PHP is widely used and is there-
fore supported by most (if not all) API’s. Another advantage of PHP is that it is a
language that many freelance programmers are familiar with and that allows hiring
programmers more easily when needed.
     To collect data from any social network account, one needs to have account iden-
tifiers (IDs). The Netherlands Museum Register does register official websites of
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
5
                 Twitter has several types of API’s. In this case the REST API is meant.
6
                 For instance: Twitter offers a service called ‘Firehose’ (accessible through 3rd parties).
7
                 3 of the 146 museum Twitter accounts had more than 3200 tweets.
museums, but does not register account-IDs of social media networks. We therefore
needed to collect and verify this information ourselves. Collecting social media ac-
count-IDs is in essence a simple task, but it proofed to be time-consuming. Because it
is difficult to automate the verification of authenticity of accounts, this was basically
done manually. The protocol that we used to collect account-IDs was: ‘search for the
museum in question on the social media websites‘. If account(s) are found: ‘look for
reference to official museum website’ and/or ‘make a selection’.
    After a set of account-IDs was collected, we started collecting as much historical
data as possible. Because collecting historical data involved large amounts of data, we
had to be careful with the request-limits, mentioned earlier in this section. We there-
fore decided to execute these scripts by hand. When the historical data was collected,
we created ‘cronjobs’ to periodically run the scripts to collect new data. After approx-
imately one month, we collected enough data to make a number of basic data views,
like tag clouds, tables and graphs using Google Chart Tools8.


4                                          Lessons Learned and Discussion
One of the main questions that we try to answer within the Museum Compass re-
search project is what social media activities bring for the cultural heritage sector. The
monitor that is being developed is meant to help answer that question. It is not the
answer to the question. Besides this monitor, other tools are being developed, specifi-
cally designed for the (Dutch) cultural heritage sector, and these tools will be linked
to the monitor.
    The first version collects data from Twitter, Facebook, and Flickr and has a num-
ber of basic and separate data representations. The data is not yet being interpreted or
combined. In that sense, the monitor cannot demonstrate its actual potential at this
moment. However, as mentioned in the introduction, this is not discussed in this pa-
per. What is discussed is our approach and choices we made during the development
this prototype.
    To collect social media accounts, one can consider these two options: (1) create an
inventory of social media platforms that are currently used by museums, or (2) select
social networks with a large user base and look which museums use them. In retro-
spection, we believe that former choice is better, because it gives a complete over-
view of the social media platforms that are used. We choose the latter, because this is
less time-consuming. The question is if this is a defendable argument. Besides that,
the protocol for verifying account-IDs should be refined, as it is too subjective.
    The final point of discussion here is our choice to collect a selection of data, in-
stead of collecting all available data. On the long-term this could be an impractical,
inefficient decision, as we may come to the conclusion that we need specific data-
elements that we left out in the first place. A solution for this is to collect and store all




	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
8
                 https://developers.google.com/chart/, checked 20-04-2011.	
  
data in a non-relational, or ‘NoSQL’ database9. This makes the monitor also easier
scalable.
    We chose develop our own tool, for which we described our reasons. However,
we do not exclude that, in the near future, other (commercial) solutions may also fulfil
the requirements. Still, we believe that the development of this tool is useful way to
get a better understanding of the subject.
    The next version of the monitor will contain data of the three other social media
platforms: Flickr, YouTube and Foursquare. At this moment we are researching the
possibilities for visualizations and for combining data-elements.


References
       1. Andreas M. Kaplan, M. H. (2009). Users of the world, unite! The challenges and
          opportunities of Social Media. Elsevier.
       2. Claartje Bunnik, E. v. (2011). Niet Tellen Maar Wegen. Amsterdam: Boekmanstudies.
       3. Erik Hekman, H. v. (2011). Bringing the Past to the Present. Utrecht.
       4. Fisher, T. (2009). ROI in social media: A look at the arguments. Journal of Database
          Marketing & Customer Strategy Management , 189–195.
       5. Groen, A. J. (2005). Knowledge Intensive Entrepeneurship in Networks: Towards a Multi-
          Level/Multi-Dimensional Approach. Enschede: Journal of Enterprising Culture.
       6. Hofer-Shall, Z. (2011). The 2011 Listening Platform Landscape. Forrester Research.
       7. Jane Finnis, S. C. (2011). Let's Get Real: How to Evaluate Online Success. Brighton.
       8. Liang, A. B. (2012). Tools and Methods for Capturing Twitter Data for Natural Disasters.
          First Monday.
       9. Murdough, C. (2009). Social Media Measurement: It's Not Impossible. Journal of
          Interactive Advertising.




