• Publications close
• Research data close
• Research software close
• Other research products close
• AU close
• Digital Humanities and Cultural Her... close

Locally-established marine protected areas (MPAs) have been proven to achieve local-scale fisheries and conservation objectives. However, since many of these MPAs were not designed to form ecologically-connected networks, their contributions to broader-scale goals such as complementarity and connectivity can be limited. In contrast, integrated networks of MPAs designed with systematic conservation planning are assumed to be more effective—ecologically, socially, and economically—than collections of locally-established MPAs. There is, however, little empirical evidence that clearly demonstrates the supposed advantages of systematic MPA networks. A key reason is the poor record of implementation of systematic plans attributable to lack of local buy-in. An intermediate scenario for the expansion of MPAs is scaling up of local decisions, whereby locally-driven MPA initiatives are coordinated through collaborative partnerships among local governments and their communities. Coordination has the potential to extend the benefits of individual MPAs and perhaps to approach the potential benefits offered by systematic MPA networks. We evaluated the benefits of scaling up local MPAs to form networks by simulating seven expansion scenarios for MPAs in the Verde Island Passage, central Philippines. The scenarios were: uncoordinated community-based establishment of MPAs; two scenarios reflecting different levels of coordinated MPA expansion through collaborative partnerships; and four scenarios guided by systematic conservation planning with different contexts for governance. For each scenario, we measured benefits through time in terms of achievement of objectives for representation of marine habitats. We found that: in any governance context, systematic networks were more efficient than non-systematic ones; systematic networks were more efficient in broader governance contexts; and, contrary to expectations but with caveats, the uncoordinated scenario was slightly more efficient than the coordinated scenarios. Overall, however, coordinated MPA networks have the potential to be more efficient than the uncoordinated ones, especially when coordinated planning uses systematic methods. shp - verde_puregion_governanceShapefile presenting the governance areas for each planning unit used in the model simulations and results described in the paper. The file can be used to present the data included in the paper to create maps.shp - verde_puregion_habitatsShapefile describing the area of habitats present within each planning unit. This file was used in the data analysis for the models described in the paper. This area of habitat described in the file are in square metres.DataZip folder containing processed data, model results and source code of the models described in the paper. The excel sheets can be used as attributes in the SHP files provided to create maps. The text file included in the folder is the the source code for the models and can be used and edited in MatLab.

Fire has a long history in Australia and is a key driver of vegetation dynamics in the tropical savanna ecosystems that cover one quarter of the country. Fire reconstructions are required to understand ecosystem dynamics over the long term but these data are lacking for the extensive savannas of northern Australia. This paper presents a multiproxy palaeofire record for Marura sinkhole in eastern Arnhem Land, Northern Territory, Australia. The record is constructed by combining optical methods (counts and morphology of macroscopic and microscopic charcoal particles) and chemical methods (quantification of abundance and stable isotope composition of pyrogenic carbon by hydrogen pyrolysis). This novel combination of measurements enables the generation of a record of relative fire intensity to investigate the interplay between natural and anthropogenic influences. The Marura palaeofire record comprises three main phases: 4600–2800 cal BP, 2800–900 cal BP and 900 cal BP to present. Highest fire incidence occurs at ~4600–4000 cal BP, coinciding with regional records of high effective precipitation, and all fire proxies decline from that time to the present. 2800–900 cal BP is characterised by variable fire intensities and aligns with archaeological evidence of occupation at nearby Blue Mud Bay. All fire proxies decline significantly after 900 cal BP. The combination of charcoal and pyrogenic carbon measures is a promising proxy for relative fire intensity in sedimentary records and a useful tool for investigating potential anthropogenic fire regimes.

Datasets used for RE (Relation Extraction) LLL IEPA HPRD50 BioInfer LLL Abstracts E_Coli RegulonDB Abstracts

This dataset contains the sentiment analysis (SA) of original tweets posted in English by users related to the 10th anniversary of the 2010 Haitian earthquake. Tweets are classified according to their polarity or not related. This classification includes supervised and unsupervised classification. This dataset compares the accuracy (ACC) of three tools for unsupervised text classification: a no-code machine learning (ML) classification platform: ‘MonkeyLearn’ and two trained models finetuned for SA: ‘troberta’ and ‘btweet’. These last ones are language models based on RoBERTa (https://aclanthology.org/2020.findings-emnlp.148/) and BERTweet (https://aclanthology.org/2020.emnlp-demos.2/) architecture, respectively. Both models are available in the platform: Hugging Face. The first author performed the supervised classification and trained the tweets on the MonkeyLearn platform at the tweet level using samples of 1, 5 and 10 per cent of the tweets in the dataset (excluded to test ACC in the prediction). This supervised classification is compared to the unsupervised classification performed by ‘MonkeyLearn’, ‘troberta’ and ‘btweet’. We can observe that the average confidence in the classification increase with the number of trained tweets in the case of ‘MonkeyLearn’ (0.39, 0.56 and 0.64) while the average confidence in their own classification by troberta (0.89) and btweet (0.92) is very high and higher than MonkeyLearn’s average confidence.

