172,783 research outputs found
Caridina kutchi Pandya & Richard 2019, sp. nov.
No full textCaridina kutchi sp. nov. (Figs. 4, 5) Material examined. Gujarat, India. Types: Caridina kutchi sp. nov. Holotype. Jagadiya Dam, Khari River, Kutch, coll. Pandya, 7.9.2014, det. Pandya & Richard, 2015, RRLC /BIO-SH/02, ♂; Paratypes. Jagadiya Dam, Khari River, Kutch, coll. Pandya, 7.9.2014, det. Pandya & Richard, 2015, RRLC /BIO-SH/02, ♂; Bhadra, Khari River, Kutch, coll. Pandya, 7.9.2014, det. Pandya & Richard, 2015, RRLC /BIO-SH/01, 2♀; Ker-vandh, Khari River, coll. Pandya, 7.9.2014, det. Pandya & Richard, 2015, RRLC /BIO-SH/03, 4♂, 2juv.; Khari river catchment, Bhojraj vandh, Gadhshisha, Kutch. coll. Pandya, 7.9.2014, det. Pandya & Richard, 2015, RRLC / Bio-Gadh /07, 2♂, 2♀. Other material examined. Sri Lanka ( Ceylon). Types: Caridina simoni Bouvier, 1904, coll. E. Simon, 1904, Lectotype, designated by Richard & Clark 2014, MNHN Na 856, ♂; Paralectotype MNHN Na 856 ♂; coll. E. Simon, 1904, exch. Paris Museum, 117-97, NHM reg. 1907.1.7.33, 1♀. Nontypes: Sri Lanka. Caridina simoni Bouvier, 1904, irrigation streams, Peradeniya, pres. R. Gurney, NHM reg.1920.2.5.11-13, 4♀; stream running in to Mahawallagunga River, Peradeniya, pres. R. Gurney, NHM reg. 1920.2.5.14-16, 1♂, 1♀ ovig., 1♀, 1 damaged specimen; Keani River, Kekirawa, Colombo, pres. D. R. R. Burt, NHM reg. 1935.5.30.26-27, 4♂, 3♀; Kalaweva, April 1932, pres. D. R. R. Burt, Department of Zoology, University College, NHM reg. 1935.5.30.15-19, 1♂ (abnormal), 4♀ ovig., 2♀; from streams running into Mahawallagunga River, pres. Dr. R. Gurney, det. W.T. Calman, NHM reg. 1947.3.18, 1♀ ovig; pres. Dr. R. Gurney, NHM reg. 1950.1.2.148, dissected parts; irrigation streams, Peradeniya, pres. Dr. R. Gurney, NHM reg. 1951.2. 17.1792/3, 1♂, 1♀; fresh water pond, Botanical Gardens, Perademiya, 17.6.1954, coll. & pres. E.S. Brown, NHM reg. 1954.10.27.1-10, 20♂, 5♀ ovig., 7♀; Ambanganga Anoiont, nr. Polonarraw, 1962, coll. & pres. C. H. Fernandes, NHM reg. 1962.8.24.104, 3♀ ovig., 1♀. India. Hindupur, S. India. coll. P. K. Sartory, pres. Mr. Scourfield, det. J. Richard & P. Cark 2009, NHM reg. 1945.vii.27.5-12, 3♂, 4♀; Madras (Chennai) area, coll. and pres. Dr. Sanjeevaraj, det. I. Gordon, 0 5. 1965. NHM reg. 1965.5.7.1-10, 31♀ ovig. Description. Adult size 15–28 mm. Carapace length 2.2–3.5 mm. Rostrum (Fig. 4a, b, c): Slender, 1.4–1.7×long as carapace, distinctly longer than antennal scale; 12–22 teeth proximally leaving 0.5–0.65 of dorsal margin unarmed distally which is interrupted by a single tooth at distal end; tip pointed and setose dorsally. 1–3 post orbital teeth present. 9–15 teeth proximally leaving 0.1–0.2 of ventral margin unarmed distally. Formula (1–3) 12–22+1/9–15. Carapace (Fig. 4a, c): Antennal spine well developed. Pterygostomian angle rounded without a spine. Mouth parts: Mandibles asymmetrical without palp. Incisor process of mandibles ending in irregular teeth, molar process truncated. Maxillula with broadly truncated lower lacinia and elongated upper lacinia bearing distinct teeth on inner margin; palp slender. Upper endites of maxilla subdivided, palp elongated, scaphognathite with long narrow posterior lobe bearing tuft of setae at truncated tip. Palp of first maxilliped rounded ending in a finger like projection. Endopod of second maxilliped with ultimate segment fused to penultimate segment; exopod longer than endopod. Third maxilliped reaching the end of second segment of antennular peduncle. Exopod reaching 2 nd segment of endopod. Epipod present. Antennular peduncle (Fig. 4a, b, c): 0.8–0.9×carapace. Stylocerite 0.6–0.75×length of basal segment. Anterolateral teeth of basal segment 0.19–0.23×second segment. 