	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
9
                 For an explanation, see, for instance
                 http://www.couchbase.com/sites/default/files/uploads/all/whitepapers/NoSQL-Whitepaper.pdf, checked
                 24-04-12.

Contenu connexe

Tendances

An Exploratory Study of the Motivations and Satisfactions on Mobile Web Browsing
An Exploratory Study of the Motivations and Satisfactions on Mobile Web BrowsingAn Exploratory Study of the Motivations and Satisfactions on Mobile Web Browsing
An Exploratory Study of the Motivations and Satisfactions on Mobile Web BrowsingRuby Kuo
 
Building and Sustaining a Community using the Social Web
Building and Sustaining a Community using the Social WebBuilding and Sustaining a Community using the Social Web
Building and Sustaining a Community using the Social Weblisbk
 
Metrics for Understanding Personal and Institutional Use of the Social Web
Metrics for Understanding Personal and Institutional Use of the Social WebMetrics for Understanding Personal and Institutional Use of the Social Web
Metrics for Understanding Personal and Institutional Use of the Social Weblisbk
 
C3 The Hyperlinked Library: Future Technologies and Their Applications
C3 The Hyperlinked Library: Future Technologies and Their ApplicationsC3 The Hyperlinked Library: Future Technologies and Their Applications
C3 The Hyperlinked Library: Future Technologies and Their Applicationslisbk
 
NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...
NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...
NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...Dlazarow
 
Anti-social Networking: Web 2.0 and Social Exclusion
Anti-social Networking: Web 2.0 and Social ExclusionAnti-social Networking: Web 2.0 and Social Exclusion
Anti-social Networking: Web 2.0 and Social ExclusionUltan O'Broin
 
Jill Freyne - Collecting community wisdom: integrating social search and soci...
Jill Freyne - Collecting community wisdom: integrating social search and soci...Jill Freyne - Collecting community wisdom: integrating social search and soci...
Jill Freyne - Collecting community wisdom: integrating social search and soci...DERIGalway
 
Assignment #10 (p2): Group coursework presentation of research
Assignment #10 (p2): Group coursework presentation of researchAssignment #10 (p2): Group coursework presentation of research
Assignment #10 (p2): Group coursework presentation of researchmedia_jojo
 
Assignment 10 group coursework presentation of research draft 2 part 1.5
Assignment 10   group coursework presentation of research draft 2 part 1.5Assignment 10   group coursework presentation of research draft 2 part 1.5
Assignment 10 group coursework presentation of research draft 2 part 1.5Abc Abc
 
VU University Amsterdam - The Social Web 2016 - Lecture 6
VU University Amsterdam - The Social Web 2016 - Lecture 6VU University Amsterdam - The Social Web 2016 - Lecture 6
VU University Amsterdam - The Social Web 2016 - Lecture 6Davide Ceolin
 
twitter_visual_chi2016
twitter_visual_chi2016twitter_visual_chi2016
twitter_visual_chi2016Cat Yao
 
Can Web Search Be Enhanced For User-Generated
Can Web Search Be Enhanced For User-GeneratedCan Web Search Be Enhanced For User-Generated
Can Web Search Be Enhanced For User-GeneratedEvan Atkinson
 
Putting Social Media to work in Information Services
Putting Social Media to work in Information ServicesPutting Social Media to work in Information Services
Putting Social Media to work in Information ServicesAnne Peoples
 
Web 2.0 -_hype_or_helpful
Web 2.0 -_hype_or_helpfulWeb 2.0 -_hype_or_helpful
Web 2.0 -_hype_or_helpfulMargarida Costa
 
VU University Amsterdam - The Social Web 2016 - Lecture 4
VU University Amsterdam - The Social Web 2016 - Lecture 4VU University Amsterdam - The Social Web 2016 - Lecture 4
VU University Amsterdam - The Social Web 2016 - Lecture 4Davide Ceolin
 
You have to start somewhere (2012)
You have to start somewhere (2012)You have to start somewhere (2012)
You have to start somewhere (2012)Anne Peoples
 
Open Educational Practices (OEP): What They Mean For Me and How I Use Them
Open Educational Practices (OEP): What They Mean For Me and How I Use ThemOpen Educational Practices (OEP): What They Mean For Me and How I Use Them
Open Educational Practices (OEP): What They Mean For Me and How I Use Themlisbk
 
Networked Social Media in Learning, Teaching and Research
Networked Social Media in Learning, Teaching and ResearchNetworked Social Media in Learning, Teaching and Research
Networked Social Media in Learning, Teaching and ResearchMaged N. Kamel Boulos
 