Appendix 14 provides a brief description and other data about 708 of the ASAGE sites documented in the ten Study Areas. It does not provide specific latitude / longitude coordinates, so as not to facilitate looting of these sites, although the spatial data are available to bona fide researchers.

This dataset contains coordinates of building damage reported by Twitter users after the 22nd March 2020 Zagreb earthquake. Coordinates are linked to tweets classified (supervised) at this level according to their polarity (Sentiment analysis) and topic (topic analysis) that they address.

Supplemental material, sj-xlsx-4-anr-10.1177_20530196221135077 for Beppu Bay, Japan, as a candidate Global Boundaries Stratotype Section and Point for an Anthropocene series by Michinobu Kuwae, Bruce P Finney, Zhiyuan Shi, Aya Sakaguchi, Narumi Tsugeki, Takayuki Omori, Tetsuro Agusa, Yoshiaki Suzuki, Yusuke Yokoyama, Hirofumi Hinata, Yoshio Hatada, Jun Inoue, Kazumi Matsuoka, Misaki Shimada, Hikaru Takahara, Shin Takahashi, Daisuke Ueno, Atsuko Amano, Jun Tsutsumi, Masanobu Yamamoto, Keiji Takemura, Keitaro Yamada, Ken Ikehara, Tsuyoshi Haraguchi, Stephen Tims, Michaela Froehlich, Leslie Keith Fifield, Takahiro Aze, Kimikazu Sasa, Tsutomu Takahashi, Masumi Matsumura, Yukinori Tani, Peter R Leavitt, Hideyuki Doi, Tomohisa Irino, Kazuyoshi Moriya, Akira Hayashida, Kotaro Hirose, Hidekazu Suzuki and Yoshiki Saito in The Anthropocene Review

Anthology paper link: https://aclanthology.org/2021.emnlp-main.270/ Abstract: Entity Alignment (EA) aims to match equivalent entities across different Knowledge Graphs (KGs) and is an essential step of KG fusion. Current mainstream methods ��� neural EA models ��� rely on training with seed alignment, i.e., a set of pre-aligned entity pairs which are very costly to annotate. In this paper, we devise a novel Active Learning (AL) framework for neural EA, aiming to create highly informative seed alignment to obtain more effective EA models with less annotation cost. Our framework tackles two main challenges encountered when applying AL to EA: (1) How to exploit dependencies between entities within the AL strategy. Most AL strategies assume that the data instances to sample are independent and identically distributed. However, entities in KGs are related. To address this challenge, we propose a structure-aware uncertainty sampling strategy that can measure the uncertainty of each entity as well as its impact on its neighbour entities in the KG. (2) How to recognise entities that appear in one KG but not in the other KG (i.e., bachelors). Identifying bachelors would likely save annotation budget. To address this challenge, we devise a bachelor recognizer paying attention to alleviate the effect of sampling bias. Empirical results show that our proposed AL strategy can significantly improve sampling quality with good generality across different datasets, EA models and amount of bachelors.

https://youtu.be/7FPiqHvlMxY The Australian Museum acquired three cartonnage masks and a foot-case in the early 20th century. The cartonnage pieces have been recently restored and they are ready for display. Quite fragile compared to wooden coffins or stone sarcophagi, many cartonnage masks survived surprisingly well preserved for some millennia. They proliferated in the Greco-Roman period of Egyptian history and hundreds were recovered from ancient cemeteries. Relatively little research was completed on these funerary accessories but in recent times scholars have begun to show more interest, as is attested by a study conducted by Dr. Carlo Rindi Nuzzolo who examined our masks in April 2022. Cartonnage (cardboard) - a word derived from French – describes a material made of layers of linen or papyrus covered with plaster and used in ancient Egyptian funerary masks and full-body casing from the First Intermediate Period (c. 2181–2055 BCE) to the Roman era (332 BC -395 AD). https://youtu.be/7FPiqHvlMxY

The global internet antiquities market exists in a complex cultural heritage framework, comprised of international law and domestic legislation. In this paper, the questions I seek to answer are the following: how do internet antiquities dealers engage with their legal obligations, and how is this engagement translated to the ethics of their businesses? This paper presents a comparative examination of 45 antiquities dealers split across three categories—internet dealers, eBay dealers and social media dealers—revealing three key insights about the internet antiquities market: firstly, that the level of legal literacy in the market is depicted as being quite poor; secondly, that the performance of legal awareness does not always correspond with ethical dealer practices; and finally, some dealers utilise a suite of justifications for their behaviours, practices and values (known as neutralisation techniques) to undermine their legal obligations. Such results confirm existing claims of the failure of self-regulation in the internet antiquities market and reveal a demand for educational campaigns targeted at raising consumer awareness by challenging misleading market narratives and highlighting the ethical and legal issues involved with the trade of cultural heritage.