10–25 segments bearing aesthetascs. First pereiopod (Fig. 5a): Dactylus 1.3–1.4×palm of propodus. Chela 3.2–3.7×long as broad. Carpus 1.7– 2.3×long as broad, with anterior excavation. Second pereiopod (Fig. 5b): Dactylus 1.5–1.9×long as palm of propodus. Chela 2.7–3.7×long as broad. Carpus 4.9–6.4×long as broad. Third pereiopod (Fig. 5c, d): Dactylus 3.0–3.7×long as broad. 7–12 marginal spines on dactylus. Propodus 4.1–5.0×long as dactylus and 10.0–12.5×long as broad with 10–14 spines along inner margin. Carpus 0.45– 0.55×long as propodus, with 1 large spine and 3–5minute spines on inner margin. Merus 1.6–2.0×carpus length. Merus with 3 large spines on posterior margin. Ischium with a spine. Fifth pereiopod (Fig. 5e, f): Dactylus3.9–5.0×long as broad with 40–50 marginal spines. Propodus 12–16×long as broad and 3.7–4.2×long as dactylus and with 10–15 spines along posterior margin. Carpus 0.4 5–0.6×propodus length and with 4–5 minute spines along inner margin. Merus 1.5–1.9×carpus length, with 2 large spines at posterior margin. Ischium with a spine. Epipod: present on 1–4 pereiopods; absent on fifth pereiopod. Setobranchs: 1 seta on all pereiopods. First male pleopod (Fig. 5g, h): Endopod 0.25–0.35×exopod, appendix interna absent. First female pleopod: Endopod 0.5–65×exopod. Second male pleopod (Fig. 5i, j): Appendix masculina 1.4–1.7×appendix interna and 0.25–0.3×endopod. 6th abdominal somite (Fig. 4a): 0.57–0.86×long as carapace. Telson (Fig. 4a, 5k, l): Narrow and tapering, 1.0–1.1×long as 6th abdominal somite. Dorsal spines 4–6 pairs (including subterminal spine). Posterior margin narrow and triangular, with a median projection, bearing 1 pair of long lateral spines and 2–3 pairs of sparsely plumose spines of equal length and shorter than laterals. Uropod (Fig. 5m): 8–12 diaeresis spinules. Preanal carina (Fig. 5n): armed with a spine. Colouration. Freshly collected specimens were light greenish transparent in colour. Type locality. Jagadiya Dam, River Khari, Kutch District (also spelt as Kachchh) Gujarat, India. Etymology. The species is named for Kutch District, Gujarat, from where the specimens were collected. Remarks. Caridina kutchi sp. nov. is distinguished by long, slender rostrum that is distinctly longer than antennal scale, the unarmed dorsal margin interrupted by a single tooth distally; pointed tip of rostrum with fine setae on the dorsal margin; telson posterior margin narrow and triangular with a median projection bearing intermediate spines of equal length that are distinctly shorter than the laterals. Caridina kutchi sp. nov. is similar to Caridina simoni Bouvier, 1904 which was described from Sri Lanka and now reported from South India (Richard and Clark 2014) in the structure of rostrum with pointed tip and the distal unarmed rostral margin interrupted by a single tooth distally. However, C. kutchi sp. nov. distinctly differs from C. simoni in telson structure. C. kutchi sp. nov. could be distinguished from C. simoni in having rostrum that is distinctly longer than antennal scale (vs. equal to or slightly longer than antennal scale in C. simoni); unarmed dorsal rostral margin interrupted by a single tooth distally (vs. unarmed dorsal rostral margin interrupted by 0–4 teeth in C. simoni); posterior margin of telson narrow and