Filename
stringlengths 22
64
| Paragraph
stringlengths 8
5.57k
|
|---|---|
Processed_Contradicting_Motivations_in_Civic_Tech_Software_D.txt | In this paper, we analyzed what motivation, motive, or goal contradictions emerge in a grassroots civic tech project in action. The main contradictions occurred inside activity systems on details of implementation or between system motives, instead of big picture motivations. Two most significant contradictions involved planning, and converging on design and technical approaches. |
Processed_Contradicting_Motivations_in_Civic_Tech_Software_D.txt | Our recommendations for practice are what perhaps pro- fessionals in established volunteering and non-government organisations already know: Have skilled and committed facilitators that can keep volunteers pointed in the same direction and sustain participant engagement. Based on our case study, it appears that it is essential to have: efficient communicators, shared task management tools, and a compro- mise on technologies that most participants can commit to. Finding an acceptable lowest common technology denominator is important, since participant skill levels in civic tech can vary [35, 36]. |
Processed_Contradicting_Motivations_in_Civic_Tech_Software_D.txt | We extend the body of knowledge in civic tech software engi- neering through two main contributions. First, we demonstrate activity theory can be used as a theoretical lens to evaluate software engineering methods in action [15, 16] as evolving, in- action activity systems. Second, we demonstrate that examining participant motives together as an activity system can shed light on contradictions that arise during civic tech or tech for good development. |
Processed_Contradicting_Motivations_in_Civic_Tech_Software_D.txt | The trade-off between contextual richness of a single case study and generalizability of a multiple case study points towards a potential limitation in our work. However, seminal literature [37, 38] acknowledge single case study as a method of inquiry, especially when undertaking novel, exploratory research to capture the complexity of the context in which the case is situated (i.e. motives and motivations within activity systems in complex civic tech environments and processes). Looking forward, we advocate for more comparisons between different civic tech software engineering activity systems for broader findings and more generalizability. |
Processed_Contradicting_Motivations_in_Civic_Tech_Software_D.txt | We thank Dr. Victoria Palacin for her conceptual contribu- tions and research contributions to the authors’ earlier research collaboration. This research is based on the line of thinking that started in collaboration with Dr. Palacin. |
Processed_Contradicting_Motivations_in_Civic_Tech_Software_D.txt | J. Saldana, The Coding Manual for Qualitative Re- searchers, 3rd edition. Los Angeles ; London: SAGE Publications Ltd, Dec. 8, 2015, 368 pp., ISBN: 978-1- 4739-0249-7. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | We present the wide-field imaging and polarimetry at ν = 20 GHz of seven of the most extended, bright (Stotal > 0.50 Jy), high-frequency selected radio sources in the southern sky with declinations δ < −30◦. Accompanying the data are brief reviews of the literature for each source. The results presented here aid in the statistical com- pleteness of the Australia Telescope 20 GHz Survey (AT20G) Bright Source Sample (Massardi et al. 2008; Burke-Spolaor et al., in prep). The data is of crucial interest for future CMB missions as a collection of information about candidate calibrator sources. We were able to obtain data for seven of the nine sources identified by our selection criteria. We report that Pictor A is thus far the best extragalactic calibrator candidate for the Low Frequency Instrument (LFI) of the Planck ESA mission due to its high level of integrated polarized flux density (∼ 0.50±0.06 Jy) on a scale of 10 arcminutes. Six out of the seven sources have a clearly detected compact radio core in our images, with either a null detection or less than 2 percent detection of polarised emission from the nuclei. Most sources with detected jets have magnetic field alignments running in a longitudinal configuration, however PKS 1333-33 exhibits transverse fields and an orthogonal change in field geometry from nucleus to jets. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | ground (CMB), in particular the current focus on detect- ing anisotropies in the polarised CMB signature. Thomp- son scattering on the Last Scattering Surface produces lin- ear polarisation and curl-free patterns termed “E-modes,” which on small angular scales provide information about structures in the early Universe. Observations of compact CMB foreground sources are necessary to distinguish be- tween real E-modes and signal scattered by gravitational lensing of the CMB on point source angular scales. Addi- tionally, the diffuse plasmas in bright, extended extragalac- tic sources can form patterns which are termed “B-mode” polarisation (Hu, Hedman, and Zaldarriaga 2003); B-mode signal in the radio lobes of extragalactic sources may be un- surprising and can be caused by complex magnetic fields in the sources and irregular substructures of different Fara- day rotation. However, the positive detection of a similar B-mode polarisation signal in the CMB is a major goal of increasingly sensitive measurements of CMB anisotropy. A B-mode signal in the CMB is predicted to occur only by CMB models with tensor (gravitational wave) fluctua- tions; these are in turn tied to the rapid period of expansion that may have occured prior to the recombination epoch. The detection of B-mode CMB anisotropies will give insight into the energy scales that occurred directly prior to infla- tion and would suggest the presence of gravitational waves. However, the amplitude of the cosmological B-mode sig- nals is extremely small, and will require careful attention to foreground signals that might mimic it. On small angu- lar scales where extended and complex foreground sources (such as radio galaxies) are resolved, removing contamina- tion due to extragalactic sources becomes particularly im- portant. For faint, confused, and unresolved extragalactic sources in CMB data, a statistical correction may become necessary to calibrate and interpret the results of a cosmic background B-mode analysis. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | Bright, highly polarised sources are also necessary for upcoming CMB instruments, in particular, the ESA Planck satellite mission’s low frequency instrument (LFI), which will observe in bands between 30 and 70 GHz (Bersanelli and Mandolesi 2000). Among the source types that have been pursued as potential flux density and po- larisation calibrators for the instrument are extragalactic sources; however, as the spectral behaviour of the bright ex- tragalactic population at high frequency is difficult to model, sources selected at lower frequency are not certain to be suitable calibrator candidates. With our observations, along with the catalogues of Massardi et al. (2008) and Burke et al. (in prep), we provide a list and homogeneous observations of bright and strongly polarised sources that are selected at a frequency closer to CMB observing frequencies than was previously available. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | The AT20G team’s confirmation followup observations to their initial transit scan survey provided a list of sources detected at 20 GHz, from which Burke et al. (in prep) se- 30◦ > δ whose lected a sample of sources at declinations total intensity exceeded 0.50 Jy. Preliminary results of an AT20G pilot survey are available in Sadler et al. (2006); a description of the blind scan and followup observations of the AT20G survey are available in Massardi et al. (2008). The 2.4 arcminute primary beam FWHM and the 80 to 200 metre interferometer spacings used in the AT20G fol- lowup did not provide accurate flux density measurements for the most extended sources, and furthermore left cases in which sources were either: 1) Fully resolved (and there- fore undetected) by the 60 metre shortest antenna spac- ing of the AT20G followup, or 2) Had compact compo- nents (hotspots, cores) which were detected as separate, <0.50 Jy sources in the followup. The Sydney Univer- sity Molonglo Sky Survey (SUMSS) (Mauch et al. 2003) and the Parkes-MIT-NRAO (PMN) Southern (Wright et al. 1994; Condon, Griffith, and Wright 1993) and Zenith Sur- veys (Wright et al. 1996) have a higher sensitivity to ex- tended radio sources. This is due to their more complete Fourier coverage for short baselines, and due to the lower observing frequency of the two surveys which allowed higher sensitivity to diffuse, steeper spectrum radio components (SUMSS at 0.843, PMN at 5 GHz). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | To recover highly resolved 20 GHz sources, we cross- referenced multiple-component AT20G sources against SUMSS images. This indicated AT20G sources which ap- peared to be compact components of one larger, extended object. We then ran a search for sources which are highly extended and thus may have been completely resolved out by the AT20G obervations. We expect any extended source with S20 > 0.5 Jy to have a 5 GHz flux of S5 > 0.9 Jy if it has a spectral index of less than -0.5, as would be expected by a conservative estimate for the spectral index of diffuse emission. We thus compiled a list of the sources in the PMN catalogue which are flagged as extended and have an inte- grated 5 GHz intensity greater than 900 mJy. By including these objects and those selected at 20 GHz, this sample has a nominal flux limit for extended sources with S20 > 0.50 Jy. However, due to the spatial resolution limit of PMN and our decision to include only sources flagged as extended in PMN, the sample is not complete for sources of scales between 2.4 and 4.2 arcminutes. The number of missing sources in this range is estimated to be between 4 and 10 based on source count extrapolations. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | Table 1. The nine sources identified by our selection criteria. Columns 1 and 2 give the common source name and the WMAP ID, respectively. Columns 3, 4 give the radio centroid position. Column 5 gives the source redshift; for redshift references, see text. Columns 6, 7 Give the angular size and calculated linear size based on a flat, ΩΛ = 0.7, Ωm = 0.3, H0 = 72 km/s/Mpc cosmology. Column 8 gives relevant details of the observations (including number of telescope pointings for observed sources). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | The resulting list from the searching process consisted of nine objects. However, we did not observe Fornax A due to its highly diffuse emission, most of which would remain undetected even in ATCA’s most compact configuration (furthermore, attempts to detect the nucleus of the galaxy at 20 GHz during unscheduled ATCA telescope time were unsuccessful down to a few milliJansky). For some sources, only subregions containing compact structure were observed. The full nine sources detected by our criteria are listed in table 1, which provides information on what part of the source was observed (if applicable). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | Below, the radio morphology of each source is described, and significant literature and interesting properties for each object are summarised. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | PKS 0131-36 (NGC612), figure 2: An asym- metric, Faranoff-Reilly II source with a bright knot in one jet at low radio frequency (Mauch et al. 2003). As- sociated at optical wavelengths with a magnitude 15 S0-type 1964). Westerlund and Smith (1966) note a radio-optical mor- phology similar to Centaurus A, with a dust lane crossing the core and an asymmetric positioning of the core along the axis of the two radio lobes. Recently, a study of HI gas in the galaxy demonstrated the existence of a huge, 140 kpc disk along NGC612’s dust lane and a faint bridge of HI emission to the nearby galaxy NGC619, indicating a possibly recent or ongoing interaction in this system (Emonts et al. 2008). The source resides near the centre of a cluster of galaxies (Schilizzi 1975). Redshift measured by Ekers et al. (1978). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | symmetrical double-lobed galaxy extending approximately one degree along its major axis. Source is associated with the peculiar S0 galaxy NGC1316 (Mills 1954; Ferguson 1989). The source resides in the Fornax cluster. The high resolution observations of Geldzahler and Fomalont (1984) across 1.5, 4.9, and 15.0 GHz indicate a steep-spectrum core and nonlinear jets. The large-scale structure of the radio source was studied by Ekers et al. (1983), who noted a bridge of emission extending between the lobes that is displaced from NGC1316’s centre; this morphological feature and the complicated dynamics of the galaxy lead to their description of a formation scenario involving multiple galaxy collisions. Caon, Capaccioli, and D’Onofrio (1994) further suggest Fornax A’s host as a merger remnant, noting tidal tails, filaments, dust lanes, and a low velocity disper- sion. Although too diffuse to make useful measurements with our telescope configuration, the source was detected in WMAP at 23 GHz with a flux of S23 = 11.7 0.1 (Bennett et al. 2003). The redshift given in the table is as measured by Longhetti et al. (1998). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | Centaurus A (NGC5128), figure 4: A very complex and highly extended radio source associated with a spherical galaxy crossed by large, absorptive dust lanes. Being the nearest and one of the first radio galaxies identified, the literature is incredibly extensive and too vast to summarise here. A seminal study of the large scale radio structure of the source from 500 MHz to 5 GHz was made by Cooper, Price, and Cole (1965); the optical identification was made by Bolton, Stanley, and Slee (1949). The redshift measurement is available from Graham (1978). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | PKS 1333-33 (IC4296), figure 5: A “triple” radio source called thus for its bright inner jets and diffuse outer lobes that are in alignment along the jet axes. The major lobes extend approximately 30 arcminutes and the inner lobes are within 2 arcminutes of the centroid (Goss et al. 1977; Jones and McAdam 1992). The radio source is associated with the E1 galaxy IC 4296, which is a companion to IC 4299 and is centrally located in the galaxy cluster Abell 3565 (Younis, Meaburn, and Stewart 1985; Huchra and Geller 1982). The host galaxy is claimed to have a dusty disk by Colbert, Mulchaey, and Zabludoff (2001) and Schmitt et al. (2002); however, Sadler and Gerhard saw no significant dust (1985) and Michard (1998) features. and Killeen, Bicknell, and Carter thorough multi-wavelength studies of the radio polarisation and infrared, rotation measures, and studies of the optical, and X-ray form of the source with an extensive analysis presented in Killeen and Bicknell (1988). The redshift was most recently measured by Smith et al. (2000). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | a spectral index of -0.73. Extrapolating their published measurements, we predict an integrated 20 GHz flux value 14.6 Jy. The redshift quoted in the table was of S20 measured by West and Tarenghi (1989). