Following an invitation by my colleage Kimmo Elo, a contemporary historian and social scientist with an interest in network analysis and text analysis, I went to meet members of Turku’s DH community, taught an introductory workshop on network data extraction and visualization and gave the first public live demo of the redeveloped histoGraph at the Connect! Perspectives in Digital Humanities workshop series.
In this blog post I highlight a few aspects which were of particular interest for me during the conference (Retro)Digitalisate – Kommentarkultur – Big Data: Zum Stand des Digitalen in den Geisteswissenschaften. The concept behind the conference was quite intrigueing: No classical paper presentations as an attempt to stop people from saying what they always say. Instead the organizers opted for a combination of panel discussions and keynotes. The conference’s subtitle describes the angle: “The state of Digital in the Humanities”. I was invited to talk about tools and the (re-)use of digital materials and would like to thank the organizers Lilian Landes and Norbert Kunz for the invitation.
As far as I could tell, the conference brought together people from very diverse fields of work (libraries, publishing, humanities research, IT, lexicology…) who mixed together in four panels with the topics “Digital publishing”, “Tools”, “Communication” and “Post-publications”. In the end however, the issues which were really up for debate found their way in all of them: What counts as Digital Humanities? Which consequences does this have for the humanities? Are the rapid changes we are experiencing changes for the worse, better or are they just changes?
The four panels made it very clear that “digital” is changing the humanities and makes me agree even more with Mareike König who calls the humanities digital already. But following the discussions also made it very clear that at this stage, old models of publications, career planning, self-identification are in a state of profound transformation which is experienced in a positive way by some and in a negative way by others. At this stage there is no way to predict where this transformation will lead the humanities and continued cuts in funding in- and outside the universities are reasons to be concerned. This is also reflected in one common angle in all of the keynotes and panelists: Sitting it out is not an option.
This blog posts highlights just a few aspects from the conference “Europa baut auf Biographien”, especially those which are particularly close to my field. I was there to talk about networks.
National Library of Vienna
National Biography projects are seen as trusted repositories for information on selected personalities in a nation’s history. In most cases, three conditions need to be met in order to be considered for an entry: significant accomplishments, death and – since these projects typically ran/run for very long periods of time and move alphabetically: having a surname which starts with the right letter. This procedure is of course pragmatic in the pre-digital world but it also meant that Konrad Bloch, a biochemist and nobel prize winner who died in the year 2000, never made it to the German national biography.
Apart from these selection biases and sometimes outdated information the biographies remain a valuable resource for scholars and the public. It was great to see how keen national and regional biographies projects in Germany, Austria and Switzerland were to interlink their entries by means of VIAF and GND ids.
Growth and enrichment of national biographies were of particular interest: Paul Arthur’s keynote referred to a recent change in biography-scholarship: A move towards a biography understood as a network of intersecting lives and activities. His own efforts in this domain for example seeks to aggregate biographical data from different parts of the population, all branded similarly: there is Indigenous Australia, Obituaries Australia, People Australia, Indigenous Australia, Women Australia, Labour Australia as well as a crowd-based project HuNI which lets the public add relations between objects from Australia’s cultural heritage.
Piek Vossen presented their work in BiographyNet and focused on the automated extraction of additional metadata by means of text analysis. Their accomplishments are impressive and a short video on the site describes the project in detail.
In my interpretation and tightly knit to the current wave of networking thinking, a number of talks circled around the idea that biographies should no longer be thought of as isolated depictions of the great but reinterpreted as collective, explicitly European efforts. This is a big, great and most of all: fundable idea.
We presented the first screenshots of the all-new histoGraph during the poster session at this year’s INSNA Sunbelt conference on Social Network Analysis in Brighton. histoGraph facilitates the crowdbased indexation and exploration of multimedia cultural heritage materials. Click on the poster to get an idea of how this works and looks like.
We have outlined our ideas for the future development of histoGraph in this short article titled Interactive Networks for Digital Cultural Heritage Collections – Scoping the Future of HistoGraph (Cite).
Network theories and methods have recently gained wide-spread attention. Graph visualizations look great and will easily grab the attention of any audience. However for those who are new to the field, there is a steep learning curve. Novices need to learn to master a variety of skills starting from network theory to systematic data collection, appropriate means to compute and visualize them and finally the challenge to link back the results of any analyses to their original research questions.
