Search

This talk presents results from our recent research analysing the relationship between digital humanities and the research library. There is clear in specific areas, such as heritage imaging and special collections, but how might this be strengthened in our wider activities to create a stronger partnership and a more equal one with the library as initiator and collaborator rather than as service provider? This relationship has different models and a significant factor is where the DH centre (or research group) is based. UCLDH is a virtual centre with membership across all faculties, and management based in Humanities, Computer Science, and the Built Environment, all within academic departments. In the USA, DH centres are often hosted in the library (for example Stanford [1] Duke [2] Cincinnati [3]) whereas elsewhere the links are more informal (Sula 2013). Muñoz (20120), MITH is another example, argues that neither DH nor libraries should be considered a service. Posner (2013) points to administrative and institutional problems that play their part. ADHO has a Libraries and DH SIG with a pre-conference, Libraries as Research Partner in Digital Humanities, at DH2019 [4]. The model in mainland China is mixed but closer to the USA. The DH research group at Peking University (PKU) is in the library [5], hosting the DH Forum there since 2016. The Shanghai (public and academic) Library has a very active DH research group (Mahony and Gao 2019) and is hosting (Chinese) DH2020. The authors are running a two-day workshop at UCL in collaboration with PKU library to explore and analyse how their practice leverages this relationship between library and research group with a focus on collaborative working, integrated and innovative approaches. We examine how cultural differences (institutional and societal) shape PKU work activities and their formulation of DH practice and wider engagement. This will also be informed by the published reports of Research Libraries UK, particularly their recent work on the creation, archiving, and preservation of tools for DH (Kamposiori 2019). [1] [2] [3] [4] [5] Kamposiori, Christina. (2017). ‘The role of Research Libraries in the creation, archiving, curation, and preservation of tools for the Digital Humanities.’ RLUK Report.&nbsp

The iSchools Consortium isa collective of Information Schools with a shared interest in the relationships between information, people, and technology. Both the University of Sheffield and UCL are active members of the iSchools network; the Information School at Sheffield hosted the 2018 annual conference and the Department of Information Studies at UCL offers Master's and PhD programmes in DH. Two main educational areas that we seek to advance are those of data science and digital humanities. This presentation presents the findings of the iSchools Consortium committee for Digital Humanities Curriculum Design with a particular focus on education management and administration. We are working to develop a more comprehensive picture of the iSchool approach to DH and to formulate our path forward in this dynamic field. We wish to capitalise on the interdisciplinary methods already embedded in iSchool curricula and to facilitate opportunities for collaboration with other institutional departments and initiatives. To that end, the committee sees itself as having a crucial role in helping to define recommendations for digital humanities coursework and instructional resources as well as envisioning long-term goals and strategies for iSchool degree programmes in the field. One motivation for this initiative is the increasing number of job announcements asking for DH graduates, which emphasises the importance of curriculum and skills development in this field. Our aim is to see how iSchools can take this forward by asking what our role might be in the DH community. What might be the unique approach of iSchools in DH and what is it that distinguishes an iSchool's DH programme as opposed to one in other faculties and the wider academy? How might accreditation work and would it be advantageous or counter-productive by restricting innovation and diversity of curricula? What is the synergy between DH and the iSchool and what is the role for the latter in this arena? These are the questions addressed here.

Coal seam microorganisms have attracted much attention for their special functions that are closely related to secondary biogenic gas formation, but the influence of subsurface microbial invasion on coal microorganisms remains poorly understood. In this study, subsurface microorganisms from soils at different depths were used to infect coal samples, and the abundance and composition of the microbial community were analyzed by quantitative polymerase chain reaction (PCR) and high-throughput sequencing. The results showed that methane production from the coal samples treated with soil extractions did not increase significantly nor did the abundance and transcriptional activity of archaeal 16S rRNA and mcrA genes. In contrast, two soil extractions significantly improved the abundance and transcriptional activity of the coal bacteria. There was no significant difference in the effects of different soil extractions on bacterial abundance and transcriptional activity, but differences in the effects on α- and β-bacterial diversity between the coal samples treated by different soil extractions were found. Soil extraction from the shallow layer soil far away from the coal mine increased the bacterial α-diversity of the coal samples and changed the bacterial community composition. In addition, this soil extraction increased various bacterial groups in the coal sample and changed the predictive metagenome functional contents of the coal samples from the bacterial sequence data. In summary, this study provides basic microbial information for subsurface microbial invasion of coal seams and helps increase understanding of the source of microorganisms in in situ coal seams and the changes in the microbial community during subsurface microbial migration.

Automatic law article prediction aims to determine appropriate laws for a case by analyzing its corresponding fact description. This research constitutes a relatively new area which has emerged from recommended algorithm. Therefore, the task is still a challenge due to the highly imbalanced long-tail data distribution and lack of discrimination in the feature representation. To deal with these challenges, we proposed a codex enhanced multi-task framework, which consists of two modules. The first one is a codex learning module that estimates the broad codex attributes related to the case fact for alleviating the long-tail issue. The other one is a Bidirectional Text Convolutional Neural Network, which predicts the law articles by considering both local and global information of the facts. These two modules are learning simultaneously through a multi-task loss function. To evaluate the performance of the method proposed in this paper, we construct a new law article prediction data by collecting the judgment documents from the China Judgement online. Experimental results on the dataset demonstrate the effectiveness of our proposed method and can outperform other comparison methods.

This study aims to better understand the relationship between the response to acid mine drainage (AMD) stress of tolerant plants and changes in root-related bacterial communities. In this study, reed stems were planted in AMD-polluted and unpolluted soils, and high-throughput sequencing was conducted to analyze the bacterial community composition in the soil, rhizosphere, rhizoplane and endosphere. The results showed that the effect of AMD pollution on root-associated bacterial communities was greater than that of rhizo-compartments. Proteobacteria were dominant across the rhizo-compartments between treatments. The microbiomes of unpolluted treatments were enriched by Alphaproteobacteria and Betaproteobacteria and depleted in Gammaproteobacteria ranging from the rhizoplane into the endosphere. However, the opposite trend was observed in the AMD pollution treatment, namely, Gammaproteobacteria were enriched, and Alphaproteobacteria and Deltaproteobacteria were mostly depleted. In addition, endophytic microbiomes were dominated by Comamonadaceae and Rhodocyclaceae in the unpolluted treatment and by Enterobacteriaceae in the AMD-polluted soils. PICRUSt showed that functional categories associated with membrane transport, metabolism and cellular processes and signaling processes were overrepresented in the endosphere of the AMD-polluted treatment. In conclusion, our study reveals significant variation in bacterial communities colonizing rhizo-compartments in two soils, indicating that plants can recruit functional bacteria to the roots in response to AMD pollution.