triangular with a median projection (vs. posterior margin of telson broad and rounded without a median projection in C. simoni); telson posterior margin bearing 2–3pairs of sparsely plumose intermediate spines of equal length and distinctly shorter than laterals spine (vs. 3–4 pairs of sparsely plumose intermediate spines either equal in length and slightly shorter than the laterals or the median pair longer and equal to laterals in C. simoni); preanal carina armed with a spine (vs. preanal carina unarmed in C. simoni). Caridina kutchi sp. nov. differs from C. babaulti, which is now reported from Gujarat, in possessing rostrum that is distinctly longer than antennal scale (vs. rostrum equal to antennular peduncle or shorter reaching middle of 3 rd antennular peduncle segment in C. babaulti); 12–22 teeth proximally leaving 0.5–0.65 of dorsal margin unarmed distally which is interrupted by a single tooth at distal end (vs. 14–25teeth proximally leaving 0.1–0.23 of dorsal margin unarmed distally in C. babaulti); 1–3 post orbital teeth present (vs. 3–7 postorbital teeth present in C. babaulti); 9–15 teeth proximally leaving 0.1–0.2 of ventral margin unarmed distally (vs. 3–8 teeth proximally leaving 0.1–0.45 of ventral margin unarmed distally in C. babaulti); carpus of first pereiopod with anterior excavation (vs. carpus of first pereiopod with deep anterior excavation in C. babaulti); telson posterior margin narrow and triangular, with a median projection (vs. telson posterior margin broad and rounded, with or without median protrusion in C. babaulti); 2–3 pairs of sparsely plumose intermediate spines of equal length and distinctly shorter than laterals (vs. 2–4 pairs or 5 sparsely plumose intermediate spines of varying length; fractionally longer or shorter than the lateral spines in C. babaulti); 8–12 uropod diaeresis spinules (vs. 12–21 uropod diaeresis spinules in C. babaulti); preanal carina armed with a spine (vs. preanal carina unarmed in C. babaulti). Caridina kutchi sp. nov. is the first Caridna species to be described from Kutch district, Gujarat state, which is known for its complex geological set up.Published as part of Pandya, Pranav J. & Richard, Jasmine, 2019, Report of Caridina babaulti Bouvier, 1918 (Crustacea: Decapoda: Caridea: Atyidae) and description of a new species Caridina kutchi sp. nov. from Gujarat, India, pp. 470-482 in Zootaxa 4568 (3) on pages 477-480, DOI: 10.11646/zootaxa.4568.3.3, http://zenodo.org/record/260166
Supplementary movies for "A unified model for the co-evolution of galaxies and their circumgalactic medium: the relative roles of turbulence and atomic cooling physics"
No full text================================================================================
Title: A unified model for the co-evolution of galaxies and their
circumgalactic medium: the relative roles of turbulence and
atomic cooling physics
Authors: Pandya V., Fielding D.B., Bryan G.L., Carr C., Somerville R.S.,
Stern J., Faucher-Giguere C.-L., Hafen Z., Angles-Alcazar D., Forbes J.C.
================================================================================
Description of contents: This Dataverse deposit includes supplementary movies for various simulated dark matter halos as described in the accepted version of the article listed above. The movies show the time evolution of the equilibrium temperature and turbulent velocity for the circumgalactic medium based on a novel semi-analytic model.
A list of the individual files and halos are given
below.