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | PKS 1610-60 , figure 7: The radio source is a wide angle head-tail galaxy with bright inner lobes and diffuse tails (mapped by Schilizzi and McAdam (1975); Jones and McAdam (1992); Christiansen et al. (1977); Gregory et al. (1994) at 408, 843, 1415, and 4850 MHz, respectively). Ekers (1970) identified the source with an E3 60:55:13.7. The galaxy galaxy at J2000 α =16:15:1.7, δ = is thought (but not yet proved) to be a central member of the local massive cluster Abell 3627 (Robertson and Roach 1990; Brown and Burns 1991; Boehringer et al. 1996). Another head-tail galaxy lies within one degree of PKS 1610-60; Jones and McAdam (1996) mapped the two sources at 1360 MHz, explaining PKS 1610-60’s morphol- ogy (sharp kinks in the jets and the wide tail separation) through probable interaction with an intercluster medium and influence from subcluster mergers. Most recently, a redshift was determined by Woudt et al. (2004). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | PKS 2153-69 , figure 8: Less than 3 arcminutes in extent, the radio structure of this Seyfert II galaxy consists of a core-lobe morphology. A presumably unrelated source lies approximately 2.5 arcminutes to the east (Ekers 1969; Christiansen et al. 1977). The host galaxy is an early type of class E0 (Schilizzi 1975; Marenbach and Appenzeller 1982), and has been shown to have an extra-nuclear ionised gas cloud with which the jets are interacting (Tadhunter et al. 1988; Tingay et al. 1996). Further deep radio and X-ray studies have been performed by Fosbury et al. (1998) and Young et al. (2005). The galaxy was modelled to have a 106y) jet precession by Lu and Zhou long-period (t (2005). Redshift given by Da Costa et al. (1991). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | PKS 2356-61 , figure 9: A Faranoff-Riley II galaxy marked by four bright regions of emission that are slightly asymmetric about a core. In addition, there is a diffuse arm of emission extending south-west from the western inner lobe (see e.g. low radio frequency im- ages1 by Christiansen et al. (1977); Jones and McAdam (1996); Subrahmanyan, Saripalli, and Hunstead (1992); Burgess and Hunstead (2006)). Identified with an E3 galaxy by Whiteoak (1972) and Schilizzi (1975). Sus- pected to be a member of galaxy cluster SC 2357-61 (Westerlund and Smith 1966; Teague, Carter, and Gray 1990). Source was detected as a hard X-ray emitter by Landi, Malizia, and Bassani (2005). Redshift measured by Loveday et al. (1996). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | east-west arms with a maximum baseline of approximately 80 metres. The array was well-suited to obtain sufficient (u,v ) coverage to sample and image large, diffuse structures at high radio frequencies over a short period of time. Each source was observed simultaneously at 16.7 and 19.4 GHz in two polarisations using linear feeds with 128 MHz band- width at each frequency. Observations took place on October 1, 2006. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | algorithm performs a maximum entropy method deconvolu- tion simultaneously for all polarisations, using morphologi- cal information from bright emission in the total intensity image to solve for the Q and U polarised Stokes images. Images are then restored using a gaussian with the same half-power beam width of the synthesised beam2. While this method does produce a morphologically correct total inten- sity image, the MEM algorithm is not optimum for comput- ing total intensities, causing integrated intensity measure- ments to be systematically high by about 1 to 20 percent. For this reason, we used CASA’s multi-scale clean (again joining mosaic subfields) to deconvolve images and gather intensity information. While CASA does compute polarisa- tion values, a known error in the data interface for ATCA polarisation information currently renders their output er- roneous. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | Position angle maps are also calculated on a per pixel basis, with ψ = 1 Q , where ψ is the observed angle of polarisation. Vector maps of fractional polarised inten- sity were formed by dividing the masked total polarisation maps by the total intensity maps; in this way, the maps will only show fractional polarisation above the detection threshold. Polarisation vectors were then calculated using lengths determined by fractional polarisation levels, and po- sition angles from PA maps. Note that the orientation of a polarisation “vector” is ambiguous by 180 degrees, and does not provide a preferred electric or magnetic field direction along the printed rods. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | Table 2. See section 4.1 for more detailed column descriptions: (1) Common source name (2-3) J2000 RA/Dec (4,5) Core and total integrated flux density and error in Jy (6) Integrated polarised intensity and error in Jy. ∗ indicates the integrated polarisation over an observed subregion (7) Fractional polarisation (100 · P18/S18) (8) 4.85-18GHz spectral index (9) Flux density as detected in WMAP maps, with †marking fluxes taken from the NEWPS catalogue (Lopez et al. 2007). Note that the Centaurus A measurements can be taken as actual values (not simply limits) when referring to the inner lobes of Centaurus A. |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | an integration over the MSclean-deconvolved maps, nor- malised by the synthesised beam. Note that this process has corrected for the primary beam. Errors were quanti- fied by the image RMS; (6) Integrated detected (> 3σp) polarised intensity in Jy (P = p(P Qi)2 + (P Ui)2, where the sums represent the integrated and beam-normalized val- ues over the P > 3σp and Stokes I > 3σ pixels in the deconvolved Stokes Q and U images). Because of possible depolarisation effects when imaging the missed source re- gions, we cannot provide accurate total integrated polari- sation measurements for sources which were incompletely observed. Such measurements, indicated with an asterisk, give the integrated value of the observed source subregion; (7) The fractional polarisation of the observed source re- gion in percentage (8) The spectral index between 4.850 and 18.0 GHz, using PMN flux densities for all sources except Pictor A, where the 5 GHz measurement of Kuehr et al. (1981) was used and Centaurus B, for which the 5 GHz measurement of Wright and Otrupcek (1990) was used ; (9) WMAP flux density at 23 GHz (Bennett et al. 2003); flux densities marked with are as given in the NEWPS cata- logue (Lopez et al. 2007). |
Processed_Wide-Field_Imaging_and_Polarimetry_for_the_Biggest.txt | frequency images; all cores have either undetected or very low values of polarisation (an upper limit on the polarisation of the cores can be calculated by taking 3 σP from Table 2). The undetected polarisation of the cores could be the result of several different effects. The low values of polarisation can be explained by a complex scattering medium surrounding the inner regions of jet formation, or by a dense plasma. Furthering this supposition are the high (& 1000 rad m−2) rotation measures noted in the inner regions of some AGN. However, studies of galactic nuclei have been yet unable to characterise the spectral properties for emission from ac- tive galactic nuclei; the complex spectra are unable to be fit by pure synchrotron self absorption or free-free spectra. An equally likely case is the beam depolarisation of close, compact regions of structure that we have not resolved due to our lower angular resolution. With milliarcsecond reso- lution, the pilot VLBA Imaging Polarimetry Survey (VIPS, Taylor et al. (2005)) has demonstrated that a sample of 24 3 percent polarisation active galactic nuclei average below at 15 GHz, however the full VIPS sample will provide a more numerically significant assessment for high resolution polarisation. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | The VIALACTEA project aims at building a predictive model of star Abstract. formation in our galaxy. We present the innovative integrated framework and the main technologies and methodologies to reach this ambitious goal. