Expansión supervisada de léxicos polarizados adaptable al contexto

El análisis de opiniones es un área en la cual múltiples disciplinas han otorgado diferentes enfoques para elaborar modelos que sean capaces de extraer la polaridad de los textos analizados. En función del dominio o categoría del texto analizado, donde ejemplos de categorías son Deportes o Banca, estos modelos deben ser modificados para obtener un análisis de opinión de calidad. En esta tesis se presenta un modelo que pretende elaborar un análisis de opiniones independiente de la categoría a analizar y un extenso estado del arte sobre análisis de opiniones. Se propone un enfoque cuantitativo que haría uso de un léxico polarizado semilla como único recurso cualitativo del modelo. El enfoque propuesto hace uso de un corpus anotado de textos por polaridad y categoría y el léxico polarizado semilla para producir un modelo capaz de elaborar un análisis de opinión de calidad en las distintas categorías analizadas y expandir el léxico polarizado semilla con términos que se adecúan a las categorías procesadas.---ABSTRACT---Sentiment analysis is an area in which multiple disciplines have given diferent approaches to make models that are able to extract the polarity of the analyzed texts. Depending on the domain or category of the analyzed text, where examples of categories are Sports or Banking, these models should be modified to obtain a good opinion analysis. This thesis presents a model that aims to develop a category independent opinion analysis model and a extensive sentiment analysis state of the art. A quantitative approach is proposed that will use a polarized lexicon as the only qualitative resource. The proposed approach uses an annotated corpus by polarity and category and a polarized lexicon seed to produce a model able to develop a good opinion analysis in the various categories analyzed and to expand the polarized lexicon seed with terms that fit the processed categories.

Is the polarity of content producers strongly influenced by the results of the event?

This paper presents an approach to compare two types of data, subjective data (Polarity of Pan American Games 2011 event by country) and objective data (the number of medals won by each participating country), based on the Pearson corre- lation. When dealing with events described by people, knowledge acquisition is difficult because their structure is heterogeneous and subjective. A first step towards knowing the polarity of the information provided by people consists in automatically classifying the posts into clusters according to their polarity. The authors carried out a set of experiments using a corpus that consists of 5600 posts extracted from 168 Internet resources related to a specific event: the 2011 Pan American games. The approach is based on four components: a crawler, a filter, a synthesizer and a polarity analyzer. The PanAmerican approach automatically classifies the polarity of the event into clusters with the following results: 588 positive, 336 neutral, and 76 negative. Our work found out that the polarity of the content produced was strongly influenced by the results of the event with a correlation of .74. Thus, it is possible to conclude that the polarity of content is strongly affected by the results of the event. Finally, the accuracy of the PanAmerican approach is: .87, .90, and .80 according to the precision of the three classes of polarity evaluated.

Hacia un Lexicón Unificado de Sentimientos basado en Unidades de Procesamiento Gráfico

