29 Apr 2013

The liquid helium coolant that enabled instruments on board the Herschel space observatory to collect extraordinary images and spectra has finally run out. Launched in 2009, the ESA mission collected unprecedented data of the cool as well as of the distant Universe. Herschel's observations have exceeded expectations, enabling scientists to learn more about how stars form, about the rates of star formation in galaxies across the cosmos, and about the origin and presence of water in different celestial bodies. While observations have come to an end and the spacecraft is to be propelled to a stable parking orbit around the Sun, where it will remain indefinitely, the science mission will continue for several years with many discoveries still to be made in the treasure trove of images and spectra collected by the observatory.

Herschel cooling system and build up. Credit: ESA/AOES Medialab (Click here for further details about this animation.)

Observations on ESA's Herschel space observatory ceased on 29 April 2013, as the supply of liquid helium coolant on board the spacecraft had completely evaporated. The most powerful infrared telescope ever flown in space, Herschel operated at far-infrared and sub-millimetre wavelengths, being sensitive to a wide range of low temperatures from a few hundred to less than ten degrees above absolute zero.

"To observe very cold regions with high sensitivity, Herschel's instruments needed to be at even lower temperatures – only a few degrees above absolute zero," explains Thomas Passvogel, Head of the Projects Department in ESA's Directorate of Science and Robotic Exploration, and former Herschel and Planck Programme Manager. "This was achieved through the use of 2300 litres of superfluid helium, which granted us over three and a half years of observations of the cool and distant Universe," he adds.

"Herschel's observations have revealed the cosmos in unprecedented detail at these wavelengths," comments Göran Pilbratt, Herschel Project Scientist at ESA. "We have very exciting results from the observatory's first few years and we are looking forward to many more exciting discoveries," he adds.

The Cygnus X star-forming complex. Credit: ESA/PACS/SPIRE/Martin Hennemann & Frédérique Motte, Laboratoire AIM Paris-Saclay, CEA/Irfu - CNRS/INSU - Univ. Paris Diderot, France

Spanning 12 billion years of star formation in the Universe

Herschel's chief goal is the study of star formation throughout the cosmos, so astronomers have used it to survey thousands of galaxies across a wide span of the Universe's history. These data revealed that galaxies in the first few billion years of cosmic history contributed many more stars than had been previously estimated using observations at shorter or longer wavelengths.

A new census of starburst galaxies across the Universe. Credit: ESA - C. Carreau/C. Casey (University of Hawai'i); COSMOS field: ESA/Herschel/SPIRE/HerMES Key Programme; Hubble images: NASA, ESA

"Herschel has allowed us to see for the first time a huge population of intensively star-forming galaxies that make up about half of the entire star formation budget of the Universe," comments Matt Griffin from Cardiff University, UK, who is the Principal Investigator of the Spectral and Photometric Imaging Receiver (SPIRE) on board Herschel. "This highlights the importance of looking at the distant Universe in the spectral range covered by Herschel. Stars form in clouds of gas and dust, and most of the energy released by forming stars is absorbed by the dust grains, causing them to warm up. We can detect this warm dust with Herschel and so measure the energy produced by the young stars."

Previous studies hinted that the high star-forming rate of galaxies in the past may be due to episodes of intense star formation, possibly triggered by mergers between massive galaxies. But Herschel has caused a re-think.

"Herschel data suggest that mergers may not be required to generate intense star formation as previously thought. The crucial point appears to be the availability of sufficient gas to make stars. Massively star-forming galaxies could be fed by gas-rich mergers, but also by other processes such as intergalactic streams of cold gas," explains Pilbratt. "The issue remains open, but further analysis of Herschel data along with follow-up observations of these galaxies with other telescopes will hold the key," he adds.