ADLUG 2008 Web 2.0 - Library 2.0 presentation
ADLUG 2008 Web 2.0 - Library 2.0 presentationADLUG 2008 Web 2.0 - Library 2.0 presentation
ADLUG 2008 Web 2.0 - Library 2.0 presentation@CULT Srl
 

Tendances (20)

An Exploratory Study of the Motivations and Satisfactions on Mobile Web Browsing
An Exploratory Study of the Motivations and Satisfactions on Mobile Web BrowsingAn Exploratory Study of the Motivations and Satisfactions on Mobile Web Browsing
An Exploratory Study of the Motivations and Satisfactions on Mobile Web Browsing
 
Building and Sustaining a Community using the Social Web
Building and Sustaining a Community using the Social WebBuilding and Sustaining a Community using the Social Web
Building and Sustaining a Community using the Social Web
 
Metrics for Understanding Personal and Institutional Use of the Social Web
Metrics for Understanding Personal and Institutional Use of the Social WebMetrics for Understanding Personal and Institutional Use of the Social Web
Metrics for Understanding Personal and Institutional Use of the Social Web
 
C3 The Hyperlinked Library: Future Technologies and Their Applications
C3 The Hyperlinked Library: Future Technologies and Their ApplicationsC3 The Hyperlinked Library: Future Technologies and Their Applications
C3 The Hyperlinked Library: Future Technologies and Their Applications
 
NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...
NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...
NGOs responding to Crisis: Using Social Media to Meet New Challenges, The Cas...
 
Anti-social Networking: Web 2.0 and Social Exclusion
Anti-social Networking: Web 2.0 and Social ExclusionAnti-social Networking: Web 2.0 and Social Exclusion
Anti-social Networking: Web 2.0 and Social Exclusion
 
Jill Freyne - Collecting community wisdom: integrating social search and soci...
Jill Freyne - Collecting community wisdom: integrating social search and soci...Jill Freyne - Collecting community wisdom: integrating social search and soci...
Jill Freyne - Collecting community wisdom: integrating social search and soci...
 
Assignment #10 (p2): Group coursework presentation of research
Assignment #10 (p2): Group coursework presentation of researchAssignment #10 (p2): Group coursework presentation of research
Assignment #10 (p2): Group coursework presentation of research
 
Assignment 10 group coursework presentation of research draft 2 part 1.5
Assignment 10   group coursework presentation of research draft 2 part 1.5Assignment 10   group coursework presentation of research draft 2 part 1.5
Assignment 10 group coursework presentation of research draft 2 part 1.5
 
VU University Amsterdam - The Social Web 2016 - Lecture 6
VU University Amsterdam - The Social Web 2016 - Lecture 6VU University Amsterdam - The Social Web 2016 - Lecture 6
VU University Amsterdam - The Social Web 2016 - Lecture 6
 
twitter_visual_chi2016
twitter_visual_chi2016twitter_visual_chi2016
twitter_visual_chi2016
 
Can Web Search Be Enhanced For User-Generated
Can Web Search Be Enhanced For User-GeneratedCan Web Search Be Enhanced For User-Generated
Can Web Search Be Enhanced For User-Generated
 
Putting Social Media to work in Information Services
Putting Social Media to work in Information ServicesPutting Social Media to work in Information Services
Putting Social Media to work in Information Services
 
Web 2.0 -_hype_or_helpful
Web 2.0 -_hype_or_helpfulWeb 2.0 -_hype_or_helpful
Web 2.0 -_hype_or_helpful
 
VU University Amsterdam - The Social Web 2016 - Lecture 4
VU University Amsterdam - The Social Web 2016 - Lecture 4VU University Amsterdam - The Social Web 2016 - Lecture 4
VU University Amsterdam - The Social Web 2016 - Lecture 4
 
You have to start somewhere (2012)
You have to start somewhere (2012)You have to start somewhere (2012)
You have to start somewhere (2012)
 
Open Educational Practices (OEP): What They Mean For Me and How I Use Them
Open Educational Practices (OEP): What They Mean For Me and How I Use ThemOpen Educational Practices (OEP): What They Mean For Me and How I Use Them
Open Educational Practices (OEP): What They Mean For Me and How I Use Them
 
Networked Social Media in Learning, Teaching and Research
Networked Social Media in Learning, Teaching and ResearchNetworked Social Media in Learning, Teaching and Research
Networked Social Media in Learning, Teaching and Research
 
ADLUG 2008 Web 2.0 - Library 2.0 presentation
ADLUG 2008 Web 2.0 - Library 2.0 presentationADLUG 2008 Web 2.0 - Library 2.0 presentation
ADLUG 2008 Web 2.0 - Library 2.0 presentation
 
