47,107 research outputs found
NOx Emissions Control Area (NECA) scenario for ports in the North Adriatic Sea
No full textIn response to global warming, the International Maritime Organisation (IMO) set rules of 50% Greenhouse Gas (GHG) reduction by 2050, from 2008 levels. Signatory countries to the IMO's regulation require frequent assessment of the contribution of GHG emissions from shipping calling at their ports or trading in their territorial waters to ensure their compliance with the regulations. This demands a rapid and accurate method to assess shipping's contribution to GHG emissions. Current methodologies for estimating emissions from ships can be described on a scale between bottom-up and top-down methods. Top-down methods provide rapid estimates – primarily based on fuel sales reports - without considering individual vessel details. Therefore, they are less accurate and do not provide a breakdown of emissions by ship types or in specific regions. Bottom-up methodologies are detailed vessel-based estimates; however, they are data and time-demanding. The Ship Emissions Assessment method (SEA) (Topic et al., 2021) fills the gap between bottom-up and top-down methods by providing an innovative hybrid solution for rapid but accurate ship emission estimation. It uses publicly available, cost-effective data sets for emission estimates. The SEA method is capable of estimating ships' emissions in designated areas to understand regulations' effectiveness and provide emission quantification evidence. This research objective was to apply the SEA method to quantify CO2, SOX and NOX exhaust emissions from containerships for the three crucial containership ports: Trieste, Rijeka and Venice, in the North of the Adriatic Sea. The SEA methodology was applied to assess emissions and forecast efficiency in scenarios of different regulatory measures. A reduction in NOx emissions was estimated for the event of the implementation of NECA in all three ports. Results showed that 447.13 tonnes of NOx could be reduced each year in the North Adriatic Sea area around the ports of Rijeka, Trieste and Venice in the event that NECA regulations are stipulated.</p
Assessment of ship emissions in coastal waters using spatial projections of ship tracks, ship voyage and engine specification data
No full textTo understand, mitigate and reduce the detrimental effects on human health and the environment from exhaust gas emissions from ships it is necessary to be able to estimate the quantity and location of these ship emissions in time.Currently, the two most commonly used ship emission assessment methods sit on opposite ends of the spectrum – the top-down approach, which provides low resolution yet efficient aggregated results however is unable to account for specific shipping activities, and the bottom-up vessel-by-vessel approach, which provides near-instantaneous ship emissions production at a high resolution – yet is data and time intensive.To address the market gap for a ship greenhouse emission estimation method that hybridises the best of both the bottom-up and top-down methods the novel Ship Emissions Assessment (SEA) method is proposed as an innovative hybrid solution.It is a cost effective and resource efficient method, presenting spatial ship emissions utilising widely accessible data, and it is precise – fulfilling the requirements needed to evaluate ship emissions reduction measures.Novel SEA method is the first in its endeavour to replace Automatic Identification System (AIS) Vessel-based raw data allocation, by using rapid analyses of readily available ship track density data and average voyage information. It combines obtained average voyage distance with voyage average speed to estimate ship activity for emission assessments - saving costs by reducing time and reliance on complex computations, especially when many ships need to be analysed simultaneously.Using the novel SEA method, a series of containerships from geographically diverse ports were sampled and assessed for emissions with comparative results confirming the representations equivalent to the detailed and data demanding bottom-up method.Subsequently, the novel SEA method was applied to containership traffic calling into the Port of Trieste, in the northern Adriatic Sea, where it demonstrated the ability to estimate and quantify historic emissions for the preceding 12 months while taking into account seasonal port traffic variations.The novel SEA method showed to be an efficient, inexpensive and accurate, easy-to-use emission assessment tool based on widely accessible data. It can be used in day-to-day shipping operations by a variety of stakeholders including port operations managers, regional traffic operators, and those non-industry, while providing the required level of technical accuracy. In comparison, existing methods are not as time and cost effective, user-friendly, nor based on easy to interpret and readily accessible data.The novel SEA method enables further global research of ship emissions, and for regional and international policy makers to effectively manage the measures needed to reach greenhouse gas emission reduction targets
Citation Author Topic Model in Expert Search
No full textThis paper proposes a novel topic model, Citation-Author-Topic (CAT) model that addresses a semantic search task we define as expert search – given a research area as a query, it returns names of experts in this area. For example, Michael Collins would be one of the top names retrieved given the query Syntactic Parsing. Our contribution in this paper is two-fold. First, we model the cited author informa-tion together with words and paper au-thors. Such extra contextual information directly models linkage among authors and enhances the author-topic association, thus produces more coherent author-topic distribution. Second, we provide a prelim-inary solution to the task of expert search when the learning repository contains ex-clusively research related documents au-thored by the experts. When compared with a previous proposed model (Johri et al., 2010), the proposed model pro-duces high quality author topic linkage and achieves over 33 % error reduction evaluated by the standard MAP measure-ment.
