Bringing you: Crowd-funding news; Grey lit to choose; Data modelling issues; National landscape views (and now west of Hormuz)

HER Forum Winter Meeting 2016, Birmingham and Midland Institute

(Post presentation questions and comments – in italicised type following each abstract – are based on notes taken at the meeting).

View from the Chair – Chris Webster (South West Heritage Trust)

The winter HER Forum meeting was held as last year in the lecture theatre at the Birmingham and Midland Institute with 60 attendees. The late withdrawal of one of the speakers caused the programme to be reshuffled slightly but also allowed for more discussion.

The morning started with a presentation on the National Historic Landscape Characterisation project, concentrating on the difficulties faced when trying to combine various rather disparate data sets created over a 20-year period. This was followed by an update on the integration of the various bibliographic databases held by the ADS into the British and Irish Archaeological Bibliography and how this might be maintained in the future. Finally, the software Digital Dig Team that has been developed by Digventures to allow on-site digital recording was demonstrated.

In the afternoon, the potential of data modelling for HER data was explored and the project to develop a system for the national HER in Qatar was described and the system demonstrated. After the close of formal business, there was a very good turn-out for post-meeting networking.

National Landscape Characterisation Project - Ben Wallace (ALGAO/ Warwickshire HER), Roger M Thomas (Historic England), Crispin Flower and Abby Hunt (Exegesis)

Roger Thomas introduced the session with an overview of HLC and the decisions behind the idea of creating a national dataset. Ben Wallace then followed this with the background to the NHLC project and the history of the involvement of ALGAO.

A summary of the completed stages of the NHLC project was presented by Abby Hunt and Crispin Flower, which have so far been Stage 1 (the set-up and initialisation of the project) and Stage 2 (a review of the existing HLC databases).

Amongst the key elements of Stage 2 was the creation of an online workspace to allow easy communication between project board members and sub-contractors, a summary of the statistics derived from all the individual supplied HLC datasets and the workflow which Exegesis used to process and clean the datasets. Some of the issues encountered during the data processing included the presence of gaps in some datasets (ranging from small, insignificant gaps to more substantial, problematic gaps) and overlapping polygons within datasets.

The process of merging all the HLC datasets into a single database (within SQL Server) was then undertaken, which resulted in 3945 pairs of unique instances of Type/Broad Type combinations across the national dataset. A term-matching exercise was carried out by Locus Consulting as part of Stage 2 of the project which matched the Type/Broad Type term pairs with terms from the Historic Characterisation Thesaurus (HCT). Approximately two thirds of the term pairs were matched to HCT terms or matched to a new candidate term, but for the remaining one third (which represented a much higher proportion of actual occurrences in the dataset), there was no simple match and peer review was required.

Using the matched terms, trial maps were created colouring polygons according to Top Term and HCT term, which fed into the development of various further options for presenting the data, including casting the data onto a grid. The outcomes of Stage 2 have been the successful collation, analysis and aggregation of the sub-regional HLC datasets, the rehearsal of various methodologies for transformation into a single NHLC dataset and gaining a greater understanding of the various issues and challenges that need to be overcome to complete the project.

Following a meeting of the Project Board at the end of Stage 2, it was agreed that the project would continue to Stages 3-5, which will culminate in the delivery of the final NHLC product, accompanied by a full report, user guide and documentation, and the deposition of the project archive.

Question (HER Officer 1) – After the data matching has been completed, will matched data be going back to HERs?

Answer (BW) – It is not certain at this stage, although I would be interested in looking into the idea. There are potential difficulties, however, in that it will not be a simple exercise and some HERs might not have time to do the work required.

Question (HER Officer 1) – Do you think it would be worthwhile for HERs to update their HLCs with the new terms?

Answer (CF) – Whilst it would be relatively easy to get a concordance table to HERs this isn’t factored into the project as it currently stands.

Question (HER Officer 2) – Will it be possible to relate the character aspects visible at the national level back to local types?

Answer (BW) – At the national level the final product won’t have that level of granularity. The idea is that, should this prove necessary, Natural England will contact the relevant HER(s) directly.