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | The Milky Way galaxy is a complex ecosystem where a cyclical transformation pro- cess brings diffuse baryonic matter into dense unstable condensations to form stars, that produce radiant energy for billions of years before releasing chemically enriched material back into the InterStellar Medium in their final stages of evolution. Although considerable progress has been made in the last two decades in the understanding of the evolution of isolated dense molecular clumps toward the onset of gravitational collapse and the formation of stars and planetary systems, a lot remains still hidden. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | The aim of the European FP7 VIALACTEA project is to exploit the combination of all new-generation surveys of the Galactic Plane to build and deliver a galaxy scale predictive model for star formation of the Milky Way. The technological objectives of the project are: to boost the scientific exploitation of ESA missions space data by devel- oping new and carefully tailored data processing tools to combine in a VO-compatible and interoperable way the new-generation Galactic Plane surveys; to build and visualize an innovative 3D representation of the Milky Way Galaxy. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | To explore very large regions of the galaxy, a new framework is implemented using advanced visual analytics techniques, data mining methodologies, machine learning paradigms and based on VO (Virtual Observatory) data representation and retrieval standards. All such specialized tools are integrated into a virtualized computing envi- ronment, resulting as an efficient and easy-to-use gateway for the scientific stakeholder community. An overview of the methodologies and technologies (see Figure 1), able to fulfil the scientific expectations of the project is presented in the following sections. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | The ViaLactea Knowledge Base (VLKB) includes a combination of storage facilities, a Relational Data Base (RDB) server and web services on top of them. The RDB is the content holder of the metadata of the stored files as well as a data resource itself. It is also the base resource on top of which the IVOA TAP service (Dowler et al. 2010) works. The goal of the VLKB RDB component is to allow easier searches and cross correlations between VIALACTEA data using the software tools the user community have at its disposal. Consuming the VLKB alongside other VO available resource is possible through the implementation of the TAP service so that the project’s commu- nity can exploit the VIALACTEA data without the need to continuously retrieving and downloading external resources. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | Currently the RDB is subdivided, for logical and functional reasons, into separate schemas (see Figure 2) containing: Hi-GAL survey (Molinari et al. 2010) catalogue sources and related band merged information; structural informations such as filament structures or bubbles; and Radio Datacubes with search and cutout services. Finally a TAP_SCHEMA is built for the VLKB TAP services, this allows to make all tables accessible using standard VO tools capable of TAP/ADQL queries. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | The integration and exploitation of data-mining and machine-learning technologies within the project enable developing new tools that incorporate data products and the astronomer’s know-how into a set of supervised workflows with decision-making capa- bilities to carry out building of Spectral Energy Distributions, distance estimate and Evolutionary classification of hundreds-of-thousands of star forming objects on the Galactic Plane. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | More in detail, the main developed data mining tools are related to: compact source extraction to obtain a more refined version of band-merged catalogues based on the positional cross-match among sources at different wavelengths (Q-FULLTREE); filamentary structure detection to refine and optimize the detection of the edges of fil- amentary structures (Riccio et al. 2015) (FilExSeC); and source kinematical distance estimation combining all available information from Galactic rotation curve, spectro- scopic survey data in molecular gas lines or 3D extinction maps in the near and mid- Infrared (MLNPQNA). |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | Real-time data interaction are performed to carry out complex tasks for multi-criteria data/metadata queries on the VLKB, subsample selection and further analysis processed over the Science Gateway, or real-time control of data fitting to theoretical models. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | Figure 3. Screenshots of the 3D Visual Analytics tool. Left: Navigable overview of the galactic plane from -60 to +60 degree. Right: Compact sources visualization. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | visualization allows to analyze single slices and extract iso-contours to investigate the region of interest. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | An infrastructure monitoring has been developed to guarantee full operation of the VIALACTEA infrastructures where a diversity of software and parallel and multi- thread jobs are running. The developed monitoring features of the Science Gateway test the infrastructures health periodically, report monitoring tests on the gateway as well as the history of the monitoring tests and send e-mail alerts on failure. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | The presented tools developed within the VIALACTEA project will be able to access data products as well as libraries of millions of radiative transfer models necessary for the science analysis in an integrated environment. Especially the emerging field of visual analytics brings data analysis and visualization into a single human-in-the- loop process. New technologies such as new 3D images of our galaxy with interactive data manipulation capabilities will provide powerful analytical, educational and inspi- rational tools for the next generation of researchers. |
Processed_Advanced_Environment_for_Knowledge_Discovery_in_th.txt | Acknowledgments. The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 607380 (VIALACTEA: the Milky Way as a star formation engine). |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Abstract This research investigates the curriculum structures of undergraduate Artificial Intelligence (AI) education across universities worldwide. By examining the curricula of leading universities, the research seeks to contribute to a deeper understanding of AI education on a global scale, facilitating the alignment of educational practices with the evolving needs of the AI landscape. This research delves into the diverse course structures of leading universities, exploring contemporary trends and priorities to reveal the nuanced approaches in AI education. It also investigates the core AI topics and learning contents frequently taught, comparing them with the CS2023 curriculum guidance to identify convergence and divergence. Additionally, it examines how universities across different countries approach AI education, analysing educational objectives, priorities, potential careers, and methodologies to understand the global landscape and implications of AI pedagogy. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | In recent years, artificial intelligence (AI) has pervaded our daily lives, driven by high-impact machine learning roles in finance, It plays integral (ML) applications. healthcare, transportation, and law enforcement, alongside various applications in computer science. Additionally, AI techniques are increasingly utilized in fields such as biology, chemistry, physics, art, and architecture, addressing complex challenges across diverse domains Garvey and Patrick (2023); Doughty et al. (2024); Williamson (2024); Nithithanatchinnapat et al. (2024); Xu et al. (2024). Concepts like “neural networks” and “deep learning” have gained widespread recognition, enabling individuals to perform tasks previously reserved for experts. