Climate change has the potential to enhance or disrupt biological systems, but currently, little is known about how organism plasticity may facilitate adaptation to localised climate variation. The bee-flower relationship is an exemplar signal-receiver system that may provide important insights into the complexity of ecological interactions in situations like this. For example, several studies on bee temperature preferences show that bees prefer to collect warm nectar from flowers at low ambient temperatures, but switch their preferences to cooler flowers at ambient temperatures above about 30° C. We used temperature sensor thermal probes to measure the temperature of outdoor flowers of 30 plant species in the Southern regions of the Australian mainland, to understand how different species could modulate petal temperature in response to changes in ambient temperature and, potentially, influence the decision-making of bees in the flowering plant’s favour. We found that flower petal temperatures respond in different ways to changing ambient temperature: linearly increasing or decreasing relative to the ambient temperature, dynamically changing in a non-linear manner, or varying their temperature along with the ambient conditions. For example, our investigation of the difference between ambient temperature and petal temperature (ΔT), and ambient temperature, revealed a non-linear relationship for Erysimum linifolium and Polygala grandiflora that seems suited to bee temperature preferences. The temperature profiles of species like Hibertia vestita and H. obtusifolia appear to indicate that they do not have a cooling mechanism. These species may therefore be less attractive to bee pollinators in changing climatic conditions with ambient temperatures increasingly above 30° C. This may be to the species’ detriment when insect-pollinator mediated selection is considered. However, we found no evidence that flower visual characteristics used by bees to identify flowers at close range, such as colour or shape, were straightforward modulators of floral temperature. We could not identify any clear link to phylogenetic history and temperature modulation either. Mapping our test flower distribution on the Australian continent however, indicates a potential clustering that suggests different flower responses may constitute adaptations to local conditions. Our study proposes a framework for modelling the potential effects of climate change and floral temperature on flower pollination dynamics at local and global scales.
A publicly available database and interactive map presenting ambient background soil concentrations and other soil characteristics for Victoria, Australia. The interactive portal include summary statistics for ambient background metal/element concentrations in surface (0.0-0.1m), sub-surface (0.3-0.6m) and deeper (>0.6m) soils, grouped by expected underlying geology and region.
Naturally enriched ambient background concentrations of metals and elements have been reported in soils of Victoria, Australia. Where natural enrichment is not accurately distinguished from anthropogenic impacts, soils can be inappropriately categorised as “contaminated waste”; resulting in unnecessary disposal to landfill. Publicly available and new background soil data was collated by RMIT University for the purpose of developing improved methods for estimation of background concentrations in soils (methods described in Mikkonen et al. 2017).
Ambient background concentrations are defined as the sum of the geogenic concentration of the element plus diffuse anthropogenic contamination that has been introduced from non-point sources. There is lack of consistency on what constitutes diffuse contamination or a point source. For the purpose of this research, ambient background concentrations include human contributions of contaminants through diffuse inputs such as atmospheric deposition of lead from the broad use of leaded fuels. However, lead impacts directly associated with an adjacent road (i.e. within 25 m of the road) were not considered representative of ambient background concentrations. Consistent with background studies undertaken across Europe, broad application of fertilisers during typical agricultural practices (excluding horticulture and application of biosolids), were considered representative of ambient background concentrations.
Crude pyrolysis bio-oil contains significant quantities of carboxylic acids which limit its utility as a biofuel. Vapour phase ketonisation of organic acids contained within biomass fast-pyrolysis vapours offers a potential pre-treatment to improve the stability and energy content of resulting bio-oils formed upon condensation. Zirconia is a promising catalyst for such reactions, however little is known regarding the impact of thermal processing on the physicochemical properties of zirconia in the context of it's corresponding reactivity for the vapour phase ketonisation of acetic acid. Here we show that calcination progressively transforms amorphous Zr(OH)4 into small tetragonal ZrO2 crystallites at 400 °C, and subsequently larger monoclinic crystallites >600 °C. These phase transitions are accompanied by an increase in the density of Lewis acid sites, and concomitant decrease in their acid strength, attributed to surface dehydroxylation and anion vacancy formation. Weak Lewis acid sites (and/or resulting acid–base pairs) are identified as the active species responsible for acetic acid ketonisation to acetone at 350 °C and 400 °C, with stronger Lewis acid sites favouring competing unselective reactions and carbon laydown. Acetone selectivity is independent of acid strength.