This flow chart is meant to help people with the hard decision whether or not it is worthwhile and feasible to engage with network analysis. You may also think of these questions as a simple way to test the quality of existing projects.
This is my first draft of this flow chart, so nothing is set in stone and I am happy to consider any criticism you leave in the comments.
To get a first idea of Social Network Analysis concepts, you may want to take a look at the Cheat Sheet: Social Network Analysis for Humanists.
Social Network Analysis concepts and methods are extremely powerful ways to describe complex social relations. The field however has developed their own their concepts, some of which require a little bit of translation. This cheat sheet should help with the very first steps.
| Actor | The entity which is described by a node, e.g. a person, an institution etc. |
| Alter/Alteri | Actors in an ego network |
| Broker | A node which is positioned e.g. between two clusters and can act as a “bottleneck” |
| Clique | A fully interconnected group of nodes |
| Community | A set of nodes which are relatively more connected to each other than to the rest of the graph |
| Co-occurrence network | A network in which the edge between nodes is based on the fact that both appear together in the same context, usually in texts. An example would be two people who are mentioned in the same paragraph. |
| Data | A set of individual pieces of information which are typically machine-readable |
| Data visualisation | The visualisation of data in general. Network visualisation is one of many subfields. |
| Dyad | A group of two connected nodes, the smallest possible network |
| Edge attribute | Data which describes a certain aspect of an edge, for example how often two actors speak to each other |
| Edge or tie or arc or link | What connects two nodes. There are slight differences between these terms but they are mostly used interchangeably |
| Ego network | A network which contains all connections of an actor to their alteri and usually also the connections between the alteri |
| Graph | (in a network and math. context): objects which are connected by links. However often used interchangeably with “network” even by experts. |
| Graph theory | The mathematical description of nodes and ties |
| Historical Network Research | SNA + historical research methods and questions |
| Hub | A node with a degree far higher than the average in a network |
| Layout algorithms | Algorithms which arrange a graph according to an underlying logical principle. Many different ones exist and can be fine-tuned by additional parameters. Layout algorithms help reveal patterns in the network data. So-called force-directed algorithms are used often: they consider ties to be “springs” which attract well-connected nodes and seek to avoid crossings between ties. The same data can look very differently with different layout algorithms and even the same algorithm will produce a different visualisation each time it is run; there is therefore not the “one” or “right” way to visualise a network. |
| Network dataset | Typically machine-readable data which contains information on edges between nodes, on node attributes, on the graph itself. |
| Network metaphor | The most common way to refer to networks in the humanities. It typically describes the observation that social relations have an effect on something or somebody without specifying it further. |
| Network theory | The theoretical approach to social networks which often inspires research and the development of algorithms |
| Network visualizations | Are visual representations of network data. Can take various forms, commonly used are node-link-diagrams and matrices |
| Node attribute | Data which describes a certain aspect of a node, for example an actor’s age or gender |
| Node centrality | Describes the extent to which a node is connected to other nodes within a network. Various algorithms exist to describe different aspects of such connectivity. To some extent centrality can be linked to abstract notions such as “influence”, “power” or “importance”. |
| Node or vertex | Refers to the object which is connected to other objects in a graph. The terms are often used interchangeably |
| Relationships | Any relationship between actors can be represented as an edge and two actors can be connected by multiple ties. When conceptualising network ties, it is important that they are defined clearly to allow comparison: What exactly is e.g. “friendship” or “collaboration” and when does it apply? |
| Software for network visualisation and analysis | Helps to create, compute, visualise, modify network data. Many different tools exist for different purposes. Gephi, NodeXL, UCINET, Pajek are among the most well-known |
| Social Network Analysis (SNA) | Is a cross-disciplinary field of research which is based on the axiom that the systematic study of relations between (mostly) humans (often exchanges of some kind) can help answer research questions. |
| Triad | Similar to a dyad, a triad describes a group of three nodes. The number of completed triads in a graph can also provide insight into the structure of the network. |
| Unipartite (or 1-mode), bipartite (or 2-mode) ndn-partite network | Unipartite networks describe relations between one type of nodes (e.g. people connected to people); bipartite networks describe relations between two types of nodes (e.g. people connected to organisations). A bipartite network can be projected into a unipartite network (people who were connected to the same organisation are now connected to each other). This concept can be expanded even further, in theory a network can have any number of types of actors. It is important to note that in these networks, nodes of the same type can never connect to each other, only to nodes of the other types. |
This post gives an overview of the peculiarities of Open Journal System, an open source publication tool and might be of help when considered together with OJS’s documentation. OJS is developed by the Public Knowledge Project has become the chosen publishing platform for CVCE. OJS sports a variety of features which are of relevance for us: Multi-language support, xml integration and a clean and customizable look to it.