Halo Animation Description
---- --------- -----------
m10q supplementary_animation_m10q.mp4 ultrafaint dwarf, Mvir~10+10Msun
m10y supplementary_animation_m10y.mp4 ultrafaint dwarf, Mvir~10+10Msun
m10z supplementary_animation_m10z.mp4 ultrafaint dwarf, Mvir~10+10Msun
m11a supplementary_animation_m11a.mp4 intermediate-mass dwarf, Mvir~10+11 Msun
m11c supplementary_animation_m11c.mp4 intermediate-mass dwarf, Mvir~10+11 Msun
m11b supplementary_animation_m11b.mp4 intermediate-mass dwarf, Mvir~10+11 Msun
m11f supplementary_animation_m11f.mp4 intermediate-mass dwarf, Mvir~10+11 Msun
m11q supplementary_animation_m11q.mp4 intermediate-mass dwarf, Mvir~10+11 Msun
m11v supplementary_animation_m11v.mp4 intermediate-mass dwarf, Mvir~10+11 Msun
m12f supplementary_animation_m12f.mp4 MW-mass halo, Mvir~10+12 Msun
m12i supplementary_animation_m12i.mp4 MW-mass halo, Mvir~10+12 Msun
m12m supplementary_animation_m12m.mp4 MW-mass halo, Mvir~10+12 Msun; also shown in the accepted article as Figure 14)
System requirements: These are all standard H.264 encoded, MP4 animations.
Additional comments: The movies were created based on a semi-analytic model that treats the co-evolution of galaxies and their circumgalactic medium. The panels and main takeaways of the movies are described in the caption of Figure 14 from the article referenced above.
===============================================================================
Galaxies Going Bananas: Inferring the 3D Geometry of High-Redshift Galaxies with JWST-CEERS
No full textThis repository provides all figures for the Astrophysical Journal article "Galaxies Going Bananas: Inferring the 3D Geometry of High-Redshift Galaxies with JWST-CEERS" by Viraj Pandya et al. We also include a machine readable version of Table 2.
Below we describe the four figure sets corresponding to Figures 7, 8, 13 and 23 in the paper as well as Table 2. This repository also includes all individual figures not comprising sets -- for a description of these, we refer the user to their corresponding captions in the paper.
Figure 7 shows corner plots from our constrained Bayesian model for 3D galaxy shapes in a single mass-redshift bin. The figure set here includes analogous corner plots for the other mass-redshift bins.
Figure 8 shows the fractional contribution of ellipsoids of different types (prolate, oblate, spheroidal) to the observed joint distribution of projected axis ratios and sizes. This figure also shows that we can use these fractional model contributions to assign 3D shape probabilities to individual observed galaxies. The figure set here includes analogous figures for other mass-redshift bins and for our model applied independently to the SE++ and Galfit data.
Figure 13 shows a histogram of 3D axis ratios (C/A vs B/A) computed as the average of 500 draws from our model posterior for every mass-redshift bin. The version in the paper is for our model applied to the SE++ data. The additional figure here is for our model applied to the Galfit data.
Figure 23 shows mock parameter recovery tests for Hamiltonian Monte Carlo applied to our Bayesian 3D galaxy shape model with different sample sizes. The version in the paper is for a mock population of ellipsoids dominated by prolate objects. The additional figures here are for additional mock populations dominated by either spheroids, oblate (axisymmetric) disks, or triaxial (oval) disks.
Table 2 summarizes the means and standard deviations of our Bayesian model as well as ellipsoid class fractions for every mass-redshift bin. The results from both of our models based on Galfit and SE++ have been combined into this single table. This is a machine readable table that can easily be read in with, e.g., the Python astropy.table module
Going Beyond Counting First Authors in Author Co-citation Analysis
No full textThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Mitomycin C in highly myopic eyes - Author reply
No full textOphthalmology. 2005 Feb;112(2):208-18; discussion 219.
Mitomycin C modulation of corneal wound healing after photorefractive keratectomy in highly myopic eyes.
Gambato C, Ghirlando A, Moretto E, Busato F, Midena E.
SourceRefractive Surgery Service and Antimetabolite Therapy Research Unit, Department of Ophthalmology, University of Padova, Padova, Italy.
Abstract
PURPOSE: To evaluate the role of topical mitomycin C in corneal wound healing (CWH) after photorefractive keratectomy (PRK) in highly myopic eyes.
DESIGN: Prospective, double-masked, randomized clinical trial.
PARTICIPANTS: Seventy-two eyes of 36 patients affected by high (>7 diopters) myopia.