(CF) – A concordance, outlining the equations made between character types at the local level, exists and could be published (although not part of the scope of the current project). This would enable users to drill down to original information.      

Question (Archaeologist 1) – Did the urban HLCs cause any problems when collating HLC data?

Answer (CF) – Urban areas have tended to be greatly simplified in the thesaurus matching process. It should be remembered that NE’s interests seldom require fine detail within urban contexts.

(BW) – There is certainly a question as to how open areas within urban environments are dealt with but this doesn’t come within the scope of this project.

Question (HER Officer 3) – Why has this been done? What are NE going to do with the data?

Answer (BW) – In general terms HLC is a body of information that should feed into NE’s work but, because the existing framework doesn’t sit at the correct level, it has tended to get eroded out of Countryside Stewardship schemes thus far. The full details of how it will eventually be utilised by HE have yet to be fully ‘bottomed out’. However, it is anticipated that its primary use will be for high level targeting and national decision making. Employment at any more local level would be less acceptable to ALGAO who will be safeguarding against anything that might be construed as inappropriate use. 
 
(RT) – A national HLC is seen as a necessary precursor to using the data at the national level.   
 
Question (HER Officer 4) – Those involved in creating the national HLC are to be congratulated. I was involved in HLC in the east midlands and correlating the various datasets looked almost impossible. However, my concern is whether the original HLC types will be accessible? If not, won’t there effectively be two completely different systems operating at the different levels resulting in detail being lost?

Answer (CF) – It is not intended that the national record will become the de facto HLC. It should still serve as an index to signpost local information. I would be worried if I thought more detailed data was going to be lost.  

Question (HER Officer 4) – How will this signposting work?

Answer (CF) – All aspects should be traceable and drill-down mechanisms will be put in place.

Question (HER Officer 4) – Is there to be a user guide?

Answer (CF) – Certainly, but it’s not clear at this stage how it will be used.  This will depend on how NE make the data available to their staff. However, new avenues of communication will certainly be opened up.

(BW) – Consistency of use is important on both sides, on the part of NE and HERs. The data is still there though, and nothing has been lost. HERs may, at any future stage, return to employing their previous character types.

Comment (CF) – A persuasive case might be made for re-doing quite a lot of the original HLC project.   

ADS Library: weaving a web of references – Jo Gilham (Archaeology Data Service)

The ADS Library is the fusion of existing datasets. These include journal and series backruns archived with the ADS, the Library of unpublished fieldwork reports (aka the Grey literature library) which is mostly populated with reports from OASIS and last but not least the British and Irish Archaeological Bibliography (BIAB) which is in itself a collection of different datasets which have been collected over the last hundred years.

The project to get these references online as a single resource has involved cleaning, mapping and enhancing the data from the different datasets. Allowing them to share the same data structure and hopefully give users as consistent information about each item listed in the library. Some records simply show the existence of a report or publication and others link out to the publication itself where available. There was some overlap in the combined datasets and we have endeavoured to merge records where appropriate in order to limit the existence of duplicates in the lists of results.

The ADS Library is intended to be a ‘live’ resource with records continuing to be added from numerous sources:

• OASIS reports will be uploaded to the library with the record of a report coming on completion of the OASIS record and the report following shortly after, format and review permitting.
• Journal and monograph records will be supplied regularly from the main archaeological publishers, including BAR, Oxbow and Taylor & Francis.
• Smaller regional publishers will be able to log in to the library and add their new publications as they become available.
• Volunteers will be able to request permission to add records.
• All users will be able to register and tag or comment on records.
It will also be possible to link to resources and the bibliographic record about resources in various ways:
• Each record in the library will have a direct URL so that it can be included in external databases
• Where there are DOIs available for a publication this will be available so that the resource can be linked to in perpetuity
• Bibliographic references within the library will be able to be exported in various bibliographic formats for import into other systems.
• Research frameworks, which in future will be updated from OASIS, will link through to the library record and publication where available

The aim is to create a definitive resource for publications in archaeology - linking the grey with the traditionally published and keeping alive the hard work already done in producing BIAB so far.