Chemo‐ and regioselectivity in a heterogeneously catalyzed cross aldol reaction were directed by tuning the nature of the sites, textural properties, and reaction conditions. Catalysts included sulfonic acid‐functionalized resins or SBA‐15 with varying particle size or pore diameter, H‐BEA zeolites, and Sn‐BEA zeotype; conditions were 25 °C to 170 °C in organic media. Benzaldehyde and 2‐butanone yielded branched (reaction at ‐CH2‐ of butanone) and linear (reaction at ‐CH3) addition and condensation products; and fission of the branched aldol led to β‐methyl styrene and acetic acid. Strong acids promoted the dehydration step, and regioselectivity originated from preferred formation of the branched aldol. Both, resins and functionalized SBA‐15 materials yielded predominantly the branched condensation product, unless particle morphology or temperature moved the reaction into the diffusion‐limited regime, in which case more fission products were formed, corresponding to Wheeler Type II selectivity. For H‐form zeolites, fission of the branched aldol competed with dehydration of the linear aldol, possibly because weaker acidity or steric restrictions prevented dehydration of the branched aldol.
How different visual systems process images and make perceptual errors can inform us about cognitive and visual processes. One of the strongest geometric errors in perception is a misperception of size depending on the size of surrounding objects, known as the Ebbinghaus or Titchener illusion. The ability to perceive the Ebbinghaus illusion appears to vary dramatically among vertebrate species, and even populations, but this may depend on whether the viewing distance is restricted. We tested whether honeybees perceive contextual size illusions, and whether errors in perception of size differed under restricted and unrestricted viewing conditions. When the viewing distance was unrestricted, there was an effect of context on size perception and thus, similar to humans, honeybees perceived contrast size illusions. However, when the viewing distance was restricted, bees were able to judge absolute size accurately and did not succumb to visual illusions, despite differing contextual information. Our results show that accurate size perception depends on viewing conditions, and thus may explain the wide variation in previously reported findings across species. These results provide insight into the evolution of visual mechanisms across vertebrate and invertebrate taxa, and suggest convergent evolution of a visual processing solution.
The test data related to the article consists of three files: one contains all bee choices for the 20 choices of each test (learning, transfer 1 and transfer 2) for bees in the unrestricted viewing condition; one contains all choices for each bee during the 80 conditioned choices of the learning phase for bees trained under both restricted and unrestricted viewing conditions; the final file contains the 20 choices each bee made during all tests (learning, transfer 1 and transfer 2) in the restricted viewing condition using the y-maze.
Evaluation of the effectiveness of chiropractic care for the Indigenous community in Kempsey.
Project revolved around embedding a chiropractic program into an existing Indigenous community health service and evaluation of the impact of the program.
Data set comprises of 14 audio interviews with participants of the program, discussing their experience and treatment.
Other data around the project is archived and only available from researcher with permission.
Data relates to published article "Multiple dispersal vectors drive range expansion in an invasive marine species", and includes microsatellite genotypes, structure input and parameter files, population genetic analysis script and dispersion model outputs.