Left: Artist's impression of a galaxy forming stars in a gentle and steady fashion. Right: Artist's impression of an Ultra-Luminous InfraRed Galaxy (ULIRG) with outflows of molecular gas. Credit: ESA/AOES Medialab

Another study based on Herschel data has assessed that these prolific galaxies can only form in the most massive haloes of dark matter – those with a mass as large as 300 billion times the mass of the Sun. Numerical simulations suggested to astronomers that galaxies take shape in the densest knots of the cosmic web – the wispy network of dark matter throughout the Universe. Herschel has allowed them to measure the mass threshold for the cradles where the most productive galaxies can form, which may be linked to the physical processes triggering the intense star-forming activity of these galaxies.

The change in star formation rates over time may also be linked to the supermassive black hole assumed to exist at the centre of most galaxies. In so-called active galaxies, the central black hole feeds on the surrounding matter at extraordinarily high rates. The radiation pressure created by this accretion triggers an outflow of material, which may drain the galaxy's reservoir of gas and quench its star formation.

"By monitoring a number of relatively nearby active galaxies with Herschel, we have detected, for the first time, vast amounts of outflowing gas that is molecular in nature," comments Albrecht Poglitsch from the Max-Planck-Institut für Extraterrestrische Physik, Germany, who is Principal Investigator of the Photodetector Array Camera and Spectrometer (PACS) on board Herschel. "This shows us that these outflows may indeed deplete a galaxy's supply to form stars, since it is from molecular gas that stars are made," he adds.

Astronomers have used Herschel's galaxy surveys, along with observations at other wavelengths, to search for a direct link between star formation and the activity of the central black hole in large samples of more distant galaxies. But these investigations have led to a variety of different results, indicating that more research will be needed to solve this particular puzzle.

Delving deep into the formation of stars and planetary systems in the Milky Way

Dark cloud in the Southern Cross not so dark. Credit: ESA and the SPIRE & PACS consortium

Observations closer to home, in our Galaxy, the Milky Way, yielded extraordinary revelations about the physical processes that give birth to stars. Herschel identified the presence of a filamentary network nearly everywhere in the interstellar medium that pervades the plane of our Galaxy, filaments which lead to the formation of stars.

"We did expect that there would be filaments in the interstellar medium, but Herschel finally gave us proof that they are omnipresent and that they are key to allowing star formation to occur," explains Matt Griffin. "Herschel's beautifully detailed images show how some clouds have developed filaments so dense that they are collapsing under their own weight and starting to form stars, while other clouds exhibit a more chaotic tangle of more tenuous filaments and are not dense enough to form stars – at least, not yet," he adds.

The emergence of filaments and their subsequent collapse into stars may also be linked to magnetic fields in interstellar clouds; Herschel's results have prompted further research into the role of magnetic fields in star formation, which can no longer be neglected.

Left: The star-forming cloud IC 5146. Credit: ESA/Herschel/SPIRE/PACS/D. Arzoumanian (CEA Saclay) for the "Gould Belt survey" Key Programme Consortium.  Right: The Polaris Flare. Credit: ESA/Herschel/ SPIRE/Ph. André (CEA Saclay) for the "Gould Belt survey" Key Programme Consortium and A. Abergel (IAS Orsay) for the "Evolution of Interstellar Dust" Key Programme Consortium

Herschel image of the star Fomalhaut and its debris disc.Credit:ESA/Herschel/PACS/ Bram Acke, KU Leuven, Belgium

Herschel has given a unique view of the subsequent stages in the formation of stars, and possibly of planetary systems around them. The image of a massive debris disc around nearby star Fomalhaut suggests a resemblance between this structure and the Kuiper belt as it may have appeared in the early days of our Solar System. The vast amount of dust in Fomalhaut's debris disc is produced by frequent collisions between comets within the disc. Herschel has detected more debris discs around other nearby stars, some of which host planets: a closer inspection of these data hints that debris discs may survive more easily in planetary systems without very massive planets like Jupiter.

"The study of extra-solar debris discs is providing us with new clues to the formation of planetary systems, and may eventually help us to achieve a more complete view of how our own Solar System formed and evolved from its parent cloud," says Albrecht Poglitsch.