Question (JG to the floor) – OASIS Standard requires validation by the HER before any report goes into the archive. However, after a time-span (to be agreed) the report will go into the library anyway. Does anyone have any opinions regarding the timescale for this? Should this be after 3 months or is this too short a period? (Contractors seem eager for speed and, under the present arrangements if an HER is validating records this does usually take place within 3 months).

Additionally, it is accepted that some reports shouldn’t go out and the contractor has the power of embargo. Should HERs have this power too?

Question (BW) – Is there any automatic tagging of those reports for which HER validation is missing?

Reply – No, however they could be distinguished by the lack of a validation tag.

Question (CW) – Would there be a problem concerning ‘restraint of trade’ for contractors?

Comment (JG) – Probably there would be mitigating circumstances since, in the majority of cases, this would form part of a planning decision.

Comment (HER Officer 5) – The HER might not wish to validate because conditions had not been fulfilled.

Comment (HER Officer 6) – It seems evident that things should have passed satisfactorily through the planning process before the report is made accessible through the library.

Comment (Assistant Inspector of Ancient Monuments, Historic England) – Isn’t there a danger of the voluntary sector feeling disenfranchised? 

Question (HER Officer 6) – Who makes the choice as to whether OASIS standard or lite is adopted?

Answer (JG) – This decision rests with the HER.

Question (HER Officer 7) – Would it be possible to embargo reports if using lite?

Answer (JG) – Not at present.

Question (CW) – For how long would the embargo last?

Answer (JG) – Periods can vary but are specified.

Question (Assistant Inspector of Ancient Monuments, Historic England) - Is an indefinite embargo possible?

Answer (JG) – Yes, but under such circumstances the report wouldn’t be catalogued anyway.

Comment (HER Officer 6) – In most cases the embargo put in place would be relevant to elements of the report rather than the work in its entirety.

Question (Historic Environment Intelligence Analyst, Historic England) – How would this fit in with the Freedom of Information Act?

Answer (JG) – The report can be requested from the HER but, in broader terms, it will be less available.

Comment (BW) – On balance problems are perhaps more likely to emerge if HERs can’t embargo.

Comment (HER Officer 7) – I would refuse to validate if I hadn’t received a copy of the report. Our terms and conditions insist on this.

Comment (Assistant Inspector of Ancient Monuments, Historic England) – It seems possible that this dialogue has a specific archaeological emphasis and won’t necessarily be seen as applicable to the built environment.

Comment (HER Officer 5) – It seems that there would be an incentive for publishers to link to the website in that it would encourage sales. 

Response (JG) – Theoretically yes, but this wouldn’t work well for smaller publishers. URLs are much more fluid.

(JG closed by saying that the system wasn’t yet in a state in which it could appear online. However, a launch would take place early in 2017).    

Digital Dig Team; adventures with a born digital, co-produced and open archaeological archive - Amanda Forster (Programme Manager, DigVentures)

Digital Dig Team can best be described as a Community Management System for archaeology projects – standing somewhere between a relational database designed to accommodate the online input of archaeological information, a Content Management Systems (such as a WordPress website), and a fully accessible, co-produced archaeological archive. DDT is based on a cloud-based, open-source software platform (L-P Archaeology’s ARK) which we have substantially enhanced to enable researchers to publish data directly from the field using any web-enabled device (such as a smartphone or tablet) into a live relational database.

Once recorded, the born-digital archive is instantly accessible via open-access on a dedicated website, and also published to social profiles of all project participants. Though other digital recording systems have been used in archaeology over a number of years (such as Intrasis, IADB or the ARK), Digital Dig Team is the first to adopt an explicitly open protocol, enabling a step change in the ease of recording data, accessibility to data once uploaded, and previously unachievable levels of participation, communication and social (and professional) shareability of project information.

Digital Dig Team enables archaeologists to build audiences (through immersive storytelling), generate revenue (through crowdfunding), harness public participation (through crowdsourcing) and improve research results by making data instantly available to a networked specialist team in the moment of discovery. Digital Dig Team, as used on several sites in Britain and Europe, can be further explored here:

http://digventures.com/digital-dig-team/

 Question (HER Officer 4) – The process you outline seems to be very rapid. Is there adequate quality control?