The establishment and subsequent spread of invasive species is widely recognized as one of the most threatening processes contributing to global biodiversity loss. This is especially true for marine and estuarine ecosystems, which have experienced signifi- cant increases in the number of invasive species with the increase in global maritime trade. Understanding the rate and mechanisms of range expansion is therefore of significant interest to ecologists and conservation managers alike. Using a combination of population genetic surveys, environmental DNA (eDNA) plankton sampling and hydrodynamic modelling, we examined the patterns of introduction of the predatory Northern Pacific seastar (Asterias amurensis) and pathways of secondary spread within southeast Australia. Genetic surveys across the invasive range reveal some genetic divergence between the two main invasive regions and no evidence of ongoing gene flow, a pattern that is consistent with the establishment of the second invasive region via a human-mediated translocation event. In contrast, hydrodynamic modelling combined with eDNA plankton sampling demonstrated that the establishment of range expansion populations within a region is consistent with natural larval dispersal and recruitment. Our results suggest that both anthropogenic and natural dispersal vectors have played an important role in the range expansion of this species in Australia. The multiple modes of spread combined with high levels of fecundity and a long larval duration in A. amurensis suggests it is likely to continue its range expansion and significantly impact Australian marine ecosystems.
"On Friday February 3rd, 2017, Aussie Rules Football changed forever.
There were twenty-four and half thousand people packed into the stands at Princes Park in Melbourne that night.
And when the siren sounded, and the ball was thrown into the air for the first time, they erupted, their roar an expression of their passion, of their joy, of long-held dreams realised.
Women have been playing football for more than 100 years. The first recorded game was played in Perth in 1915. More games followed, albeit sporadically, until state based leagues began launching from the early 1980s. The first, the Victorian Women’s Football League in 1981.
In 2010, the league commissioned a report into women’s football, a series of exhibition matches starting in 2013 grew from that report.
It was the success of those exhibition matches that encouraged the AFL to bring forward their plans for a national women’s football league and so the AFLW was born.
There are thousands of stories from that inaugural season. Feats of strength and skill and endurance, both on field and off. Tales of resilience and determination, of dreams dashed and rebuilt. But their genesis was that warm Friday night in February when thirty-two women clad in the navy blue of Carlton and the black and white stripes of Collingwood took to the field.
The First Friday in February will take you back to that night, reliving that first AFLW game and all that it meant through the memories of women who were there.
They each have a story, their story, of that First Friday in February."
This submission contains the raw rheometer files from time sweep and creep test measurements of waste activated sludge (total solids content = 4.2 wt %). The sludge samples were measured at various gas flow rates (0.54 LPM and 1.5 LPM). The dynamic time sweep measurements were generated from the sludge sample first without gas at 0.08% strain across 25 minutes. Then the sample was conditioned and the gas was injected for 20 minutes before the measurement at similar conditions. Similarly, the dynamic creep test measurements were generated from the sludge sample first without gas at 0.89pa stress across 10 minutes.
The measurements were carried out on a DHR-3 rheometer (TA-Instruments), equipped custom made cup.
For detail procedure refer: Bobade, V, et al. 2017, 'Impact of gas injection on the apparent viscosity and viscoelastic property of waste activated sewage sludge', Water Research, vol. 114, pp. 296-307.
Detailed experimental protocol is recorded in the .tri (TRIOS) file attached with each data set. This file can be read using the TRIOS software made available on TA-Instruments website (http://www.tainstruments.com/support/software-downloads-support/downloads/). Rheometer geometry dimensions and other details are also obtainable from these files.
This repository contains the dataset used to analyze user preferences of podcast summaries. The study is described in this paper.
We provide all the releasable data:
Query biased summaries, for both text and audio channels
Relevance and preference judgments
We also release a software package to download the copyrighted content: Audio podcasts (mp3), Manual transcripts.
We address the challenge of extracting "query biased audio summaries" from podcasts to support users in making relevance decisions in spoken document search via an audio-only communication channel. We performed a crowdsourced experiment that demonstrates that transcripts of spoken documents created using Automated Speech Recognition (ASR), even with significant errors, are effective sources of document summaries or "snippets" for supporting users in making relevance judgments against a query. In particular, results show that summaries generated from ASR transcripts are comparable, in utility and user-judged preference, to spoken summaries generated from error-free manual transcripts of the same collection. We also observed that content-based audio summaries are at least as preferred as synthesized summaries obtained from manually curated metadata, such as title and description. We describe a methodology for constructing a new test collection which we have made publicly available.