Investigating the origin of water in planetary systems

Another highlight from Herschel is the study of the chemical composition of cosmic objects through high-resolution spectroscopy, and in particular the search for water in a variety of environments.

Water in L1544. Credit: ESA/Herschel/ SPIRE/HIFI/Caselli et al.

Studies based on Herschel data have indicated that a cold pre-stellar core in the vast Taurus molecular cloud hosts a reservoir of water ice amounting to a few million times the amount of water stored in Earth's oceans, and that the massive protoplanetary disc surrounding nearby star TW Hydrae contains a water vapour supply equivalent to several thousand times Earth's oceans.

"These results show that there is plenty of water available to enrich the surface of future planets, should they form in similar discs around stars other than the Sun," says Frank Helmich from SRON Netherlands Institute for Space Research, the Netherlands, who is the Principal Investigator of the Heterodyne Instrument for the Far Infrared (HIFI) on board Herschel.

Herschel image and spectrum of comet 103P/Hartley 2 and its orbit. Credit: ESA/AOES Medialab; Herschel/HssO Consortium

Herschel has also contributed to  the debate about the origin of water in the Solar System, and on Earth in particular. Astronomers believe that water was not present on the surface of our planet during its early days, but was brought here by comets and asteroids which stored water as ice. Asteroids were considered to be the primary source, but the case for comets strengthened after Herschel detected water in comet 103P/Hartley 2 with the same isotopic composition as Earth's oceans. Later observations of comet C/2009 P1 Garradd from Herschel found a different isotopic composition so the role of comets is still uncertain.

"The comparison between the composition of water in comets and on Earth's oceans may have far-reaching implications: if comets did contribute to enriching our planet with water, they might also be responsible for the delivery of other species, such as the carbon- and nitrogen-based compounds that are of great importance to the emergence of life," explains Helmich.

A peek at the future

Astronomers worldwide have so far used Herschel data to produce over 600 scientific papers published in refereed journals. But this is only the beginning: the careful analysis of Herschel's images and spectra will generate many more science discoveries and new insights for years to come.

"While the observations have ceased and the spacecraft will soon be propelled to a stable orbit around the Sun where it will remain indefinitely, Herschel as a science mission is by no means over," comments Pilbratt. "Most of Herschel's data are publicly available already. We will continue supporting the community exploiting the data, collecting and producing the best possible data products in the form of maps, spectra, and various catalogues to benefit all astronomers working with Herschel. We are looking forward to the multitude of great discoveries that are still ahead of us."

Notes for editors

Herschel is an ESA space observatory sensitive to far-infrared and sub-millimetre wavelengths, with science instruments provided by European-led Principal Investigator consortia and with important participation from NASA. Launched on 14 May 2009, Herschel is equipped with a primary mirror 3.5 m across, making it the largest, most powerful infrared telescope ever flown in space.

Its payload consists of three instruments: the Photodetector Array Camera and Spectrometer (PACS), the Spectral and Photometric Imaging Receiver (SPIRE), and the Heterodyne Instrument for the Far Infrared (HIFI). All three instruments were cooled to -271°C inside a cryostat filled with 2300 litres of liquid superfluid helium, which weighed 335 kg – nearly 10 percent of Herschel's mass at launch. After almost four years of observations, the mission finally exhausted its coolant today.

Herschel has executed over 35 000 scientific observations, amassing more than 25 000 hours worth of science data from about 600 observing programmes. A further 2000 hours of calibration observations also contribute to the rich dataset.

Herschel will continue communicating with its ground stations for a few weeks after the helium is exhausted, during which a range of technical tests will be performed. Finally, in May, it will be propelled into its long-term stable orbit around the Sun, where it will remain indefinitely, and safe from Earth impact for several hundreds of years.