In this blog post I describe the process of setting up Transactions in Digital Humanities, the first publication by CVCE which publishes papers from last year’s DHLU conference.
I cannot really comment on the installation process itself since thankfully a colleague of mine in the IT department took on this task. There were a few hickups at first: some of the labels were not displayed correctly as shown in the screenshot below (not that this is not the final design).
OJS features a range of different user profiles, each with associated rights and access levels. Starting with “reader” and “author”, the roles include section editors, layout editors, proofreaders, editors, a journal manager and finally an admin. OJS explain their respective roles rather well. In an ideal world you would have a team of people, each with their respective task working on such a project who would seamlessly interact with each other and pass on tasks from one step to the next. In the real world, one person needs to manage a range of tasks. Thankfully, OJS lets users take on more than one role.
This can get confusing however: Once logged in you put on your author hat and submit an article using a detailed and nicely configurable submission process. This process includes the acknowledgement of copyright issues, metadata self-classification, file upload etc. Once an article is submitted, a section editor or an editor needs to assign a reviewer to the article. To do this, you need to click (in my mind: counter intuitively) on “My Journals” in the right navigation pane.
This brings you back to an overview page which bundles the different roles associated with your user id and provides handy updates on the progress of different publications.
As an editor you are free to assign reviewers, add their respective recommendations (accept/reject) and forward the task to layouters and proof-readers who are able to upload the edited versions of the files they have been working on. OJS stores these versions separately, albeit only the admin has access to the quite rudimentary file browser. OJS offers a quite handy, yet easily overlooked navigation through the publishing process:
You might find it hard to find the “Complete” commands which are associated with each task in the user interface, again in my mind counter intuitively, they are positioned in the middle of the page and are easily overlooked.
Once everything is proof-read and layouted, OJS asks you upload the final version of the file. I chose to edit an HTML file and to upload it directly in the Layout section, selecting the “Galley” option. Not that there seems to be an annoying bug which causes problems when parsing H2 (second level) captions. Replacing them with H3 (third level) captions in an editor solves this.
The test file I worked with came in ODT format. Images embedded in the text and exported to HTML will contain machine-created filenames which will not be recognized by OJS. Instead I embedded images manually in an editor, for example:
IMG SRC=”figure_1.jpg” NAME=”graphics1 WIDTH=741 HEIGHT=536>
The file will be uploaded and displayed above. Click “Edit” next to the uploaded file to change the label of the link to the uploaded file (I stick with “HTML”. Albeit a bit hidden, this editing section is quite useful for any revisions. Here you can replace the uploaded file in case you run into formatting problems and crucially: Upload pictures which accompany the text.
Note that once you close this editing section, you will jump back to the previous page but end up in the deviously similarly looking Copyediting section above the Layout section.
Continue by “Completing” the Proofreading section and the article should be published.
Helpful links:
The Talk of Europe Creative Camp was organised by Max Kemman and Astrid van Aggelen and featured research projects and presentations by 16 researchers in Computer Science and the Humanities – a big thanks for all their hard work!
What do you do when you have all the debates in the EU parliament from 1992 onwards digitized, in RDF and you are just one SPARQL query away from them? If you have one week to work with specialists in network data visualization and management? First of all you need to bridge the gap between domain experts and ask how this data can be used to answer research questions in European studies and Computer Science. This turned out to be as hard as expected.
This week nevertheless yielded a method which helps us to detect unexpected speaker appointments by Members of the European Parliament (created by Fintan McGee of our neighbouring research institute CRP Lippmann) and an import functionality for Nodegoat which makes it easy to pull and visualise data from the Talk of Europe’s SPARQL endpoint (created by Pim van Bree and Geert Kessels of Lab1100). For CVCE this was an excellent opportunity to experiment with new ways of extracting and visualising information from structured data repositories.