Attached file provides supplementary data for linked article.
If a paper-based analytical device (mu-PAD) could be made by printing indicators for detection of heavy metals in chemical symbols of the metals in a style of the periodic table of elements, it could be possible for such mu-PAD to report the presence and the safety level of heavy metal ions in water simultaneously and by text message. This device would be able to provide easy solutions to field-based monitoring of heavy metals in industrial wastewater discharges and in irrigating and drinking water. Text-reporting could promptly inform even nonprofessional users of the water quality. This work presents a proof of concept study of this idea. Cu(II), Ni(II), and Cr(VI) were chosen to demonstrate the feasibility, specificity, and reliability of paper-based text-reporting devices for monitoring heavy metals in water.
Two main sources of data for species distribution models (SDMs) are site-occupancy (SO) data from planned surveys, and presence-background (PB) data from opportunistic surveys and other sources. SO surveys give high quality data about presences and absences of the species in a particular area. However, due to their high cost, they often cover a smaller area relative to PB data, and are usually not representative of the geographic range of a species. In contrast, PB data is plentiful, covers a larger area, but is less reliable due to the lack of information on species absences, and is usually characterised by biased sampling. Here we present a new approach for species distribution modelling that integrates these two data types. We have used an inhomogeneous Poisson point process as the basis for constructing an integrated SDM that fits both PB and SO data simultaneously. It is the first implementation of an Integrated SO-PB Model which uses repeated survey occupancy data and also incorporates detection probability. The Integrated Model's performance was evaluated, using simulated data and compared to approaches using PB or SO data alone. It was found to be superior, improving the predictions of species spatial distributions, even when SO data is sparse and collected in a limited area. The Integrated Model was also found effective when environmental covariates were significantly correlated. Our method was demonstrated with real SO and PB data for the Yellow-bellied glider (Petaurus australis) in south-eastern Australia, with the predictive performance of the Integrated Model again found to be superior. PB models are known to produce biased estimates of species occupancy or abundance. The small sample size of SO datasets often results in poor out-of-sample predictions. Integrated models combine data from these two sources, providing superior predictions of species abundance compared to using either data source alone.
Associate Professor Adrian Miles was co–director of the non/fictionLab, and Deputy Dean Learning and Teaching in the School of Media and Communication at RMIT University, in Melbourne, Australia. He was formerly the Program Director of the consilience Honours lab. He was also a senior new media researcher in the InterMedia Lab at the University of Bergen, Norway.
Miles had extensive experience in the development, implementation and assessment of innovative pedagogies in new media and media studies education. Miles undertook substantial research and development in desktop networked interactive technologies, using 'off the shelf' software and systems to develop new knowledge genres and objects.
His academic research on hypertext and networked interactive video is widely published. Research interests included hypertext, new pedagogies for media education, digital humanities and media, digital video poetics, interactive documentary, and computational nonfiction. There is a strong theoretical theme through all this work (Deleuze's cinema philosophy, recent research on new materialism, and avatars of these.) He was the first or second person in the world to videoblog.
This research data collection will comprise of a listing of his physical academic book collection, an archive of his publicly available vogmae site, data from his work archives, and images of his office and book collection. This archive is currently being put together by RMIT staff.
Attached file provides supplementary data for linked article.
Related paper seeks to provide novel insights into the effects of ethnic communities on immigrants’ entrepreneurial activities. We investigate to what extent the decision of an employed immigrant to become an entrepreneur is associated with his or her embeddedness in ethnic networks in the host region. We capture such embeddedness through various mechanisms. Using longitudinal-registered data from Sweden and employing a logit model, we find that merely being located in an ethnic community does not have an influence on immigrant entrepreneurship; rather, what matters is being located in ethnic communities that have a high share of entrepreneurs themselves.