Entradas & Dimopoulos: Public science on the web
Entradas & Dimopoulos: Public science on the webEntradas & Dimopoulos: Public science on the web
Entradas & Dimopoulos: Public science on the web
 

En vedette

Over de ontwikkeling van een effectieve mediastrategie
Over de ontwikkeling van een effectieve mediastrategieOver de ontwikkeling van een effectieve mediastrategie
Over de ontwikkeling van een effectieve mediastrategieHU-Crossmedialab
 
Manon jacobs gradassignment_1529185
Manon jacobs gradassignment_1529185Manon jacobs gradassignment_1529185
Manon jacobs gradassignment_1529185HU-Crossmedialab
 
Social Media Metrics for the Cultural Heritage Sector
Social Media Metrics for the Cultural Heritage SectorSocial Media Metrics for the Cultural Heritage Sector
Social Media Metrics for the Cultural Heritage SectorHU-Crossmedialab
 

En vedette (6)

Museumscapes
MuseumscapesMuseumscapes
Museumscapes
 
Over de ontwikkeling van een effectieve mediastrategie
Over de ontwikkeling van een effectieve mediastrategieOver de ontwikkeling van een effectieve mediastrategie
Over de ontwikkeling van een effectieve mediastrategie
 
Experts en leken
Experts en lekenExperts en leken
Experts en leken
 
Narrativestructurein3 d
Narrativestructurein3 dNarrativestructurein3 d
Narrativestructurein3 d
 
Manon jacobs gradassignment_1529185
Manon jacobs gradassignment_1529185Manon jacobs gradassignment_1529185
Manon jacobs gradassignment_1529185
 
Social Media Metrics for the Cultural Heritage Sector
Social Media Metrics for the Cultural Heritage SectorSocial Media Metrics for the Cultural Heritage Sector
Social Media Metrics for the Cultural Heritage Sector
 

Similaire à Social Media Metrics for the Cultural Heritage sector

HANDLING YOUR MEDIA PRESENCE - THINK TANKS AND SOCIAL NETWORKS
HANDLING YOUR MEDIA PRESENCE  - THINK TANKS AND SOCIAL NETWORKSHANDLING YOUR MEDIA PRESENCE  - THINK TANKS AND SOCIAL NETWORKS
HANDLING YOUR MEDIA PRESENCE - THINK TANKS AND SOCIAL NETWORKSDr Danica Radovanovic
 
Tn T Horizons April 28 2008
Tn T Horizons April 28 2008Tn T Horizons April 28 2008
Tn T Horizons April 28 2008brettssu
 
Innovative approaches to analyses of online social networks
Innovative approaches to analyses of online social networksInnovative approaches to analyses of online social networks
Innovative approaches to analyses of online social networksJakob Jensen
 
A review for the online social networks literature
A review for the online social networks literatureA review for the online social networks literature
A review for the online social networks literatureAlexander Decker
 
WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...
WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...
WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...Ramine Tinati
 
Modelling the Media Logic of Software Systems
Modelling the Media Logic of Software SystemsModelling the Media Logic of Software Systems
Modelling the Media Logic of Software SystemsJan Schmidt
 
Literature Review of Information Behaviour on Social Media
Literature Review of Information Behaviour on Social MediaLiterature Review of Information Behaviour on Social Media
Literature Review of Information Behaviour on Social MediaDavid Thompson
 
Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...
Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...
Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...NewMediaMK
 
Technical artifacts as intermediaries?!
Technical artifacts as intermediaries?!Technical artifacts as intermediaries?!
Technical artifacts as intermediaries?!Nele Heise
 
Defining the IT artefact in social media for eParticipation: An Ensemble view
Defining the IT artefact in social media for eParticipation: An Ensemble viewDefining the IT artefact in social media for eParticipation: An Ensemble view
Defining the IT artefact in social media for eParticipation: An Ensemble viewMarius Rohde Johannessen
 
Sdi, communities and social media
Sdi, communities and social mediaSdi, communities and social media
Sdi, communities and social mediaWirelessInfo
 
Research Design for the Study of Social Media Use by Dutch Development Organi...
Research Design for the Study of Social Media Use by Dutch Development Organi...Research Design for the Study of Social Media Use by Dutch Development Organi...
Research Design for the Study of Social Media Use by Dutch Development Organi...Anand Sheombar
 
MOVING: Applying digital science methodology for TVET
MOVING: Applying digital science methodology for TVETMOVING: Applying digital science methodology for TVET
MOVING: Applying digital science methodology for TVETMOVING Project
 