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | collaborated with the Association for the Advancement of Artificial Intelligence (AAAI) to integrate AI more compre- hensively into the latest revision of their curricular guide- lines, known as CS2023 Eaton and Epstein (2024). Within the AI Knowledge Area (KA), twelve knowledge units were identified, with five units partitioned among the Computer Science (CS) Core, the KA Core, and electives (Fundamental Issues, Search, Fundamental Knowledge Representation and Reasoning (KRR), Machine Learning, Applications, and Societal Impact). Given these significant updates to AI edu- cation standards, understanding and comparing the curricu- lum structures of AI education across universities worldwide becomes imperative. My research aims to analyze these curricula to identify best practices, gaps, and opportunities for improvement, ensuring that AI education remains aligned with industry demands and technological advancements. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | In an era marked by rapid advancements in artificial intelligence (AI), the pivotal role of education in preparing the next generation of AI professionals has become increasingly evident Zhou et al. (2021); Brusilovsky (2024); Zhou et al. (2022). Given the prevalence of AI, it has focus in the education of naturally become a central computer scientists, and as AI continues to permeate various aspects of society, the vision and strategies adopted by universities in shaping AI education programs hold profound implications for the future workforce and innovation landscape Wang and Yu (2023); Druga et al. (2022); Gibellini et al. (2023). However, the fast pace of AI developments poses a significant challenge in crafting a curriculum that keeps up with the latest advancements over an extended period. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | universities in the field of artificial intelligence education? We delve into the contemporary prospects of AI through a comprehensive analysis of the diverse course structures adopted by leading universities. Through this examination, we aim to understand the prevailing trends, methodologies, and priorities in AI education, revealing the nuanced approaches taken by different institutions. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | By addressing these research questions, our research strives to contribute to a deeper understanding of AI education, its evolving contours, and its global impact. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | This section includes both data collection and analytical metrics used in the study, providing a comprehensive approach to understanding the educational landscape of AI programs across selected universities. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Utilizing insights from both the “2024 AI Index Report” by Stanford University Stanford University (2024) and the “QS World University Rankings by Subject 2024: Data Science and Artificial Intelligence Quacquarelli Symonds (QS) (2024),” a rigorous selection process was undertaken to identify leading universities at AI education. Institutions from nine countries spanning four continents are included for detailed examination. These regions encompass the United States and Canada in North America, the United Kingdom and Italy in Europe, China (encompassing both mainland China and the Hong Kong Special Administrative Region), in addition to Singapore, Malaysia, and India in Asia, Australia in Oceania. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | The data collection methodology for this study involved a thorough investigation of university websites, with a specific focus on institutions listed in the QS World University Rankings Quacquarelli Symonds (QS) (2024). Among the 72 universities distinguished for their prowess in these domains, particular emphasis was placed on identifying universities offering undergraduate programs in Artificial Intelligence (AI). |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | The selection criteria prioritized universities that explicitly offered bachelor’s degree programs specializing in AI or closely related fields. Additionally, universities with prominent AI interdisciplinary research departments or programs incorporating AI education for undergraduates were included in the data collection process. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Each university’s official website served as the primary source of information for gathering data on the curriculum structure and course offerings. For universities offering undergraduate AI programs, the units comprising the curriculum were meticulously documented. In cases where AI was not explicitly offered as a standalone bachelor’s degree program, relevant courses closely aligned with AI principles, such as machine learning, data mining, or computer science courses with significant AI content, were identified and included in the dataset. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | The data collection process involved thorough navigation of university websites, exploration of program descriptions, examination of course catalogs, and consultation of academic departments or faculty profiles where necessary. Special attention was paid to ensuring the accuracy and completeness of the collected data, focusing on capturing the diverse settings and approaches to AI education across different institutions. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | In the following section, we employ various analytical techniques to gain insights into our dataset of course names. We utilize methods such as word clouds to visualize popular words in course names, bar charts to depict the distribution of courses across different categories and the popularity of various AI courses, scatter plots to illustrate clusters within the course dataset, and comparative analysis across continents. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | where pij and qij are the probabilities of data points i and j being neighbors in the high-dimensional and low- dimensional spaces, respectively. The clusters are visualized in a scatter plot using the 2D t-SNE coordinates. Each point represents a course, color-coded by its cluster label. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | To complement the detailed methodology for Scatter Plot Clustering, a similar approach is employed to visualize the distribution of AI courses across continents using two different tokenizers. For the TF-IDF tokenizer, course names are converted into TF-IDF vectors, capturing the importance of each term relative to the entire corpus. These vectors undergo dimensionality reduction via Truncated SVD to retain essential information while minimizing the number of features. KMeans clustering is applied to identify course clusters, and TSNE is used for further dimensionality reduction to visualize the data in a 2D space. In contrast, the BERT tokenizer employs a deep contextualized word embedding model to tokenize course names into dense vectors that represent the semantic meaning of each term within the context of the sentence. These vectors are projected onto a lower-dimensional space using TSNE for visualization. KMeans clustering is then applied to identify clusters of courses, and the scatter plot is generated with courses color-coded by continent. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | This section presents the analytical results derived from the collected data, offering insights into the educational landscape of AI programs across the selected universities. The analysis includes a detailed examination of curriculum structures, identification of core AI topics, and a comparative analysis of AI education across different regions. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Most AI-related undergraduate programs are structured as bachelor’s degree programs, designed to equip students with foundational knowledge and practical skills in AI. Additionally, some programs offer specialized tracks or concentrations within the broader field of AI, serving as major pathways for students pursuing careers in AI research or related areas such as Computer Science (CS). |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | institutions, core units in AI education programs are depicted as compulsory requirements, core modules, or essential units for understanding foundational knowledge in this field. These core units are represented in the stacked bar chart as the purple portion of the bars, indicating their universal inclusion in the curriculum. In addition, most programs offer optional units in AI, referred to as “options” or “electives,” providing students with flexibility and choice in their course selection. For example, the University of Hong Kong offers different concentrations covering multiple units to explore interconnected subjects such as AI in Business Technology and Finance, and AI in Smart City. This variety of optional units allows students to tailor their education to their interests and career aspirations, fostering specialization in specific areas of AI. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | While some universities offer comprehensive AI programs with a wide range of core and optional units, others may only refer to a few core units. This variation in program structure may be attributed to differences in institutional focus, program objectives, or the integration of AI into broader academic disciplines. In the case of universities like the University of California, Berkeley (UCB) and the University of Washington, their AI programs primarily focus on core units related to AI reflecting a specialized approach to the subject matter. Similarly, institutions such as Stanford University and Harvard University offer specialized programs like the Artificial Intelligence Professional Certificate Program, which may emphasize core units tailored to specific professional objectives rather than offering a broad range of elective options. Furthermore, universities like the University of British Columbia (UBC) may structure their AI education as certificate or micro-certificate programs, limiting the curriculum to core units necessary for certification. This streamlined approach aligns with the program’s objective of providing targeted training in AI skills within a shorter timeframe. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Overall, the analysis underscores the universal presence of core units in AI education programs, alongside the provision of optional units for specialization. The varying approaches to AI education reflect the diverse needs and objectives of students and institutions, resulting in different program structures tailored to meet specific academic, professional, and certification requirements. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Focus on Fundamental AI Concepts: The largest cluster located on the top right of the figure suggests a strong emphasis on fundamental AI concepts across universities. Keywords such as “Introduction to AI,” “Fundamentals of AI,” and “AI Theory” dominate this cluster, indicating a focus on foundational issues in AI education. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Programming-Related Courses: The scatter plot reveals a significant cluster situated in the bottom left quadrant, indicating a concentration of programming- related courses. Notable examples within this cluster encompass “Python Programming,” “Game Program- ming,” and “Structured Programming,” underscoring the pivotal role of programming as a requisite skill for AI students. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Machine Learning and Deep Learning Foundations: In the upper top section of the plot, two clusters coalesce to represent machine learning and deep learning courses, forming a cohesive entity. Prevalent keywords such as “Machine Learning,” “Deep Learning,” and “Applied Deep Learning” underscore the widespread inclusion of these subjects across university curricula, emphasizing their foundational significance in AI education. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Mathematical Requirements: The clusters in the lower right of the graph (labeled 2 and 6), are associated with mathematical concepts such as “Mathematics,” “Probability,” and “Statistics.” This suggests a strong emphasis on the mathematical prerequisites necessary for studying AI. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Specialized AI Subfields: Additional clusters repre- senting specialized AI subfields are scattered through- out the plot. Clusters for “Natural Language Pro- cessing (NLP),” “Ethics/AI and Society,” “Computer Vision,” and ”Robotics” indicate a diverse range of course offerings covering various applications and societal implications of AI, underscoring an increased emphasis on the practical applications of AI and their societal and ethical significance. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | the most commonly offered course, with a total of 16 out of 29 institutions providing it, including both the first and sixth most popular courses. Following closely are natural language processing and computer vision, each offered by 8 universities. This trend underscores the significance of these domains in AI education. Additionally, programming- related courses such as “Introduction to Programming” and fundamental issues like “Fundamentals of AI” are prevalent, aligning well with the guidance outlined in CS2023. These findings reflect the emphasis on foundational concepts and practical skills in AI curricula across institutions. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | To gain insights into the core components of undergraduate AI education across different universities and compare them with the AI knowledge areas outlined in the CS2023 curriculum guidance, we have collected the names of core units from various undergraduate programs. These core units represent the foundational pillars upon which AI education is built, encompassing essential topics and concepts integral to understanding and applying AI principles. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Prominence of Foundational Concepts: The word cloud prominently features terms like “machine learn- ing”, “Programming”, “Algorithm”, “Introduction” and “fundamentals,” indicating their central role in AI education. These terms likely represent core courses that lay the groundwork for understanding AI princi- ples, algorithms, and methodologies as recommended in CS2023. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | cover topics such as AI in healthcare, finance, robotics, or other specific domains, aligning with CS2023’s emphasis on the application of AI techniques in diverse contexts. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Presence of Elective Units: Additionally, units such as “computer vision” and “natural language processing” stand out, although to a lesser extent compared to the basic concept. Their appearance suggests that while not required as core units in the CS2023 Artificial Intelligence Knowledge Area, these topics are recognized as important units within AI curricula. Their inclusion reflects a focus on specialized areas of AI, allowing students to deepen their knowledge and expertise based on their interests and career aspirations. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Overall, the word cloud provides a snapshot of AI education that aligns with the CS2023 guidance, encompassing a bal- ance of foundational knowledge and practical applications. The distribution of terms suggests a holistic approach to AI education, aiming to equip students with a broad understand- ing of core principles while also offering opportunities for in-depth exploration of specialized topics. This alignment with CS2023 ensures that graduates possess the requisite knowledge and skills to contribute effectively to the evolving field of artificial intelligence in various professional settings. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | In this section, we aim to conduct a comparative analysis across different continents into the similarities and differences in the distribution of AI courses. Firstly, we employ the aforementioned stacked bar chart to stack various categories of courses from each university, facilitating the visualization and comparison of curriculum patterns across different continents. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | From Figure 5, we can find that Asian universities stand out with the highest number of courses and variety in their curriculum structures. Specifically, six out of eight Asian universities provide more than 28 courses, reflecting a rich and diverse curriculum. Contrasting the Asian universities’ rich curriculum offerings, universities from other continents, particularly North America, present a contrasting pattern with a lower average number of courses. For example, Oceania universities offer an average of around 21 courses, while those in the North America provide approximately 16 courses on average. This disparity can be attributed to the fact that the majority (6 out of 8) of programs offered by North American universities focus on AI research-related courses or certificate programs. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | highlights the significant differences in course offerings among universities across continents, with Asian universities leading in terms of course variety and richness in their curriculum structures. This observation aligns well with the result of “2024 AI Index Report” Stanford University (2024). The report shows that the UK has more AI study programs despite having fewer universities overall compared to the US based on the Studyportals (2024). (In 2023, the UK had 744 English-language AI study programs, while the US had 667.) The factors like Studyportals’ coverage and the structure of higher education could definitely contribute to this. In the UK, there might be a more concentrated effort to offer specialized AI programs, whereas in the US, students interested in AI might pursue it within broader computer science programs. This nuanced understanding helps paint a clearer picture of AI education across different countries. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | and Europe. While these continents exhibit some dense clusters, indicating a diverse range of course offerings, there is also a broader dispersion of courses, suggesting a broader coverage of AI topics. In contrast, universities in North America and Oceania demonstrate a more focused distribution, with fewer dense clusters. This concentration of courses suggests a focus on specific areas within AI education rather than a broader coverage of topics. Similarly, the scatter plot generated using the BERT tokenizer also showcases a dispersed pattern of course distributions across Asia and Europe. However, the dispersion tends to cover a broader area, indicating a more diverse range of offerings. Universities in North America and Oceania exhibit a focused distribution, with dense clusters in specific areas. This focus suggests a prioritization of particular topics or domains within AI education. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Comparing the scatter plots generated using TF-IDF and BERT tokenizers, we observe consistent patterns across continents. Both tokenizers reveal dispersed distributions in Asia and Europe, indicating a diverse range of course offerings, while North America and Oceania exhibit more focused distributions. This suggests that universities in these regions prioritize specific areas within AI education, with varying degrees of breadth and depth in their course offerings. The analysis of scatter plots generated using different tokenizers provides valuable insights into the distribution of AI courses across continents. By comparing patterns observed in TF-IDF and BERT tokenizers, we gain a nuanced understanding of regional emphases and priorities in AI undergraduate education, highlighting areas of diversity and specialization across geographical regions. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Based on our observations, most AI undergraduate programs around the world require students to have advanced- level qualifications in mathematics as part of their entry requirements. For example, the University of Birmingham in the UK requires students to include A-level Mathematics with a grade A, while the University of Technology Sydney in Australia requires Higher School Certificate (HSC) Mathematics Advanced. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | In addition to the mathematical entry requirements, students are typically required to take compulsory units related to mathematics within AI undergraduate programs. it was Among 29 institutions analyzed, found that: indicating a strong 21 courses include “mathematics”, emphasis on mathematical concepts and principles. 13 courses include “statistics”, highlighting the importance of statistical analysis and inference in AI applications. indicating 12 courses include the word “probability”, the significance of probability theory in understanding uncertainty and making informed decisions in AI. 12 courses include “calculus”, underscoring the necessity of calculus in modeling and optimizing algorithms and systems in AI. Overall, these findings suggest that AI undergraduate programs place a significant emphasis on mathematical proficiency, with students expected to have a strong foundation in mathematical principles upon entry and to further develop their mathematical skills through compulsory math-related courses throughout the program. This emphasis reflects the central role that mathematics plays in the theory, development, and application of artificial intelligence. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | rewarding careers as data scientists, given the strong overlap between AI and data science disciplines. With their expertise in machine learning algorithms, statistical analysis, and programming, AI graduates are well-suited to extract valuable insights from large datasets and drive data-driven decision-making across various industries. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | roles, AI graduates frequently pursue careers as machine learning engineers, leveraging their skills to develop and deploy machine learning models that power intelligent systems and applications. Additionally, computer vision engineers are in high demand, as industries increasingly rely on image recognition technology for tasks such as facial recognition, object detection, and augmented reality applications. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Furthermore, AI graduates often succeed in roles such as software engineers, developers, and IT consultants, where their AI expertise enhances product development, system optimization, and technology implementation. Additionally, opportunities abound in cloud computing engineering and cybersecurity analysis, as organizations seek to leverage AI technologies to enhance security and scalability in cloud- based environments. As the AI field continues to grow, specialized roles such as AI ethicists, AI architects, and AI policy advisors are emerging to address the ethical, legal, and societal implications of AI technologies. Graduates with a strong foundation in AI principles and ethics are well-positioned to shape the future of AI responsibly and sustainably. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Overall, AI graduates have a wealth of career opportu- nities, ranging from traditional roles in data science and machine learning to emerging roles in AI ethics, policy, and governance. By leveraging their skills and expertise, AI graduates can make significant contributions to the advance- ment and responsible deployment of AI technologies across various domains. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Future research should explore the longitudinal evolution of AI curricula to understand how educational programs adapt to emerging technologies and industry demands. Additionally, examining the impact of these educational structures on student outcomes, such as employability and innovation capacity, will provide deeper insights into the effectiveness of various curriculum designs. By continuing to analyze and refine AI educational programs, we can better prepare students for the dynamic and rapidly evolving field of artificial intelligence. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | The author(s) received no financial support for the research, authorship, and/or publication of this article. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Brusilovsky P (2024) AI in education, learner control, and human-ai collaboration. Int. J. Artif. Intell. Educ. 34(1): 122–135. DOI: 10.1007/S40593-023-00356-Z. URL https://doi.org/ 10.1007/s40593-023-00356-z. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Doughty J, Wan Z, Bompelli A, Qayum J, Wang T, Zhang J, Zheng Y, Doyle A, Sridhar P, Agarwal A, Bogart C, Keylor E, K¨ult¨ur C, Savelka J and Sakr M (2024) A comparative study of ai-generated (GPT-4) and human-crafted mcqs in programming education. In: Herbert N and Seton C (eds.) Proceedings of the 26th Australasian Computing Education Conference, ACE 2024, Sydney, NSW, Australia, 29 January 2024- 2 February 2024. ACM, pp. 114–123. DOI:10.1145/ 3636243.3636256. URL https://doi.org/10.1145/ 3636243.3636256. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Druga S, Otero N and Ko AJ (2022) The landscape of teaching the 27th resources for ai education. ACM Conference on on Innovation and Technology in Computer Science Education Vol. 1, ITiCSE ’22. New York, NY, USA: Association for Computing Machinery. ISBN 9781450392013, p. 96–102. DOI:10.1145/3502718.3524782. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Eaton E and Epstein SL (2024) Artificial intelligence in the CS2023 undergraduate computer science curriculum: Rationale and challenges. In: Wooldridge MJ, Dy JG and Natarajan S (eds.) Thirty-Eighth AAAI Conference on Artificial Intelligence, AAAI 2024, Thirty-Sixth Conference on Innovative Applications of Artificial Intelligence, IAAI 2024, Fourteenth Symposium on Educational Advances in Artificial Intelligence, EAAI 2014, February 20-27, 2024, Vancouver, Canada. AAAI Press, pp. URL 23078–23083. https://doi.org/10.1609/aaai.v38i21.30352. Garvey and Patrick G (2023) Artificial intelligence and art: A university curriculum course for undergraduates. In: 2023 14th IIAI International Congress on Advanced Applied Informatics (IIAI-AAI). pp. 667–670. DOI:10.1109/IIAI-AAI59060.2023. 00131. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Gibellini G, Fabretti V and Schiavo G (2023) Ai educa- tion from the educator’s perspective: Best practices for an inclusive ai curriculum for middle school. In: Extended Abstracts of the 2023 CHI Conference on Human Factors in Computing Systems, CHI EA ’23. New York, NY, USA: Association for Computing Machinery. ISBN 9781450394222. DOI:10.1145/3544549.3585747. URL https://doi-org.ezproxy-b.deakin.edu. au/10.1145/3544549.3585747. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Nithithanatchinnapat B, Maurer J, Deng XN and Joshi KD (2024) Future business workforce: Crafting a generative ai-centric curriculum today for tomorrow’s business education. SIGMIS Database 55(1): 6–11. DOI:10.1145/3645057.3645059. URL https://doi-org.ezproxy-b.deakin.edu. au/10.1145/3645057.3645059. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Wang Y and Yu F (2023) Visualization and analysis of mapping knowledge domains for ai education system studies. In: 2023 11th International Conference on Information and Education Technology (ICIET). pp. 475–479. DOI:10.1109/ICIET56899. 2023.10111115. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Xu M, Du H, Niyato D, Kang J, Xiong Z, Mao S, Han Z, Jamalipour A, Kim DI, Shen X, Leung VCM and Poor HV (2024) Unleashing the power of edge-cloud generative ai in mobile IEEE Communications networks: A survey of aigc services. Surveys & Tutorials 26(2): 1127–1170. DOI:10.1109/COMST. 2024.3353265. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Zhou X, Tong Y, Lan X, Zheng K and Zhan Z (2021) Ai education In: in massive open online courses: A content analysis. |
Processed_Comparative_Analysis_Vision_of_Worldwide_AI_Course.txt | Zhou Y, Zhan Z, Liu L, Wan J, Liu S and Zou X (2022) International prospects and trends of artificial intelligence education: A content analysis of top-level AI curriculum across countries. In: Proceedings of the 6th International Conference on Digital Technology in Education, ICDTE 2022, Hangzhou, China, September 16-18, 2022. ACM, pp. 337–343. DOI:10.1145/ 3568739.3568796. URL https://doi.org/10.1145/ 3568739.3568796. |
Processed_Poincaré_recurrence_theorem_and_the_strong_CP-prob.txt | Sergei N. Nedelko ‡ Bogoliubov Laboratory of Theoretical Physics, JINR, 141980 Dubna, Russia (Dated: December 8, 2017) The existence in the physical QCD vacuum of nonzero gluon condensates, such as hg2F 2i, re- quires dominance of gluon fields with finite mean action density. This naturally allows any real number value for the unit “topological charge” q characterising the fields approximating the gluon configurations which should dominate the QCD partition function. If q is an irrational number then the critical values of the θ parameter for which CP is spontaneously broken are dense in R, which provides for a mechanism of resolving the strong CP problem simultaneously with a correct implementation of UA(1) symmetry. We present an explicit realisation of this mechanism within a QCD motivated domain model. Some model independent arguments are given that suggest the relevance of this mechanism also to genuine QCD. |
Processed_Poincaré_recurrence_theorem_and_the_strong_CP-prob.txt | which, and this is important, neither forbids space-time variation of the strength tensor and the fields Ba µ(x) nor devalues the importance of topological (singular) pure gauge field defects of various dimensions distributed throughout the entire Euclidean space-time [4, 5]. |
Processed_Poincaré_recurrence_theorem_and_the_strong_CP-prob.txt | is well-defined on and, as is known from phenomenol- ogy and lattice QCD, is a nonzero constant, related to the topological susceptibility χ of the QCD vacuum [7]. In both cases of fields satisfying Eq. (6) and dilute su- perpositions of instantons, Q can be expressed in terms of a mean action density. There is a crucial difference however between instanton based approximations of the fields with infinite action and backgrounds (as in the do- main model for instance [5, 8, 9, 10]) based on Eq. (6). The literal use of instantons to approximate fields with infinite action assigns integer values of topological charge to the fields from and thus transports integer units ν of topological charge from Eq. (3) into QCD with con- densates where such integers are irrelevant. Instead, ap- proximations based on Eq. (6) leave the freedom to have any real values for Q for genuine gluon configurations and unit ”topological charge” q for the fields approximating them in a particular model of QCD vacuum. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.