This collection comprises of print and electronic versions of catalogues related to exhibitions held in these two spaces between 2000-2018.
Australia's RMIT University is ranked 16th in the world for art and design, which positions it as the top art school in the country. The School of Art fosters innovative and diverse art practice and research within visual art, fine art, public art, and arts management
PROJECT SPACE is a gallery that links prominent exemplars of practice-based research with Melbourne’s creative communities through a dynamic program of contemporary art projects. Housed in RMIT University's Building 94, it runs a dynamic program of exhibitions and events from February to December.
SPARE ROOM is a gallery space adjacent to PROJECT SPACE in RMIT's Building 94. Both of these gallery programs are aligned such that exhibitions launch and run together and are often conceptually linked. SPARE ROOM acts as a small space for large ideas—it is a keyhole into a rich variety of potential conversations, from blue skies to deep dives.
This is a collection of holdings pertaining to Abbas Kiarostami, an internationally acclaimed and prominent Iranian filmmaker. He was involved in over 40 films, and is considered to be part of a generation of filmmakers called the Iranian New Wave. His films include Close-Up (1990), Taste of Cherry (1997), The Wind Will Carry Us (1999), Ten (2002) and Certified Copy (2010).
Insect Records in the City of Melbourne from the Little Things that Run the City Project. This data identifies what individual species were found in the Little Things that Run the City Project and is not a count of all species recorded. This data was obtained from Luis Mata, a postdoctoral researcher at RMIT. The data is owned by the City of Melbourne and RMIT. The project started in October 2014 with the first survey of insects for this project in parks and gardens in the City of Melbourne. More surveys were conducted between January and March 2015. Insect species were sorted and identified from April to September 2015. This is the raw data for the report that was published in late 2016.
Some records are lacking information on the species classification level, and may have the correct genus information accompanied with a number in the identification notes column - this donates species 1, 2, or 3 of that genus where the species could not be identfifed. All records lack a species name as this was not provided to the City of Melbourne in the original dataset.
This project explored neighbourhood experiences of residents in two ethno-religiously diverse suburbs in Melbourne’s north, Fawkner and Broadmeadows. The two localities were chosen because they both have large Muslim minorities (25 and 30 per cent respectively at the time of the 2011 Census, and 32 and 36 at the time of the 2016 census) and the project’s primary focus was on the impact of (primarily Muslim) ‘religious visibility’ on the local bridging social capital.
Bridging social capital is an important aspect of social capital, especially in large, diverse and socially anonymous urban contexts. It refers to interactions and connections among people with different demographic, ethno-cultural and socio-economic characteristics. Bridging social capital is crucial for (but not limited to) local community cohesion, which translates into friendliness, neighbourliness and safety of (sub)urban communities.
Literature on social cohesion and social capital in the context of ethnically diverse Western cities is extensive and its findings are varied, depending on the specific characteristics of the local context under investigation, as well as wider national and international contexts at any given time. Some studies found that urban ethnic diversity tends to decrease social capital and social cohesion, while other studies came up with different conclusions. One of the reasons for the inconclusive findings is not just real differences between localities but also methodological difficulty of precisely measuring social capital and social cohesion. Our study was informed by theoretical and methodological insights of Australian and overseas studies, as well as our own earlier research on diverse neighbourhoods and the impact of ‘visible difference’ in Australian urban contexts. The project also built on our recent (2012-13) empirical research in Melbourne’s diverse north.
The project employed mixed methodology of data collection and analysis. We started with the background analysis of the Census and other available quantitative information about the two localities. The empirical data collection for the project developed in three stages:
• key informant interviews (May-June 2016)
• the survey of residents (September-November 2016)
• follow-up interviews with residents (December 2016-March 2017).