Researching Social Media – Big Data and Social Media Analysis
Researching Social Media – Big Data and Social Media AnalysisResearching Social Media – Big Data and Social Media Analysis
Researching Social Media – Big Data and Social Media AnalysisFarida Vis
 
Galli Guarneri Huhtamaki - Vertigo (2009)
Galli Guarneri Huhtamaki - Vertigo (2009)Galli Guarneri Huhtamaki - Vertigo (2009)
Galli Guarneri Huhtamaki - Vertigo (2009)Luca Galli
 
The Emerge Show02 Ng Ti P
The Emerge Show02 Ng Ti PThe Emerge Show02 Ng Ti P
The Emerge Show02 Ng Ti PGeorge Roberts
 
HADOOP based Recommendation Algorithm for Micro-video URL
HADOOP based Recommendation Algorithm for Micro-video URLHADOOP based Recommendation Algorithm for Micro-video URL
HADOOP based Recommendation Algorithm for Micro-video URLdbpublications
 

Similaire à Social Media Metrics for the Cultural Heritage sector (20)

Visualizing the Digital Humanities, Deic2012
Visualizing the Digital Humanities, Deic2012Visualizing the Digital Humanities, Deic2012
Visualizing the Digital Humanities, Deic2012
 
HANDLING YOUR MEDIA PRESENCE - THINK TANKS AND SOCIAL NETWORKS
HANDLING YOUR MEDIA PRESENCE  - THINK TANKS AND SOCIAL NETWORKSHANDLING YOUR MEDIA PRESENCE  - THINK TANKS AND SOCIAL NETWORKS
HANDLING YOUR MEDIA PRESENCE - THINK TANKS AND SOCIAL NETWORKS
 
Tn T Horizons April 28 2008
Tn T Horizons April 28 2008Tn T Horizons April 28 2008
Tn T Horizons April 28 2008
 
Innovative approaches to analyses of online social networks
Innovative approaches to analyses of online social networksInnovative approaches to analyses of online social networks
Innovative approaches to analyses of online social networks
 
A review for the online social networks literature
A review for the online social networks literatureA review for the online social networks literature
A review for the online social networks literature
 
Intercultural
InterculturalIntercultural
Intercultural
 
WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...
WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...
WSI Stimulus Project: Centre for longitudinal studies of online citizen parti...
 
Modelling the Media Logic of Software Systems
Modelling the Media Logic of Software SystemsModelling the Media Logic of Software Systems
Modelling the Media Logic of Software Systems
 
Literature Review of Information Behaviour on Social Media
Literature Review of Information Behaviour on Social MediaLiterature Review of Information Behaviour on Social Media
Literature Review of Information Behaviour on Social Media
 
Cook et al
Cook et alCook et al
Cook et al
 
Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...
Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...
Broadcasting Entrepreneurial Brainchild Based on Some Social Media Experiment...
 
Technical artifacts as intermediaries?!
Technical artifacts as intermediaries?!Technical artifacts as intermediaries?!
Technical artifacts as intermediaries?!
 
Defining the IT artefact in social media for eParticipation: An Ensemble view
Defining the IT artefact in social media for eParticipation: An Ensemble viewDefining the IT artefact in social media for eParticipation: An Ensemble view
Defining the IT artefact in social media for eParticipation: An Ensemble view
 
Sdi, communities and social media
Sdi, communities and social mediaSdi, communities and social media
Sdi, communities and social media
 
Research Design for the Study of Social Media Use by Dutch Development Organi...
Research Design for the Study of Social Media Use by Dutch Development Organi...Research Design for the Study of Social Media Use by Dutch Development Organi...
Research Design for the Study of Social Media Use by Dutch Development Organi...
 
MOVING: Applying digital science methodology for TVET
MOVING: Applying digital science methodology for TVETMOVING: Applying digital science methodology for TVET
MOVING: Applying digital science methodology for TVET
 
Researching Social Media – Big Data and Social Media Analysis
Researching Social Media – Big Data and Social Media AnalysisResearching Social Media – Big Data and Social Media Analysis
Researching Social Media – Big Data and Social Media Analysis
 
Galli Guarneri Huhtamaki - Vertigo (2009)
Galli Guarneri Huhtamaki - Vertigo (2009)Galli Guarneri Huhtamaki - Vertigo (2009)
Galli Guarneri Huhtamaki - Vertigo (2009)
 
The Emerge Show02 Ng Ti P
The Emerge Show02 Ng Ti PThe Emerge Show02 Ng Ti P
The Emerge Show02 Ng Ti P
 