To explain our work a short intro is needed: Graphs like the one below are so-called 1-mode networks. This means that there is only one type of nodes (in this case people) connected to each other. Imagine that the people in this network are connected to each other because they all are members of the same sports club.
In this example, a 1-mode network visualisations represents group affiliations by showing links between all members of a group. This works fine here but very quickly becomes cluttered when a group has a large number of members.
You can also represent this kind of information using a 2-mode network; a network which has two types of actors (people, sports club). It is important to mention that in 2-mode networks ties can only exist between actors of different types (person -> sport club), never within one type of actors (person -> person).
2-mode networks are also sometimes called affiliation networks or bimodal networks and are particularly well-suited to identify overlaps between groups as illustrated in the graph above. In both examples it is clear that Laura is the person who is part of both groups. 2-mode visualisations are however often leaner than 1-mode networks since they represent an affiliation of a person by one tie alone and merely imply the links between members of a group.
Both 1-mode and 2-mode networks can be described mathematically as well. There are numerous ways to describe how central a node is in a network and how different nodes cluster together. Fintan’s work focuses on algorithms which are used for 2-mode networks and networks with more than 2 types of actors, so-called multimodal networks. 2-mode networks can be projected into 1-mode networks. This means that instead of 2 types of nodes we end up with only one. A tie between these nodes is added when – in our example – two people are members of the same club. This works just fine in the above examples. But as as Fintan highlights, projections can lead to a loss of information: One actor may be a member of three clubs and end up with the same connections as someone who only attends two clubs. Redundancy is one way to measure this loss of information.
Getting the opportunity to speak is a privilege in the EU parliament and very often the same people are chosen to speak on their area of expertise. But sometimes they don’t. Domain experts Frédéric Allemand (CVCE) and Bjørn Høyland (Oslo University) showed interest to detect the latter. It turned out that Redundancy, a concept developed in graph theory might do just that, as Fintan discovered during the week.
Drastically simplified, Redundancy describes how irreplaceable a node is in a 2-mode graph. In other words: Nodes which can be removed from a graph without changing its overall structure when it is projected into a single mode are considered to be highly redundant.
The graph below was created by Fintan and shows connections between Members of the European Parliament who spoke on certain agenda items in the year 2010 (527 nodes, 595 edges). An interactive version of this graph, realized with the sigmajs plugin for Gephi, is also available.
Click here to launch an interactive version of this Graph
Only those MEPs are shown who have a low level of redundancy (<0.5) which means that speakers with very atypical combinations of topics on which they speak will be highlighted. An example is Mike Nattrass (United_Kingdom) who spoke on contrasting subjects such as:
Ban on commercial whaling (debate)
Implementation of the first railway package directives (debate)
Progress made on resettling Guantanamo detainees and on closing Guantanamo (debate)
Protection of animals used for scientific purposes (debate)
Welfare of laying hens (debate)
This is where our journey ends and domain experts need to take a close look at the data to identify the meaning of the graph and link back their observations to the context knowledge which is and remains essential for the humanist analysis of the parliamentary debates.
For our second project, Geert Kessels and Pim van Bree of Lab1100 have developed a new feature for Nodegoat which allows users to load data directly from the Talk of Europe SPARQL endpoint (or any other). This will make it significantly easier to explore the debate data for anyone who can write a query. Nodegoat might well be the to-date easiest way to set up a relational database, fill it with data and visualise the data as a social or spatial network and to observe changes over time.
That said, it is no trivial task to write such a query. But its advantage lies in the precision with which one can download a very particular constellation of data points.
User can define custom import templates which define how the data will be organized and linked to each other in Nodegoat.
Once these two steps are sorted, the data is ready to be visualised. Nodegoat supports the visualisation of social graphs but also combinations of social graphs and maps. You can check out a few examples on their website and use this User Guide to setup your own project. Sadly we ran out of time to test this new feature and develop queries together with our domain expert.
All in all, this was a great opportunity to start a collaboration between the CVCE DH Lab, CVCE EIS and Fintan of CRP Lippmann and Geert and Pim of Lab1100.