The key informant interviews (the total of 16) were conducted by principal investigators and targeted professionals and services providers working in Fawkner and Broadmeadows. These interviews helped identify key issues and sharpened our research focus. They also informed the development of the key research instrument, the survey questionnaire. The total of 301 residents participated in the survey.
Once the questionnaire was drafted and piloted, seven community-based bilingual research assistants (BRAs) were engaged to administer the survey face-to-face to a sample of residents of the two suburbs. The sample was balanced by suburb, gender and religion (Muslim / non-Muslim). The sample was appropriately diverse in terms of participants’ ethnicity, length of residence in their suburb and socio-economic background. The sample is not representative for the population of the two suburbs, because Muslims were overrepresented in our sample.
The follow-up in-depth interviews (the total of 36) were conducted with residents of the two suburbs who previously took part in the survey and expressed interest to be interviewed later on. The sample included people from a variety of ages, ethnic backgrounds, religions and walks of life.
The large data set consisted of quantitative and qualitative data. Quantitative data were checked for quality, entered into the Statistical Package for Social Sciences (SPSS) and cleaned. We used descriptive and inferential statistical methods to analyse the data. The narrative interview data were professionally transcribed and thematically analysed. The quantitative and qualitative analyses are complementary and combining them is the best way to contribute to an in-depth understanding of the phenomena under investigation.
Ubiquitous data Exploration (UTE) is a mobile sensor data collection, annotation and exploration platform. Our platform facilitates rapid prototyping of data mining experiments by using a flexible and do-it-yourself approach. The platform allows researchers to quickly design and deploy applications on mobile devices in order to record sensor data and the corresponding ground-truth information. The platform is supported by a web interface for designing data collection experiments, synchronizing and storing the sensor data with the corresponding labels, and sharing data.
This repository is dedicated to Mobi-UTE, the iOS-app component of UTE platform.
For the Android-app component of UTE platform, the link is here: https://github.com/cruiseresearchgroup/mobiute-android
For the backend-app component of UTE platform, the link is here:
Water vapor-weighted mean temperature, Tm, is a vital parameter for retrieving precipitable water vapor (PWV) from the zenith wet delay (ZWD) of Global Navigation Satellite Systems (GNSS) signal propagation. In this study, the Tm at 368 GNSS stations for 2000-2012 were calculated using three methods: (1) temperature and humidity profiles from ERA-Interim, (2) the Bevis Tm-Ts relationship, and (3) the Global Pressure and Temperature 2 wet model. Tm derived from the first method was used as a reference to assess the errors of the other two methods. Comparisons show that the relative errors of the Tm derived from these two methods are in the range of 1-3% across more than 95% of all the stations. The PWVs were calculated using the aforementioned three types of Tm and the GNSS-derived ZWD at 107 stations. Again, the PWVs calculated using Tm from the first method were used as the reference of the other two PWVs. The root-mean-square errors of these two PWVs are both in the range of 0.1-0.7 mm. The second method is recommended in real-time applications, since its performance is slightly better than the third method. In addition, the linear trends of the PWV time series from the first method were also used as the reference to evaluate the trends from the other two methods. Results show that 13% and 23% of the PWV trends from the respective second and third methods have a relative error of larger than 10%. For climate change studies, the first method, if available, is always recommended.
This is supplementary data for the published paper on Nano Letters.
The exhibition of plasmon resonances in two-dimensional (2D) semiconductor compounds is desirable for many applications. Here, by electrochemically intercalating lithium into 2D molybdenum disulfide (MoS2) nanoflakes, plasmon resonances in the visible and near UV wavelength ranges are achieved. These plasmon resonances are controlled by the high doping level of the nanoflakes after the intercalation, producing two distinct resonance peak areas based on the crystal arrangements. The system is also benchmarked for biosensing using bovine serum albumin. This work provides a foundation for developing future 2D MoS2 based biological and optical units
Attached file provides supplementary data for linked article.