HADOOP based Recommendation Algorithm for Micro-video URL
HADOOP based Recommendation Algorithm for Micro-video URLHADOOP based Recommendation Algorithm for Micro-video URL
HADOOP based Recommendation Algorithm for Micro-video URL
 

Social Media Metrics for the Cultural Heritage sector

  • 1. Social Media Metrics for the Cultural Heritage Sector Developing a Prototype Thijs Waardenburg Media Technology group, LIACS, Leiden University, the Netherlands Crossmedia Business research group, University of Applied Sciences Utrecht, the Netherlands [thijs.waardenburg@hu.nl] Abstract: The online presence of organizations is long gone from being just a web page. Social media have enabled easy and inexpensive interaction between millions of individuals and communities. This has not gone unnoticed by cul- tural heritage institutes. The question is what all these social media activities bring. Even if an institute knows what it tries to achieve online, the metrics of- ten consist of confusing accumulation of statistics, across several systems and reveal little about online user behaviour, engagement and satisfaction. In the re- search project Museum Compass a prototype of a social media monitor is de- veloped, which will contain data of current and historic online activities on Fa- cebook, Twitter, YouTube, Foursquare and Flickr of all registered Dutch muse- ums. The first version of this monitor has been developed, and we believe that this is a good moment to discuss – mostly in a practical sense – our general ap- proach and preliminary results. Keywords: social media, analytics, metrics, cultural heritage, museums 1 Introduction Social media have enabled easy, inexpensive interaction between millions of individ- uals and communities. This has not gone unnoticed by archives, museums, libraries and cultural heritage institutes and is visible in a broad array of initiatives and exper- iments with social media, in and outside these institutes (Hekman and Van Vliet, 2011). The question is what all these social media activities bring. To answer that question, one has to start with measuring these activities. However, current metrics solutions often consist of confusing accumulation of statistics, across several systems, that reveal “little about online user behaviour, engagement and satisfaction” (Finnis, 2011). The Crossmedia Business research group of the University of Applied Sciences Utrecht runs a project called “Museum Compass”. Within this project, several ser- vices and tools are being developed to support Dutch museum professionals to use crossmedia solutions more effectively and efficiently. One of the tools is a prototype of a monitor for social media activities of Dutch registered museums1. The main goal                                                                                                                 1 The Netherlands Museum Register, http://www.museumregisternederland.nl/, checked 11-04-2012.
  • 2. of this monitor is to offer museum professionals better insight in the effects of social media usage. It is being developed in the context of the Museum Compass project, and is therefore focused on museums, or more generally the ‘cultural heritage sector’. However, it is developed in such a way that it can also be used for other sectors. In that sense, the word ‘museums’ in this paper can also be substituted with, for in- stance, ‘festivals’ or ‘retailers’. The first version of the monitor has been developed, and we believe that this is a good moment to discuss – mostly in a practical sense – our general approach and the choices we made for the development of this prototype. In this paper, the actual data that is being collected with the monitor is not discussed. The organization of this pa- per is as follows. Section two is focused on related work, in section three the ap- proach and choices that we made are described. The conclusions and discussion are described in the remainder of this paper. 2 Related Work Many articles, books, papers, etc. are written about social media, and various defini- tions are proposed. For instance, Kaplan and Haenlein (2009) define social media as “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content”. Brussee and Hekman (2010) propose a more philosophical definition. Ac- cording to them social media are “highly accessible media”. They say that social me- dia are not characterized by the role (digital) technology takes an sich, but by the accessibility of the media supply chain to the general public. However, most defini- tions agree that current social media are digital by nature. The digital nature allows automatically collecting and viewing data of social me- dia activities. Commonly used terms for this are ‘social monitoring’, ‘social media analytics’, and ‘social media metrics’. Many commercial of-the-shelf solutions can be found for this. For instance, an overview of more than 220 solutions can be found at the website wiki.kenburbary.com2. These solutions range from relatively simple and free of charge, to extensive and costly. Hofer-Shall (2011) recognizes three catego- ries: (1) “Social Dashboards: web-based tools that focus primarily on managing and analysing social media data”. (2) “Multichannel Analytics Providers: analytics infrastructures that mine social media data along with other structured and unstructured data sources”. (3) “Listening Service Partners: vendors with proprietary social analytics tools and professional consulting teams that write custom research reports based on social media data.” Most of the solutions of the first category are ‘one-size-fits-all’ tools, and they collect data of one, or a limited number of platforms. The solutions of the second category, as Hofer-Shall describes, mine both structured and unstructured data, mean- ing that besides quantitative data, also qualitative data and data from traditional media is collected and analysed. According to Kaplan (2009) this is important, because                                                                                                                 2 http://wiki.kenburbary.com/, checked 11-04-2012.  