The PACS instrument contains an imaging photometer (camera) and an imaging spectrometer. The camera operates in three bands centred on 70, 100, and 160 μm, respectively, and the spectrometer covers the wavelength range between 51 and 220 μm. PACS has been developed by a consortium of institutes led by MPE (Germany) and including UVIE (Austria); KU Leuven, CSL, IMEC (Belgium); CEA, LAM (France); MPIA (Germany); INAF-IFSI/OAA/OAP/OAT, LENS, SISSA (Italy); IAC (Spain). This development has been supported by the funding agencies BMVIT (Austria), ESA-PRODEX (Belgium), CEA/CNES (France), DLR (Germany), ASI/INAF (Italy), and CICYT/MCYT (Spain).

The SPIRE instrument contains an imaging photometer (camera) and an imaging spectrometer. The camera operates in three wavelength bands centred on 250, 350 and 500 µm, and so can make images of the sky simultaneously in three sub-millimetre colours; the spectrometer covers the wavelength range between 194 and 671 μm. SPIRE has been developed by a consortium of institutes led by Cardiff Univ. (UK) and including Univ. Lethbridge (Canada); NAOC (China); CEA, LAM (France); IFSI, Univ. Padua (Italy); IAC (Spain); Stockholm Observatory (Sweden); Imperial College London, RAL, UCL-MSSL, UKATC, Univ. Sussex (UK); Caltech, JPL, NHSC, Univ. Colorado (USA). This development has been supported by national funding agencies: CSA (Canada); NAOC (China); CEA, CNES, CNRS (France); ASI (Italy); MCINN (Spain); SNSB (Sweden); STFC and UKSA (UK); and NASA (USA).

The HIFI instrument is a very high-resolution heterodyne spectrometer and operates in seven bands covering the wavelength range between 157 and 625 µm. HIFI has been designed and built by a consortium of institutes and university departments across Europe, Canada, and the United States under the leadership of SRON Netherlands Institute for Space Research, the Netherlands, with major contributions from Germany, France, and the US. HIFI Consortium members are: CSA, U. Waterloo (Canada); CESR, LAB, LERMA, IRAM (France); KOSMA, MPIfR, MPS (Germany); NUI Maynooth (Ireland); ASI, IFSI-INAF, Osservatorio Astrofisico di Arcetri-INAF (Italy); SRON, TUD (Netherlands); CAMK, CBK (Poland); Observatorio Astronómico Nacional (IGN), Centro de Astrobiología (CSIC-INTA) (Spain); Chalmers University of Technology - MC2, RSS & GARD, Onsala Space Observatory, Swedish National Space Board, Stockholm University - Stockholm Observatory (Sweden); ETH Zurich, FHNW (Switzerland); Caltech, JPL, NHSC (USA).

CONTACTS

Thomas Passvogel
Herschel and Planck Programme Manager
ESA, The Netherlands
Email: Thomas.Passvogelesa.int
Phone: +31-71-5655962

Göran Pilbratt
Herschel Project Scientist
Research and Scientific Support Department
Science and Robotic Exploration Directorate
ESA, The Netherlands
Email: gpilbrattrssd.esa.int
Phone: +31-71-5653621

Matt Griffin
Principal Investigator of SPIRE
School of Physics and Astronomy
Cardiff University, UK
Email: Matt.Griffinastro.cf.ac.uk
Phone: +44-29-208-74203

Albrecht Poglitsch
Principal Investigator of PACS
Max-Planck-Institut für extraterrestrische Physik (MPE)
Garching, Germany
Email: alpogmpe.mpg.de
Phone: +49-89-30000-3293

Frank Helmich
Principal Investigator for HIFI
SRON Netherlands Institute for Space Research
Groningen, The Netherlands
Email: F.P.Helmichsron.nl
Phone: +31-50-363-4799
 

Last Update: 29 Apr 2013

Source:  ESA -

SCIENCE AND TECHNOLOGY ,  HERSCHEL 

Astrophysics > Cosmology and Extragalactic Astrophysics

Inflationary paradigm in trouble after Planck2013

Anna Ijjas, Paul J. Steinhardt, Abraham Loeb

(Submitted on 9 Apr 2013)