Answer – The system is open and data does go live immediately. However, there is a weekly cycle of checking in which incorrect material is amended.

Question (AL) – Concerning the use of hand-held technology and i-pads on site: I would be nervous of employing equipment of this type in mud and rain. How is this overcome?

Answer – Since we work outside the commercial pressures of planning and development we can operate at a slower pace. Given that these are training exercises we avoid working in bad weather. One practical issue that does impact on this methodology is that of internet reception. However, we can employ booster boxes and, so far, reception has been achieved at all of our investigations, no matter how remote.

Question (AL) – In terms of context sheets being used by inexperienced excavators, is there a possibility that they could be over-simplified in the quest for ease of use?

Answer – As a matter of practice there are always experienced staff on hand to answer questions and all venturers are encouraged to consult them. The IT system also gives access to a ‘help’ section. All work is subject to periodic checking.

Question (Archaeologist 2) – Are your projects based on initiatives such as LPs?

Answer – Yes.

Question (Archaeologist 2) – How do you think that your work improves upon them?

Answer – Its advantages are in the immediacy of the work and in its community element. At its core it is a democratising process in which people work at the same level using the same equipment.  

MAID in England- Results of the research project – Adam Lodoen (Bournemouth University)

(NB – The following summary, as submitted, includes a map not reproduced here. The full version is available on the file store of the HER Forum JiscMail site (HER Forum Dec 2016)).

In 2011, a collaborative doctoral studentship was awarded by the AHRC under the acronym “MAID in England” (Modelling Archaeological Intervention Data for landscape interpretation and strategic planning in England).

The project was set up to tackle a problem which has existed for more than 20 years:  The abundance of archaeological data, which is a relatively new development, opens up many opportunities, but no coherent strategies for data analysis currently exist in order to realise them in practice. The MAID project investigated one possible way to analyse archaeological data using computer algorithms and predictive modelling techniques. It explored through a series of experimental models what is possible to achieve using predictive algorithms when applied to existing archaeological data. It also explored how such an algorithm should be constructed.

It is fair to say that the MAID project only opens up a research field, and much more research needs to be done. However, the research clearly demonstrates that the use of algorithms in conjunction with predictive modelling techniques has a great potential for analysing existing archaeological data.  Two different kinds of models can be created (depending on the available data):

• Conditional models, predicting the character of the archaeology given some archaeological investigation or intervention (if an excavation finds archaeological remains, what is the character of the discovered remains likely to be?). It should be possible to create these kinds of models using available HER with relatively little modification

• Unconditional models, predicting the character of the archaeology given some archaeological investigation or intervention as well as the probability of finding anything in the first place. To create these models, it may be necessary to combine information from more than one dataset (e.g. HER data and AIP data)

Unconditional models could be used to characterise the archaeological resource in a given area in a detailed way, to create detailed predictions useful for assessing risk to the archaeological resource from development, and to understand how different parts of the landscape were used in the past.

The results of the research project could potentially be used as the groundwork for the creation of an algorithm or piece of software useful for archaeological database management, and as basis for further research.The modelling used data from research-based and commercially-driven archaeological interventions held in Historic Environment Records (HERs) and by the Archaeological Investigations Project (AIP). Data from the Hertfordshire HER relating to the medieval and Bronze Age periods, and data from the Cornwall HER relating to the medieval periods were used for the research. The research was carried out by Adam Lodoen (PhD Candidate, Bournemouth University) in collaboration with Historic England (EH).

Question (Archaeologist) – Have the models been tested by fieldwork?

Answer – Whilst the models have their origins in fieldwork they have yet to be tested in this way. They should be tested but this would only be possible with the resources of a larger project. 

Question (Archaeologist) – Studies of this type are often subject to unintentional biases and significant incremental error. Do you think these have been avoided?