Probabilistic Author-Topic Models for Information Discovery
Get PDFWe propose a new unsupervised learning technique for extracting information from large text collections. We model documents as if they were generated by a two-stage stochastic process. Each author is represented by a probability distribution over topics, and each topic is represented as a probability distribution over words for that topic. The words in a multi-author paper are assumed to be the result of a mixture of each authors' topic mixture. The topic-word and author-topic distributions are learned from data in an unsupervised manner using a Markov chain Monte Carlo algorithm. We apply the methodology to a large corpus of 160,000 abstracts and 85,000 authors from the well-known CiteSeer digital library, and learn a model with 300 topics. We discuss in detail the interpretation of the results discovered by the system including specific topic and author models, ranking of authors by topic and topics by author, significant trends in the computer science literature between 1990 and 2002, parsing of abstracts by topics and authors and detection of unusual papers by specific authors. An online query interface to the model is also discussed that allows interactive exploration of author-topic models for corpora such as CiteSeer
Antisemitism in Serbia in the past and present
No full textPredavanje Tamare Stojanović “Antisemitizam u Srbiji u prošlosti i sadašnjosti” je drugo predavanje četvrtog kursa pod nazivom "Antisemitizam" koji se realizuje u okviru projekta „Holokaust u zemljama bivše Jugoslavije“. Ovaj projekat je prvi ovakav projekat u regionu i ima za cilj da pruži dublje razumevanje Holokausta na prostoru bivše Jugoslavije, podstakne kritičko mišljenje, promoviše toleranciju i empatiju, kao i da istakne konstantnu važnost obrazovanja o Holokaustu u regionu. Projekat se sastoji od četiri zasebna online kursa sa po tri predavanja i jednog šestodnevnog seminara uživo, koji se organizuje nakon završetka četvrtog kursa u Beogradu i Skoplju. Svi kursevi, kao i seminar, bave se temom Holokausta i antisemitizma na prostoru bivše Jugoslavije. Svaki od četiri online kursa obuhvata dragocena opšta znanja iz pomenutih tematskih okvira, ali i fokus na pojedine delove regiona i iskustvo Holokausta u njima.Tamara Stojanović's lecture "Antisemitism in Serbia in the past and present" is the second lecture of the fourth course entitled "Antisemitism", which is being realised within the project "Holocaust in the countries of the former Yugoslavia". This project is the first of its kind in the region and aims to deepen understanding of the Holocaust within the former Yugoslavia, encourage critical thinking, promote tolerance and empathy, and emphasise the ongoing importance of education about the Holocaust in the area. The project includes four online courses, each with three lectures, and one six-day live seminar, organised after the end of the fourth course in Belgrade and Skopje. All courses, as well as the seminar, address the topic of the Holocaust and anti-Semitism in the former Yugoslav territory. Each of the four online courses provides valuable general knowledge from the specified thematic frameworks, with a particular focus on certain parts of the region and their experiences of the Holocaust.Trajanje 1:43:09 minuta (duration 1:43:09 minutes).Projekat realizuje organizacija Haver Srbija u saradnji sa Fondom za Holokaust Jevreja iz Makedonije, uz podršku Claims Conference. Koordinatori projekta su Tamara i Aleksandar Stojanović. (The project is implemented by the Haver Serbia organization in cooperation with the Holocaust Fund of the Jews from Macedonia, with the support of the Claims Conference. The project coordinators are Tamara and Aleksandar Stojanović).YouTube link [https://www.youtube.com/watch?v=7T0x_h58a1U&list=PLhnejJShQmj6qZbBOZJmuu-BK74Qp4JVQ&index=7
Learning Author Topic Models from Text Corpora