The recent Planck satellite combined with earlier results eliminate a wide spectrum of more complex inflationary models and favor models with a single scalar field, as reported in the analysis of the collaboration. More important, though, is that all the simplest inflaton models are disfavored by the data while the surviving models -- namely, those with plateau-like potentials -- are problematic. We discuss how the restriction to plateau-like models leads to three independent problems: it exacerbates both the initial conditions problem and the multiverse-unpredictability problem and it creates a new difficulty which we call the inflationary "unlikeliness problem." Finally, we comment on problems reconciling inflation with a standard model Higgs, as suggested by recent LHC results. In sum, we find that recent experimental data disfavors all the best-motivated inflationary scenarios and introduces new, serious difficulties that cut to the core of the inflationary paradigm. Forthcoming searches for B-modes, non-Gaussianity and new particles should be decisive.

Subjects:	Cosmology and Extragalactic Astrophysics (astro-ph.CO); High Energy Physics - Phenomenology (hep-ph); High Energy Physics - Theory (hep-th)
Cite as:	arXiv:1304.2785 [astro-ph.CO]
	(or arXiv:1304.2785v1 [astro-ph.CO] for this version)

Submission history

From: Anna Ijjas [view email]
[v1] Tue, 9 Apr 2013 20:15:58 GMT (253kb)

Download:

• PDF
• PostScript
• Other formats

SOURCE:

myADS Personal Notification Service
Jorge Poveda
Sat Apr 27 02:51:57 2013
arXiv e-prints database