METHODS: In each patient, one eye was randomly assigned to PRK with intraoperative topical 0.02% mitomycin C application, and the fellow eye was treated with a placebo. Postoperatively, mitomycin C-treated eyes received artificial tears (3 times daily, tapered in 3 months), whereas the fellow eye was treated with fluorometholone sodium 2% and artificial tears (3 times daily, tapered in 3 months).
MAIN OUTCOME MEASURES: Uncorrected visual acuity (UCVA) and best-corrected visual acuity (BCVA), contrast sensitivity, manifest refraction, and biomicroscopy. Contrast sensitivity was determined using the Pelli-Robson chart. Corneal confocal microscopy documented CWH.
RESULTS: Mean follow-up was 18 months (range, 12-36). No side effects or toxic effects were documented. At 12-month follow-up examination, UCVAs (logarithm of the minimum angle of resolution) were 0.4+/-0.48 and 0.5+/-0.53 (P = .03) in mitomycin C-treated eyes and corticosteroid-treated eyes, respectively. At 1 year, corneal haze developed in 20% of corticosteroid-treated eyes, versus 0% of mitomycin C-treated eyes. At 12, 24, and 36 months, corneal confocal microscopy showed activated keratocytes and extracellular matrix significantly more evident in untreated eyes (Ps = 0.004, 0.024, and 0.046, respectively).
CONCLUSION: Topical intraoperative application of 0.02% mitomycin C can reduce haze formation in highly myopic eyes undergoing PRK.
Comment in
Ophthalmology. 2006 Feb;113(2):357; author reply 357-8
Co-discovery - a staff/student collaborative evaluation of the value of Broadening within the undergraduate student journey; the case for language learning
No full textThis was a collaborative student and staff enterprise between two staff and three UG student researchers – Akeisha Brown, Chandni Pandya and Robert Irnazarow – who together co-constructed and co-delivered the planning, execution and dissemination of the project’s objectives and outputs
Dispelling the Myths Behind First-author Citation Counts
No full textWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
After Belonging: Architecture, Nation, Difference
No full textThis book breaks new ground in demystifying the relationship between architecture, nationhood, and other forms of collective identity. It attempts to extricate the oppressive ideology of national identity entrenched within the very idea of architecture. Authors investigate themes such as cosmopolitanism, diaspora, geopolitics, globalisation, hybridity, and race. Certain chapters expose highly regulated environments which support cultural hegemony, such as the context of a hostel for ‘coloured colonial seamen’ in London, the illusionary rhetoric of ‘authenticity’ used to legitimise architectural conservation, and the role of the mosque as mediator between a post-war, multi-racial Britain, and ideas of nationhood. Others engage subjects at the urban scale, including the phenomena of universities transcending their nation-building roots to become agents of cosmopolitan urbanism, and how the discursive context of a high-profile yet unrealised modernist office-block in the City of London sustained a culture of British faux-nationalism. Remaining chapters adopt a postcolonial lens, with one examining how particular works of literary fiction reimagine notions of ‘place’ within an emerging intercultural nation, and another exploring the tense relationship between identitarian form and affective atmospheres to suggest the possibility of anti-essentialist experiences of architecture.