  • 3. “what is true for different types of Social Media also holds for the relationship be- tween Social Media and traditional media: Integration is key!” Murdough (2009), on the other hand, stresses that “the important rule is to focus on just a few metrics for each objective so that program evaluation remains simple and one does not end up in "analysis paralysis."”. The third category implies the need for customized research reports. The fact that social media research reports are written for specific sectors supports this observation. Two examples of these kind of reports are the Dutch social media monitor for healthcare (“Social Media Monitor Zorg”)3 and the British Cul- ture24 report, titled “Let’s Get Real: How to Evaluate Online Success?”4. 3 Approach and Choices The Museum Compass research is similar to, and partly inspired by the Culture24 research. An important difference is that the Culture24 research is based on social media data from a selection of museums, and at a given moment. The goal of the monitor that we develop is to continually monitor social media activities of a ‘whole’ sector, and mine its history. As mentioned in the previous section, there are many (commercial) ‘off-the-shelf’ solutions available to monitor social media. However, we chose to develop our own monitor for the following reason: developing our own monitor offers better opportunities to experiment, customize, and learn, in order to get a better understanding of the subject. This is also recognized by, for instance, Bruns and Liang (2012), as they explain that “for more sophisticated research programmes, and for the tracking and study of larger-scale datasets over longer time periods, more advanced and usually custom-made tools and methods are required.” As a research group we do not have the intention to develop production software. The monitor that we develop will therefore be a prototype that, in the end, may serve as a basis to develop production software (which is more reliable, maintainable, scal- able, etc.). 3.1 Requirements We identified three basic requirements for the monitor, which can also be considered as development phases: (1) Measuring social media activities of museums (what are museums sending?). (2) Measuring impact of social media activities of museums (how does ‘the pub- lic’ respond on these activities?). (3) Serving as a social media benchmark, so that museums can compare them- selves to one another on this area (how do museums relate to one another, with regard to social media activities?). On the basis of popularity (‘maturity’) and type we chose the following five social media platforms: Facebook (‘social networking’), Twitter (‘micro blogging’), Flickr (‘photo sharing’), YouTube (‘video sharing’), and Foursquare (‘location sharing’).                                                                                                                 3 http://www.socialmediamonitorzorg.nl/, only available in Dutch, checked 18-04-2012. 4 http://weareculture24.org.uk/projects/action-research/how-to-evaluate-success-online/, checked at 18- 04-2012.  
  • 4. Regarding the handling of data of social media platforms, we identified the fol- lowing requirements for our monitor: (1) Data needs to be collected and stored in a database (‘backend’). (2) Data needs to be interpreted and combined where possible. (3) Data needs to be presented (‘frontend’, in the form of a ‘dashboard’). We started with collecting and storing publicly available data of social media activi- ties of museums. 3.2 Data Collection The above-mentioned platforms all offer an application programmers interface (API) that enables basically anyone to develop an application that exchanges data with the platform. Each API (i.e. platform) has its specific set of data-elements. Which data- elements can be accessed depends on authentication levels. We recognize three gen- eral levels of data access: (1) No authentication: ‘Access token’ and account-approval are not needed. On- ly basic account data can be received (e.g. account name, account-ID, profile image, etc.). (2) One-sided authentication. ‘Access token’ is needed, but account-approval is not needed. Publicly available / visible account-data can be received (e.g. messages, number of fans, etc.). (3) Two-sided authentication. Application needs to be registered at platform, ‘access token’ and account-approval are needed. Extensive account-data can be received (e.g. friend list, private messages, etc.). The number of data-elements can be fairly comprehensive. A Facebook ‘post’, for instance, consists of 26 data-elements like ‘id’, ‘from’, ‘to’, ‘message’, ‘likes’, etc. For practical reasons, we chose to collect a selection of data-elements. The data- transfer rate varies per platform. For example, the Twitter API5 has a limit of 350 data requests an hour (called ‘rate limit’). Flickr has a limit of 500 requests. However, we established that data request limits are not always very clear, and for no obvious rea- son can be turned down. Facebook does not offer a clear insight in request limits at all. We anticipated on this by carefully planning our ‘cronjobs’ for data-requests. An easier, but much costlier solution for this is to buy the data6, which has the advantage that one actually receives all historical data. For instance, it turns out that the free API of Twitter ‘only’ provides 3200 historical tweets per account7. Due to the financial consequence, buying data was not an option for us at this moment. To actually collect and structure the data we used the scripting language PHP in combination with the database software of MySQL. PHP is widely used and is there- fore supported by most (if not all) API’s. Another advantage of PHP is that it is a language that many freelance programmers are familiar with and that allows hiring programmers more easily when needed. To collect data from any social network account, one needs to have account iden- tifiers (IDs). The Netherlands Museum Register does register official websites of                                                                                                                 5 Twitter has several types of API’s. In this case the REST API is meant. 6 For instance: Twitter offers a service called ‘Firehose’ (accessible through 3rd parties). 7 3 of the 146 museum Twitter accounts had more than 3200 tweets.