Answer – The modelling process must stick closely to the question ‘what is the character of the archaeology’? In doing so outcomes are less likely to be affected by bias. Adjustments can be made for patterns resulting from different intervention techniques.

Survival bias is, however, more difficult to exclude. It is nonetheless possible though, to assess what is presently apparent. In Cornwall the situation was less problematic as the modelling was primarily based on earthwork evidence.

Comment (Archaeologist) – Having worked on palaeo-hydrology modelling myself I’ve noticed that drawing together data combinations can quickly exaggerate error and bias.

Reply – It must be acknowledged that this possibility can’t be excluded. The process can, though, be employed with certain caveats in place. Patterns across historical periods are certainly evident, particularly in some areas (for example Hertfordshire).

Developing a National HER for Qatar - Richard Cuttler (MOSPA Heritage Consultancy Ltd)

The project to develop a national HER for the state of Qatar was undertaken between 2008 and 2015. It was a tendered project and, from an early stage, it was quite apparent that very little archaeological data of any sort presently existed for the area. This extended even to the level of monument type thesauri and an established framework of prehistoric and historic periods. Other information held by various Qatari government departments was available, however, some of which was capable of being usefully employed. 

The aim of the initiative was to develop a systemised inventory geared to a range of uses relating to planning and tourism and which could also integrate with the work of other national ministries. It was seen as important to operate within an existing framework of standards for which MIDAS Heritage served as the foundation. However, whilst its concepts were transferable, actual thesauri were not. To develop a thesaurus which was locally applicable required local experience of both landscape and archaeology. 

The new database needed to be capable of further expansion and updating and to work in conjunction with the datasets of other government ministries. There was also a requirement that it should be a bilingual system (Arabic/English).

Preliminary efforts began in 2008 with remote sensing work including Google Earth and de-classified satellite images. On arriving ‘on the ground’, however, the team received an excel spread sheet of officially recognised archaeological sites (which at that stage amounted to 60 in total). This was developed and expanded using field work done by a team (numbering between 15 and 30 at any one time) in the months between September and March.

Another exciting facet of the survey related to the development of a maritime database. This mapped out a range of submerged landscapes and located the remains of 7 shipwreck sites. In addition it was possible to include analyses made of cores taken in advance of submarine oil drilling.

Statutory protection for heritage sites already existed in Qatar and this extended to all of the sites on the official list. However, the concept of an HER was poorly developed in the middle-east as a whole, and cementing this new idea within the mind-set of the local administration, plus explaining its potential uses and benefits became a significant element of the project’s goals.

The resulting HER dataset used open source software sitting on a webserver accessed through an HTML web browser. It held a total of 7000 site records and 25,000 photographic images. The system went live in 2013. A range of possible future developments were envisaged including a translation module and websites for public access. These, however, have been put ‘on hold’ following a drop in oil prices.

Question (HER Officer 8) – Are all records available in both Arabic and English?

Answer – Yes, the output uses the same information (and details regarding photographs) which is all stored in coded format. This can translate into both English and Arabic. The drop down menus will be in whichever language option you choose. However, the free text remains a slight problem. This is not automatically translated and Google Translate in its present form isn’t a viable option. The basis of this translation process at present depends on coded pick-lists.

Question (HER Officer 8) – How is the system presently being maintained and used?

Answer – Extensive training was given to those who currently use the system. Maintenance and updating of the database is being undertaken, but on a slightly ‘ad hoc’ basis. The data is being used for the production of websites and leaflets whilst the system is employed for storing reports.

Comment (Archaeologist 2 – formerly part of the project team) – The value of the system is that it forms a seamless interface which anyone can use. Reports and the activities of individuals and groups can be tracked using the same software and the various disparate elements combined.

Response (RC) – The key to this lay in making the whole thing as intuitive as possible.

Comment (Archaeologist 2) – The more complex elements were integrated and made available through wizards. Using these required very little knowledge or prior experience. It was one of the project’s goals to create a software package with could be used on a small budget by any third-world country seeking to create an HER.

Question (Assistant Inspector of Ancient Monuments, Historic England) – Have any other countries taken up the software so far?

Answer – Not as yet.