Temperature, solar radiation and water are major important variables in ecosystem models which are measurable via wireless sensor networks (WSN). Effective data analysis is necessary to extract significant spatial and temporal information. In this work, information regarding the long term variation of seasonal field environment conditions is explored using Hilbert-Huang transform (HHT) based analysis on the wireless sensor network data collection. The data collection network, consisting of 36 wireless nodes, covers an area of 100 square kilometres in Yanqing, the northwest of Beijing CBD, in China and data collection involves environmental parameter observations taken over a period of three months in 2011. The analysis used the empirical mode decomposition (EMD/EEMD) to break a time sequence of data down to a finite set of intrinsic mode functions (IMFs). Both spatial and temporal properties of data explored by HHT analysis are demonstrated. Our research shows potential for better understanding the spatial-temporal relationships among environmental parameters using WSN and HHT.
This repository contains resources used and described in the paper. The repository is structured as follows:
data/: Dataset used for this paper.
code/: Include the algorithm implementation.
paper/: PDF of paper.
C++ with Visual Studio 2012 Please run from main.cpp. All setting are in the main.cpp. Please place the data file in the same direction.
The data include sample data comprises of the location of the online parking slots and the parking events. Please contact me if you needs the complete data.
The on-street parking system is an indispensable part of civic projects, as it provides travellers and shoppers with parking spaces. With the recent in-ground sensors deployed throughout the Melbourne CBD, there is a significant problem on how to use the sensor data to manage parking violations and issue infringement notices efficiently in a short time-window. In this paper, we use a large real-world dataset with on-street parking sensor data from the local city council, and establish a formulation of the Travelling Officer Problem with a general probability-based model. We propose two solutions using a spatiotemporal probability model for parking officers to maximize the number of infringing cars caught with limited time cost. Using real-world parking sensor data and Google Maps road network information, the experimental results show that our proposed algorithms outperform the existing patrolling routes.
Event detection on Twitter has attracted active research. Although existing work considers the semantic topic structure of documents for event detection, the topic dynamics and the semantic consistency are under-investigated. In this paper, we study the problem of topical event detection in tweet streams. We define topical events as the bursty occurrences of semantically consistent topics. We decompose the problem of topical event detection into two components: (1) We address the issue of the semantic incoherence of the evolution of topics. We propose to improve topic modelling to filter out semantically inconsistent dynamic topics. (2) We propose to perform burst detection on the time series of dynamic topics to detect bursty occurrences. We apply our proposed techniques to the real world application by detecting topical events in public transport tweets. Experiments demonstrate that our approach can detect the newsworthy events with high success rate.
Provided link supports the dataset used for this paper.
This repository contains resources used and described in the paper.
The repository is structured as follows:
NOTE: this should follow the folders you have in this repository
algorithms/: Formal description of algorithm for entity normalization and sentence clustering.
data/: Dataset used for this paper.
code/: Evaluation script.
evaluation/: Evaluation script.
presentation/: PDF of paper presentation in certain conference or venue.
In this paper, Extreme Learning Machine (ELM) is demonstrated to be a powerful tool for electricity consumption prediction based on its competitive prediction accuracy and superior computational speed compared to Support Vector Machine (SVM). Moreover, ELM is utilized to investigate the potentials of using auxiliary information such as electricity-related factors and environmental factors to augment the prediction accuracy obtained by purely using the electricity consumption factors. Furthermore, we formulate a combinatorial optimization problem of seeking an optimal subset of auxiliary factors and their corresponding optimal window sizes using the most suitable ELM structure, and propose a Discrete Dynamic Multi-Swarm Particle Swarm Optimization (DDMS-PSO) to address this problem. Experimental studies on a real-world building dataset demonstrate that electricity-related factors improve accuracy while environmental factors further boost accuracy. By using DDMSPSO, we find a subset of electricity-related and environmental factors, their respective window sizes, and the number of hidden neurons in ELM which leads to the best prediction accuracy.