  • 5. museums, but does not register account-IDs of social media networks. We therefore needed to collect and verify this information ourselves. Collecting social media ac- count-IDs is in essence a simple task, but it proofed to be time-consuming. Because it is difficult to automate the verification of authenticity of accounts, this was basically done manually. The protocol that we used to collect account-IDs was: ‘search for the museum in question on the social media websites‘. If account(s) are found: ‘look for reference to official museum website’ and/or ‘make a selection’. After a set of account-IDs was collected, we started collecting as much historical data as possible. Because collecting historical data involved large amounts of data, we had to be careful with the request-limits, mentioned earlier in this section. We there- fore decided to execute these scripts by hand. When the historical data was collected, we created ‘cronjobs’ to periodically run the scripts to collect new data. After approx- imately one month, we collected enough data to make a number of basic data views, like tag clouds, tables and graphs using Google Chart Tools8. 4 Lessons Learned and Discussion One of the main questions that we try to answer within the Museum Compass re- search project is what social media activities bring for the cultural heritage sector. The monitor that is being developed is meant to help answer that question. It is not the answer to the question. Besides this monitor, other tools are being developed, specifi- cally designed for the (Dutch) cultural heritage sector, and these tools will be linked to the monitor. The first version collects data from Twitter, Facebook, and Flickr and has a num- ber of basic and separate data representations. The data is not yet being interpreted or combined. In that sense, the monitor cannot demonstrate its actual potential at this moment. However, as mentioned in the introduction, this is not discussed in this pa- per. What is discussed is our approach and choices we made during the development this prototype. To collect social media accounts, one can consider these two options: (1) create an inventory of social media platforms that are currently used by museums, or (2) select social networks with a large user base and look which museums use them. In retro- spection, we believe that former choice is better, because it gives a complete over- view of the social media platforms that are used. We choose the latter, because this is less time-consuming. The question is if this is a defendable argument. Besides that, the protocol for verifying account-IDs should be refined, as it is too subjective. The final point of discussion here is our choice to collect a selection of data, in- stead of collecting all available data. On the long-term this could be an impractical, inefficient decision, as we may come to the conclusion that we need specific data- elements that we left out in the first place. A solution for this is to collect and store all                                                                                                                 8 https://developers.google.com/chart/, checked 20-04-2011.  
  • 6. data in a non-relational, or ‘NoSQL’ database9. This makes the monitor also easier scalable. We chose develop our own tool, for which we described our reasons. However, we do not exclude that, in the near future, other (commercial) solutions may also fulfil the requirements. Still, we believe that the development of this tool is useful way to get a better understanding of the subject. The next version of the monitor will contain data of the three other social media platforms: Flickr, YouTube and Foursquare. At this moment we are researching the possibilities for visualizations and for combining data-elements. References 1. Andreas M. Kaplan, M. H. (2009). Users of the world, unite! The challenges and opportunities of Social Media. Elsevier. 2. Claartje Bunnik, E. v. (2011). Niet Tellen Maar Wegen. Amsterdam: Boekmanstudies. 3. Erik Hekman, H. v. (2011). Bringing the Past to the Present. Utrecht. 4. Fisher, T. (2009). ROI in social media: A look at the arguments. Journal of Database Marketing & Customer Strategy Management , 189–195. 5. Groen, A. J. (2005). Knowledge Intensive Entrepeneurship in Networks: Towards a Multi- Level/Multi-Dimensional Approach. Enschede: Journal of Enterprising Culture. 6. Hofer-Shall, Z. (2011). The 2011 Listening Platform Landscape. Forrester Research. 7. Jane Finnis, S. C. (2011). Let's Get Real: How to Evaluate Online Success. Brighton. 8. Liang, A. B. (2012). Tools and Methods for Capturing Twitter Data for Natural Disasters. First Monday. 9. Murdough, C. (2009). Social Media Measurement: It's Not Impossible. Journal of Interactive Advertising.                                                                                                                 9 For an explanation, see, for instance http://www.couchbase.com/sites/default/files/uploads/all/whitepapers/NoSQL-Whitepaper.pdf, checked 24-04-12.