My main aim for this week was to start building more of a profile for each accused witch by trying to upload as much information as possible about them. I then started to focus on the data available for the people related to the trial too in order to build up more information. By the end of the week I feel as I am ready to move away from working with the database and begin to think more about my visualisations.

Adding more data for the accused witches

Aside from my main project of locating the residences for each accused witch, it is important to add as much information about each accused witch onto Wikidata. The added data will help build up their Wikidata profile which can then be accessed from the queried maps. The different datasets added can also act as different map layers which can help with creating a more interesting visualisation such as gender.

To pick the most appropriate dataset to be added to Wikidata for the accused witches I had to think about firstly, what data would be the most useful to know about the witches?  Secondly, what datasets already had an appropriate property on Wikidata?

The properties are used to describe the item that is added to Wikidata for example: the property of Charge (P1595) with the item Witchcraft (Q259745) can be used to highlight that the accused witch had been charged of witchcraft. Although there are many properties already added to Wikidata, the witchcraft data is more niche and so are yet to be properties added for some of the data mentioned in the Witchcraft database such as Torture Type. Therefore, some of the very interesting data from the database will not be able to be added onto Wikidata until a property is produced that is appropriate for the data. As Wikidata is a community based website, the properties that are proposed must be agreed on my the Wikidata community. Furthermore, the new properties should also be able to be useful for other datasets rather than a new property only being of use for the upload of one dataset. Therefore, it can take a long time period for a property to be agreed on.

Here is a TEDx Talk video by Martin Poulter explaining the concept of Wikidata in more detail :

This week I added the information to Wikidata for what each accused witch had been charged with and the ethnicity of accused witch. Aside from these datasets, I looked though the database and thought about what other data would be useful to be added to Wikidata and what properties would need to be produced.

Looking at The People Associated with The Trial

Within the Scottish Witchcraft Database, there is not only data related to the accused witches but there is also a wealth of data regarding the people associated with the witch trials with over 2000 people mentioned. Many of these people were involved in the accused witches trials by being different positions in court such as being a Sheriff or Minister. Therefore, The People Associated with The Trial had a large influence on the outcome of the trials. There is very limited data uploaded to Wikidata for these people with just their name and gender added. Therefore, I wanted to add their residence and occupation onto Wikidata as another data source that I could also work with.

This week I started trying to locate the people’s residence. It had been assumed that there would be a large crossover between the residences of the accused witches and the people associated with the trial. However, I have found that many of the people associated with the trials lived in other small settlements. The different residence locations will make the process much harder to locate all of these residences with a similar process needing to be adopted to search for locations as was conducted for the accused witches.

I then decided it would be best to add the easier locations for the people’s residence and leave the more difficult locations for a future search once more of the other datasets are added. So far I have managed to upload 1350 out of 1981 residences for the people associated with the trial which is a good start.

See the residence map here:

Next week I am going to start looking at the people associated with the trials occupations too so that I can look for trends in the data between location and occupation.

Trying to locate more place-names

At the start of the week I was given another list of place-name suggestions along with ideas of useful people/organisations to email in order to find the more difficult place-names for the accused witches residences.

One of my more interesting place-name conversations with researchers at Sabhal Mòr Ostaig has helped unravel an interesting story behind one of the accused witches whose residence was Feull (mistaken for Fowl ) on the Isle of Skye and the underlying motive related to the witches accusation. To summerise the case of  Marie Nian Innes  explained by Dr Domhnall Uilleam Stiùbhart (senior lecturer from Sabhal Mòr Ostaig), the accusation has been thought to be highly related to Highland politics during this time period with a new Chief of Clan Ranald being appointed. The witchcraft accusation could have been the new Clan Chief trying to prove himself after the estate fell into serious debt. The accused woman was then imprisoned in Castle Tioram in Moidart within the Clan Ranald region, then later Marie found refuge in Feull territory.

Hearing the stories behind these accusations adds an extra layer of life to the database.

The list of unfound locations is continuing to get smaller with just 32 places to be found however, some more work is going to need to be done to find the remaining few.

SPARQL and visualisation skype tutorials

The query service used to find and visualise data from Wikidata relies on a basic knowledge of SPARQL Queries. However, there are a lot of examples on the query service which can be used as stencils to edit for your own data. But I thought it would be useful to have a greater knowledge of SPARQL to write more advanced queries.

Therefore, I was able to receive a skype tutorial from Navino Evans (co-founder of Histropedia) to understand how to write some more advanced parts of SPARQL Queries to work with the witchcraft data in different ways.

Here is an annotated query for finding the count of a property:

Here is an annotated query for the cluster around a specified property:

I then also had another meeting with Martin Poulter (Wikimedian in Residence for the Bodleian Library) related to different ideas for the visualisation of the geographical data. It was suggested that Leaflet would be a good place to start for starting to visualise the data however, this would need a basic knowledge of Java Script and PHP coding. I think that ArcGIS Online would also be another option for the visualisation as an easier program to use without a knowledge of coding being needed. Within the next few weeks I should get the chance to explore these platforms for finding effective visualisations.

Open Refine Tutorial

Aside from my witchy data, this week I have been trying to produce a tutorial teaching people how to use OpenRefine a spreadsheet manipulation software used to upload data onto Wikidata. I aim to get my tutorial edited and uploaded online next week to share the use of this software.