Together, these perspectives propose an alternative vision of the City, where neither state-sponsored identity politics nor right-wing populism determine the cultural context within which architects design for our collective urban experience. This book will be of interest to researchers and advanced students of Architecture, Anthropology, History, Human Geography, Politics, Sociology, and Urban Studies
Mikroblogien kirjoittajien demografia- ja sentimenttianalyysi
No full textAbstractOwing to the peculiar nature of the discourse on Twitter, developing analytical frameworks to derive useful insights from Twitter remains challenging as evidenced by the poor performance at tasks such as reliable demographic inference, affect estimation, and event detection. One of the focal problems lies in analyzing short texts in general, and tweets in particular. The analysis is as such made difficult because of the vagaries of the linguistic expressions and Twitter further exacerbates this by enabling the use of emojis, hashtags, URLs, and embedded media. While the previous research has demonstrated ways of extracting useful information from individual tweet-texts to some extent, a detailed and thorough investigation of the role of metadata has not yet been systematically performed. Furthermore, a majority of the previous work has paid little or no attention to the emerging role of deep learning approaches in Twitter-based analytics. These observations motivate this thesis, which aims to enhance machine understanding of tweets towards deriving deeper insights from the public data on Twitter and inform the scientific objectives of this thesis. First, this thesis sets out to empirically investigate the impact and efficacy of deep learning approaches integrating message-text and metadata leveraging on the distributed semantic representations of textual entities. Second, the thesis contributes towards improving capturing enhanced semantics from tweets by harnessing external, open-sourced knowledge graphs and other crowd-sourced lexical resources. Third, the role of the user-created metadata, such as hashtags and URLs, in machine understanding of tweets is examined and quantified. At the same time, computational models are introduced to derive conversational, topical, and temporal contexts of tweets and utilize them in machine learning models to improve Twitter-based analytics. Validation of the proposed novel machine learning models integrating the diverse footprints of users’ online activity/behavior is achieved by employing them in various case study applications. In addition, the datasets and the tools developed during this thesis have been made available publicly for the scientific community.Original papersOriginal papers are not included in the electronic version of the dissertation.Pandya, A., & Oussalah, M. (2017). Novel semantics-based distributed representations for message polarity classification using deep convolutional neural networks. Proceedings of the 9th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management, 71–82. https://doi.org/10.5220/0006500800710082Self-archived versionPandya, A., Oussalah, M., Monachesi, P., & Kostakos, P. (2020). On the use of distributed semantics of tweet metadata for user age prediction. Future Generation Computer Systems, 102, 437–452. https://doi.org/10.1016/j.future.2019.08.018Self-archived versionPandya, A., Oussalah, M., Monachesi, P., Kostakos, P., & Loven, L. (2018). On the use of URLs and hashtags in age prediction of Twitter users. 2018 IEEE International Conference on Information Reuse and Integration (IRI), 62–69. https://doi.org/10.1109/IRI.2018.00017Self-archived versionKostakos, P., Pandya, A., Kyriakouli, O., & Oussalah, M. (2018). Inferring demographic data of marginalized users in Twitter with computer vision APIs. 2018 European Intelligence and Security Informatics Conference (EISIC), 81–84. https://doi.org/10.1109/EISIC.2018.00022Self-archived versionPandya, A., Oussalah, M., Kostakos, P., & Fatima, U. (2020). MaTED: Metadata-assisted Twitter event detection system. Communications in Computer and Information science, 1237, 402–414. https://doi.org/10.1007/978-3-030-50146-4_30Self-archived versionKostakos, P., Sprachalova, L., Pandya, A., Aboeleinen, M., & Oussalah, M. (2018). Covert online ethnography and machine learning for detecting individuals at risk of being drawn into online sex work. 2018 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 