El cerebro crítico Posted: 24 Apr 2013 10:14 AM PDT Referencia: Physics.APS.org . por Dietmar Plenz, 22 de abril 2013 Un modelo describe el cerebro como un sistema cerrado en una transición de fase, puede capturar la dinámica global de la actividad cerebral observada en los experimentos de resonancia magnética funcional. La neurociencia está al borde de un avance sin precedentes en la obtención instantánea de los mapas detallados de actividad neuronal durante las funciones cerebrales superiores, como el reconocimiento de objetos y la toma de decisiones. Estos mapas vienen a ser posibles gracias a una rápida fertilización cruzada entre los campos (biología molecular, óptica, imagen, procesos de microfabricación y nanotecnologías), y se consideran cruciales para dar sentido a las actividades complejas que emergen de los miles de millones de neuronas que crean los billones de conexiones que constituyen el cerebro humano. - Imagen: (columna izquierda) experimentos de imagen de resonancia magnética funcional (fMRI) han revelado que el cerebro en reposo está organizado en varias áreas en las que se correlacionan las fluctuaciones de las actividades cerebrales, las así llamadas redes en estado de reposo (RSN). De arriba a abajo: visual Medial (VisM), visual lateral (VisL), auditiva (Aud) y sensorio-motor (SM) RSNs. (Columnas de la derecha) Resultados del trabajo de Haimovici et al. [1] que muestran que un modelo simple puede reproducir las propiedades estadísticas de las RSN sólo si el modelo está sintonizado con la criticidad (at TC). Ya se están llevando a cabo proyectos pertinentes de investigación a gran escala o están siendo planificados a nivel internacional: elProyecto Humano Conectoma y BRAIN (Brain Research través Avanzando neurotecnologías innovadoras) en EE.UU., el proyecto Cerebro Humano en Europa, y el proyecto Brainnetome en China. El objetivo general de estas grandes empresas es la de explicar el funcionamiento del cerebro, mediante la adquisición e integración de una información detallada sobre la estructura del cerebro y su comportamiento dinámico.  Sin embargo, estos esfuerzos ¿tendrán éxito? El conocimiento de tan finos detalles estructurales del cerebro y las observaciones de las actividades neuronales puede no ser suficiente si la aparición de formas del comportamiento colectivo no se captura correctamente. Ariel Haimovichi, de la Universidad de Buenos Aires en Argentina, y sus compañeros de trabajo [1] escribe en Physical Review Letters, que el fenómeno cooperativo desempeña un papel clave en la determinación de la dinámica cerebral, al mostrar que el cerebro en su estado de reposo (es decir, cuando no está realizando una tarea explícita) es un sistema en criticidad. Los sistemas críticos se pueden definir como sistemas que están cerca de un punto crítico, generalmente identificado como frontera de transición entre las fases de orden-desorden. Muchos sistemas complejos, lejos del equilibrio y composición de un gran número de elementos que interactúan han sido modelados con éxito como críticos: ejemplos notables de esto van desde las redes de interacción de genes hasta los mercados financieros.  En la criticidad, estos sistemas pueden evitar quedar atrapados en uno de dos casos extremos: un estado desordenado (cuando las interacciones son demasiado débiles y el sistema está dominado por el ruido) o un estado ordenado global en el que se bloquean todos los elementos (cuando las interacciones son demasiado fuertes y un sistema es completamente estático).  Ningún estado soporta el dualismo esencial de un sistema complejo como el cerebro para funcionar, debe mantener un cierto orden para garantizar un funcionamiento coherente (p. ej, generar un comportamiento reproducible en respuesta a un determinado estímulo) permitiendo al mismo tiempo un cierto grado de desorden que haga disponible la flexibilidad (o sea, adaptarse a las diferentes condiciones externas). En cambio, tal dualismo es posible gracias an la criticidad. Aunque son posibles muchos grados de orden/desorden, el sutil equilibrio entre el orden y el desorden en la criticidad se manifiesta en ciertas propiedades estadísticas generales: Los sistemas críticos exhiben correlaciones espaciales y temporales que son de gran amplitud (es decir, en escalas que son más grandes que aquellas las que las interacciones mutuas tengan efecto) y seguir las distribuciones de la ley de potencias. Durante la última década, numerosos estudios han encontrado que el cerebro en reposo presenta muchas características propias de un estado crítico, como la escala de la ley de potencias de aludes neuronales (estallidos intermitentes de actividad eléctrica observados en la corteza cerebral) [2 y 3], correlaciones espaciales y temporales de largo alcance no evidentes de la actividad neuronal [4], y la ampliación anómala de las fluctuaciones de la actividad con el tamaño de la red [5]. La investigación llevada a cabo en mi grupo ha demostrado que las redes cerebrales, por estar en estado crítico, optimizan su respuesta a los entradas y maximizan su capacidad de procesamiento de información [6]. La principal conclusión de los autores es que la información estructural y las dinámicas neuronales locales tienen que integrarse en el marco de la criticidad para explicar la actividad cerebral. Dado que una de las señas de identidad de la criticidad es la invariancia de escala (la propiedad según la cual se observan similares dinámicas/propiedades en todas las escalas de longitud), su enfoque podría ser relevante para los diferentes niveles de la organización cerebral, que van desde pequeños microcircuitos corticales, a las columnas corticales hasta el cerebro completo, proporcionando de este modo un marco unificador para la interpretación de la acumulación de información sobre una estructura y dinámica neuronal. Si bien en el presente estudio los autores tuvieron que afinar su modelo para un determinado umbral de criticidad, el mayor desafío será entender cuáles son los mecanismos físicos y biológicos exactos por los que el cerebro llega a la criticidad y la mantiene durante el aprendizaje y el desarrollo. - Publicación: “Brain Organization into Resting State Networks Emerges at Criticality on a Model of the Human Connectome”. Ariel Haimovici, Enzo Tagliazucchi, Pablo Balenzuela, y Dante R. Chialvo. Phys. Rev. Lett. 110, 178101 (2013). Publicado el 22 de abril 2013, en PDF. - Autor: Dietmar Plenz received his Ph.D. in Biology in 1993 at the Max Planck Institute of Biological Cybernetics, Germany. - Section on Critical Brain Dynamics, National Institute of Mental Health, NIH, Bethesda, MD 20892, USA Fuente: Pedro Donaire, Bitnavegantes