No full textWe propose a new unsupervised learning technique for extracting information about authors and topics from large text collections. We model documents as if they were generated by a two-stage stochastic process. An author is represented by a probability distribution over topics, and each topic is represented as a probability distribution over words. The probability distribution over topics in a multi-author paper is a mixture of the distributions associated with the authors. The topic-word and author-topic distributions are learned from data in an unsupervised manner using a Markov chain Monte Carlo algorithm. We apply the methodology to three large text corpora: 150,000 abstracts from the CiteSeer digital library, 1,740 papers from the Neural Information Processing Systems Conference (NIPS), and 121,000 emails from a large corporation. We discuss in detail the interpretation of the results discovered by the system including specific topic and author models, ranking of authors by topic and topics by author, parsing of abstracts by topics and authors, and detection of unusual papers by specific authors. Experiments base
Novel Probabilistic Frameworks for Author-Level Topic Modeling
Get PDFThe increasing complexity of textual data in modern applications, such as social media and academic literature analysis, needs improved topic modeling techniques that capture sparsity, variability, and nuanced author-topic relationships. Because of their rigorous assumptions and inadequate adaptability in representing various data, traditional models generally fail to address these shortcomings. We present two novel probabilistic models, Author Dirichlet Multinomial Allocation with Generalized Distribution (ADMAGD) and Author Beta-Liouville Multinomial Allocation (ABLiMA) to overcome these drawbacks while strengthening the state of author-specific topic modeling. To depict complex author-topic relationships, ADMAGD incorporates the Generalized Dirichlet distribution. For datasets with uneven or absent topic representations, ABLiMA uses the Beta-Liouville distribution to adjust for topic distribution variability and sparsity. By comparing these models to common datasets like the NIPS and 20 Newsgroups datasets, the research presented here demonstrates how well these models manage sparsity, capture complex theme preferences, and generate coherent subjects. The results show that the models can be applied to many situations. Coherence measure and author-topic relationship visualizations further validate their interpretability and usefulness
Topic-based author cocitation analysis: A preliminary exploration
No full textMade available in DSpace on 2017-07-27T15:57:48Z (GMT). No. of bitstreams: 2
3.2_29_Bu-Topic-based Author Cocitation Analysis.pdf: 1489892 bytes, checksum: ee84d908b256d9bf79249ddacb58617f (MD5)
license.txt: 4813 bytes, checksum: 715c4321821a960fa1a1e91d2ac7ebce (MD5)
Previous issue date: 2017Author cocitation analysis (ACA) plays a significant role in mapping knowledge domains. However, it has been criticized to be relatively less informative because topic- and semantic-level information of citations has seldom been integrated into ACA. This poster aims to improve the traditional ACA by combining topical information of cocited authors with author cocited counts, which is called topic-based ACA. Author-Conference-Topic (ACT) model is adopted in this research to calculate topic distributions of authors. Compared with traditional ACA, topic-based ACA shows a better clustering ability in visualization and mines more details in knowledge domain mappings
Authorship Attribution with Author-aware Topic Models
No full textAuthorship attribution deals with identifying the authors of anonymous texts. Recently, we found that the Latent Dirichlet Allocation (LDA) topic model can be used to improve authorship attribution accuracy. We build on this finding and show that employing a previously-suggested Author-Topic (AT) model outperforms LDA when applied to scenarios with many authors. In addition, we define a model that combines LDA and AT by representing authors and documents over two disjoint topic sets, and show that our model outperforms LDA, AT and support vector machines on datasets with two to 19,320 authors
Literary Studies Topic Modeling: Metadata and Modeling Outputs
No full textNew analytical approaches, like topic modeling, can illuminate subtle transformations, revealing concepts, frequently taken for granted, to be more variable than scholars have assumed. In this study, the corpus that was modeled included 21,367 JSTOR articles and 13,221 distinct author names resulting in the 150-topic model
- …