1096–1099. https://doi.org/10.1109/ASONAM.2018.8508276Self-archived versionTiivistelmäTwitter-pohjainen analytiikka on noussut useiden tieteenalojen työkalupakkiin viime vuosina. Kuitenkin, järjestelmällisten analyysikokonaisuuksien kehitys on mikroblog-keskustelujen erityisluonteen vuoksi haastavaa. Analysointimenetelmien heikko suorituskyky on todettu useissa sovelluskohteissa, kuten kirjoittajien väestörakenne- ja tunnetila-analyyseissa taikka tehtävissä, joissa mikrobloggauksista pyritään havaitsemaan tärkeitä tapahtumia. Analyysit pitäisi suorittaa hyvin lyhyistä tekstipätkistä, tässä tutkimuksessa erityisesti mikroblogauksista. Omalaatuisten ja persoonallisten kielellisten ilmaisujen, mutta myös Twitterin emojien, metatietotagien, ulkoisten linkkien (url) ja upotettujen kuvien sekä videoiden käyttö monipuolistaa ongelmakenttää. Aikaisemmissa tutkimuksissa on onnistuttu johtamaan hyödyllistä tietoa yksittäisistä mikroblogauksista jossain määrin, mutta metatietojen roolia ja merkitystä ei ole vielä järjestelmällisesti eikä yksityiskohtaisesti tutkittu. Lisäksi syväoppimisen hyödyntämistä Twitter-pohjaisten datojen analyyseissa on tutkittu vähän tai ei ollenkaan. Tämän väitöskirjan tavoitteena on parantaa tietokoneiden valmiuksia käsitellä mikroblogauksia siten, että nykyistä parempi ja merkityksellisempi julkisten Twitter-aineistojen koneellinen ymmärtäminen olisi mahdollista. Ensinnäkin, tutkimuksessa testataan empiirisesti syväoppivan mallin vaikuttavuutta sekä tehokkuutta ym. tekstikokonaisuuksien hajautetun semanttisen esitysmuodon integroinnissa. Toiseksi, työssä parannetaan mikroblogauksien sisältöanalyysia ulkoisten, avoimen lähdekoodin tietograafien sekä muiden joukkoistettujen sanastojen avulla. Kolmanneksi tutkitaan ja kvantifioidaan käyttäjien luomien metadatojen, kuten metatietotagien ja ulkoisten linkkien roolit analyysikehikoissa. Työssä esitellään laskennalliset mallit mikroblogauksien keskusteluun, aihepiiriin sekä aikaan liittyvien asiayhteyksien päättelemiseksi ja käytetään näitä malleja koneoppimismallien suorituskyvyn parantamiseksi Twitter-dataan pohjautuvassa analytiikassa. Mikroblogaajien verkkokäyttäytymisen perusteella saadun monimuotoisen aineiston integrointi tapahtuu koneoppivien mallien avulla. Työssä käytetyt aineistot sekä tutkimuksessa kehitetyt työkalut on saatettu julkiseksi tiedeyhteisön käyttöön.OsajulkaisutOsajulkaisut eivät sisälly väitöskirjan elektroniseen versioon.Pandya, A., & Oussalah, M. (2017). Novel semantics-based distributed representations for message polarity classification using deep convolutional neural networks. Proceedings of the 9th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management, 71–82. https://doi.org/10.5220/0006500800710082Rinnakkaistallennettu versioPandya, A., Oussalah, M., Monachesi, P., & Kostakos, P. (2020). On the use of distributed semantics of tweet metadata for user age prediction. Future Generation Computer Systems, 102, 437–452. https://doi.org/10.1016/j.future.2019.08.018Rinnakkaistallennettu versioPandya, A., Oussalah, M., Monachesi, P., Kostakos, P., & Loven, L. (2018). On the use of URLs and hashtags in age prediction of Twitter users. 2018 IEEE International Conference on Information Reuse and Integration (IRI), 62–69. https://doi.org/10.1109/IRI.2018.00017Rinnakkaistallennettu versioKostakos, P., Pandya, A., Kyriakouli, O., & Oussalah, M. (2018). Inferring demographic data of marginalized users in Twitter with computer vision APIs. 2018 European Intelligence and Security Informatics Conference (EISIC), 81–84. https://doi.org/10.1109/EISIC.2018.00022Rinnakkaistallennettu versioPandya, A., Oussalah, M., Kostakos, P., & Fatima, U. (2020). MaTED: Metadata-assisted Twitter event detection system. Communications in Computer and Information science, 1237, 402–414. https://doi.org/10.1007/978-3-030-50146-4_30Rinnakkaistallennettu versioKostakos, P., Sprachalova, L., Pandya, A., Aboeleinen, M., & Oussalah, M. (2018). Covert online ethnography and machine learning for detecting individuals at risk of being drawn into online sex work. 2018 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 1096–1099. https://doi.org/10.1109/ASONAM.2018.8508276Rinnakkaistallennettu versioAcademic dissertation to be presented with the assent of the Doctoral Training Committee of Information Technology and Electrical Engineering of the University of Oulu for public defence in the Tönning auditorium (L4), Linnanmaa, on 20 November 2020, at 12 noonAbstract