Nuevo algoritmo para evaluar y clasificar la literatura científica Posted: 24 Apr 2013 03:40 AM PDT Referencia: Kurzweilai.net, 23 de abril 2013 Mantenerse al día con la literatura científica actual es una tarea de gigantes, teniendo en cuenta que cada día se publican cientos de miles de documentos. Puestos a ello, los investigadores de la Universidad Estatal de Carolina del Norte han desarrollado un programa informático con capacidad de evaluar y clasificar los artículos científicos de campo. Los investigadores utilizan un algoritmo de minería de texto que dé prioridad a los trabajos de investigación para leer e incluir en su base de datos comparativa toxicogenómica (CTD), una base de datos pública mantenida y codificada de forma manual proveniente de los datos de la literatura científica, y que describe cómo los productos químicos ambientales interactúan con los genes que afectan a la salud humana. "Desde 1926, se han publicado sobre la toxicidad de metales pesados ​​más de 33.000 artículos científicos", explica el Dr. Allan Peter Davis, gestor del proyecto de bioconservación CTD en la estatal de Carolina del Norte, que trabajó en el proyecto y es co-autor principal de un artículo sobre este trabajo. "Simplemente no podemos leer y codificarlo todo. Y, con la ayuda de este nuevo algoritmo, ya no tenemos que hacerlo." Para ayudar a seleccionar los artículos más relevantes para su inclusión en el CTD, Thomas Wiegers, investigador bioinformático de la NC State y otro co-autor principal del informe, desarrollaron un sofisticado algoritmo como parte de un proceso de minería de textos. La aplicación evalúa el texto entre miles de trabajos y asigna una puntuación de relevancia a cada documento. "Esta clasificación al conjunto de artículos ayuda a separar el grano de la paja, por así decirlo," dice Wiegers. Pero, ¿en qué medida un algoritmo puede determinar los mejores trabajos? Para probar esto, los investigadores de extrajeron texto de 15.000 artículos y enviaron una muestra representativa a su equipo de bioconservadores para leer manualmente y evaluarlos por sí mismos, sin saber nada de la puntuación asignada. "Los resultados fueron impresionantes", comenta Davis. Los bioconservadores estuvieron de acuerdo con el algoritmo en un 85 por ciento de las veces respecto a los documentos más puntuados. Utilizar el algoritmo para clasificar documentos permitió a los bioconservadores centrarse en los trabajos más relevantes, aumentando la productividad en un 27 por ciento y en contenidos novedosos en un 100 por ciento. "Es un tremendo paso para ahorrar tiempo", explica Davis. "Con esto podemos distribuir nuestros recursos de manera mucho más efectiva haciendo que el equipo se centre en los documentos más informativos." Siempre hay valores atípicos en este tipo de experimentos: puede ocurrir que el algoritmo asigne una puntuación muy alta a un artículo que el bioconservador humano descarte rápidamente como irrelevante. El equipo que analizó los valores extremos a menudo fue capaz de ver un patrón en el por qué el algoritmo identificaba erróneamente un determinado documento como importante. "Ahora podemos volver atrás y modificar el algoritmo para dar cuenta de eso y afinar el sistema", dice Wiegers. "No estamos aún en ese punto en que un equipo pueda leer y extraer todos los datos relevantes por sí mismo", concluye Davis, "pero si con este proceso de minería de texto nos dirigirnos hacia los artículos más informativos es un gran primer paso. " Este trabajo fue apoyado por el Instituto Nacional de Ciencias de Salud Ambiental. - Imagen: CTD minería de texto, crédito:. Allan Peter Davis et al / PLoS ONE - Referencias: Allan Peter Davis et al., Text Mining Effectively Scores and Ranks the Literature for Improving Chemical-Gene-Disease Curation at the Comparative Toxicogenomics Database, PLoS ONE, 2013, DOI: 10.1371/journal.pone.0058201 (open access). Fuente: Pedro Donaire. Bitnavegantes.