Owing to the peculiar nature of the discourse on Twitter, developing analytical frameworks to derive useful insights from Twitter remains challenging as evidenced by the poor performance at tasks such as reliable demographic inference, affect estimation, and event detection. One of the focal problems lies in analyzing short texts in general, and tweets in particular. The analysis is as such made difficult because of the vagaries of the linguistic expressions and Twitter further exacerbates this by enabling the use of emojis, hashtags, URLs, and embedded media. While the previous research has demonstrated ways of extracting useful information from individual tweet-texts to some extent, a detailed and thorough investigation of the role of metadata has not yet been systematically performed. Furthermore, a majority of the previous work has paid little or no attention to the emerging role of deep learning approaches in Twitter-based analytics. These observations motivate this thesis, which aims to enhance machine understanding of tweets towards deriving deeper insights from the public data on Twitter and inform the scientific objectives of this thesis. First, this thesis sets out to empirically investigate the impact and efficacy of deep learning approaches integrating message-text and metadata leveraging on the distributed semantic representations of textual entities. Second, the thesis contributes towards improving capturing enhanced semantics from tweets by harnessing external, open-sourced knowledge graphs and other crowd-sourced lexical resources. Third, the role of the user-created metadata, such as hashtags and URLs, in machine understanding of tweets is examined and quantified. At the same time, computational models are introduced to derive conversational, topical, and temporal contexts of tweets and utilize them in machine learning models to improve Twitter-based analytics. Validation of the proposed novel machine learning models integrating the diverse footprints of users’ online activity/behavior is achieved by employing them in various case study applications. In addition, the datasets and the tools developed during this thesis have been made available publicly for the scientific community.Tiivistelmä
Twitter-pohjainen analytiikka on noussut useiden tieteenalojen työkalupakkiin viime vuosina. Kuitenkin, järjestelmällisten analyysikokonaisuuksien kehitys on mikroblog-keskustelujen erityisluonteen vuoksi haastavaa. Analysointimenetelmien heikko suorituskyky on todettu useissa sovelluskohteissa, kuten kirjoittajien väestörakenne- ja tunnetila-analyyseissa taikka tehtävissä, joissa mikrobloggauksista pyritään havaitsemaan tärkeitä tapahtumia. Analyysit pitäisi suorittaa hyvin lyhyistä tekstipätkistä, tässä tutkimuksessa erityisesti mikroblogauksista. Omalaatuisten ja persoonallisten kielellisten ilmaisujen, mutta myös Twitterin emojien, metatietotagien, ulkoisten linkkien (url) ja upotettujen kuvien sekä videoiden käyttö monipuolistaa ongelmakenttää. Aikaisemmissa tutkimuksissa on onnistuttu johtamaan hyödyllistä tietoa yksittäisistä mikroblogauksista jossain määrin, mutta metatietojen roolia ja merkitystä ei ole vielä järjestelmällisesti eikä yksityiskohtaisesti tutkittu. Lisäksi syväoppimisen hyödyntämistä Twitter-pohjaisten datojen analyyseissa on tutkittu vähän tai ei ollenkaan. Tämän väitöskirjan tavoitteena on parantaa tietokoneiden valmiuksia käsitellä mikroblogauksia siten, että nykyistä parempi ja merkityksellisempi julkisten Twitter-aineistojen koneellinen ymmärtäminen olisi mahdollista. Ensinnäkin, tutkimuksessa testataan empiirisesti syväoppivan mallin vaikuttavuutta sekä tehokkuutta ym. tekstikokonaisuuksien hajautetun semanttisen esitysmuodon integroinnissa. Toiseksi, työssä parannetaan mikroblogauksien sisältöanalyysia ulkoisten, avoimen lähdekoodin tietograafien sekä muiden joukkoistettujen sanastojen avulla. Kolmanneksi tutkitaan ja kvantifioidaan käyttäjien luomien metadatojen, kuten metatietotagien ja ulkoisten linkkien roolit analyysikehikoissa. Työssä esitellään laskennalliset mallit mikroblogauksien keskusteluun, aihepiiriin sekä aikaan liittyvien asiayhteyksien päättelemiseksi ja käytetään näitä malleja koneoppimismallien suorituskyvyn parantamiseksi Twitter-dataan pohjautuvassa analytiikassa. Mikroblogaajien verkkokäyttäytymisen perusteella saadun monimuotoisen aineiston integrointi tapahtuu koneoppivien mallien avulla. Työssä käytetyt aineistot sekä tutkimuksessa kehitetyt työkalut on saatettu julkiseksi tiedeyhteisön käyttöön
- …