Esta tesis presenta un modelo, una metodología, una arquitectura, varios algoritmos y programas para crear un lexicón de sentimientos unificado (LSU) que cubre cuatro lenguas: inglés, español, portugués y chino. El objetivo principal es alinear, unificar, y expandir el conjunto de lexicones de sentimientos disponibles en Internet y los desarrollados a lo largo de esta investigación. Así, el principal problema a resolver es la tarea de unificar de forma automatizada los diferentes lexicones de sentimientos obtenidos por el crawler CSR, porque la unidad de medida para asignar la intensidad de los valores de la polaridad (de forma manual, semiautomática y automática) varía de acuerdo con las diferentes metodologías utilizadas para la construcción de cada lexicón. La representación codificada de la estructura de datos de los términos presenta también una variación en la estructura de lexicón a lexicón. Por lo que al unificar en un lexicón de sentimientos se hace posible la reutilización del conocimiento recopilado por los diferentes grupos de investigación y se incrementa, a la vez, el alcance, la calidad y la robustez de los lexicones. Nuestra metodología LSU calcula un valor unificado de la intensidad de la polaridad para cada entrada léxica que está presente en al menos dos de los lexicones de sentimientos que forman parte de este estudio. En contraste, las entradas léxicas que no son comunes en al menos dos de los lexicones conservan su valor original. El coeficiente de Pearson resultante permite medir la correlación existente entre las entradas léxicas asignándoles un rango de valores de uno a menos uno, donde uno indica que los valores de los términos están perfectamente correlacionados, cero indica que no existe correlación y menos uno significa que están inversamente correlacionados. Este procedimiento se lleva acabo con la función de MetricasUnificadas tanto en la CPU como en la GPU. Otro problema a resolver es el tiempo de procesamiento que se requiere para realizar la tarea de unificación de la intensidad de la polaridad y con ello alcanzar una cobertura mayor de lemas en los lexicones de sentimientos existentes. Asimismo, la metodología LSU utiliza el procesamiento paralelo para unificar los 155 802 términos. El algoritmo LSU procesa mediante cargas iguales el subconjunto de entradas léxicas en cada uno de los 1344 núcleos en la GPU. Los resultados de nuestro análisis arrojaron un total de 95 430 entradas léxicas donde 35 201 obtuvieron valores positivos, 22 029 negativos y 38 200 neutrales. Finalmente, el tiempo de ejecución fue de 2,506 segundos para el total de las entradas léxicas, lo que permitió reducir el procesamiento de cómputo hasta en una tercera parte con respecto al algoritmo secuencial. De estos resultados se concluye que al lograr un lexicón de sentimientos unificado que permite homogeneizar la intensidad de la polaridad de las unidades léxicas (con valores positivos, negativos y neutrales) deriva no sólo en el análisis semántico del corpus basado en los términos con una mayor carga de polaridad, o del resumen de las valoraciones o las tendencias de neuromarketing, sino también en aplicaciones como el etiquetado subjetivo de sitios web o de portales sintácticos y semánticos, por mencionar algunas. ABSTRACT This thesis presents an approach to create what we have called a Unified Sentiment Lexicon (USL). This approach aims at aligning, unifying, and expanding the set of sentiment lexicons which are available on the web in order to increase their robustness of coverage. One problem related to the task of the automatic unification of different scores of sentiment lexicons is that there are multiple lexical entries for which the classification of positive, negative, or neutral P, N, Z depends on the unit of measurement used in the annotation methodology of the source sentiment lexicon. Our USL approach computes the unified strength of polarity of each lexical entry based on the Pearson correlation coefficient which measures how correlated lexical entries are with a value between 1 and - 1 , where 1 indicates that the lexical entries are perfectly correlated, 0 indicates no correlation, and -1 means they are perfectly inversely correlated and so is the UnifiedMetrics procedure for CPU and GPU, respectively. Another problem is the high processing time required for computing all the lexical entries in the unification task. Thus, the USL approach computes a subset of lexical entries in each of the 1344 GPU cores and uses parallel processing in order to unify 155,802 lexical entries. The results of the analysis conducted using the USL approach show that the USL has 95,430 lexical entries, out of which there are 35,201 considered to be positive, 22,029 negative, and 38,200 neutral. Finally, the runtime was 2.505 seconds for 95,430 lexical entries; this allows a reduction of the time computing for the UnifiedMetrics by 3 times with respect to the sequential implementation. A key contribution of this work is that we preserve the use of a unified sentiment lexicon for all tasks. Such lexicon is used to define resources and resource-related properties that can be verified based on the results of the analysis and is powerful, general and extensible enough to express a large class of interesting properties. Some applications of this work include merging, aligning, pruning and extending the current sentiment lexicons.

Fission yeast orb6, a ser/thr protein kinase related to mammalian rho kinase and myotonic dystrophy kinase, is required for maintenance of cell polarity and coordinates cell morphogenesis with the cell cycle

The molecular mechanisms that coordinate cell morphogenesis with the cell cycle remain largely unknown. We have investigated this process in fission yeast where changes in polarized cell growth are coupled with cell cycle progression. The orb6 gene is required during interphase to maintain cell polarity and encodes a serine/threonine protein kinase, belonging to the myotonic dystrophy kinase/cot1/warts family. A decrease in Orb6 protein levels leads to loss of polarized cell shape and to mitotic advance, whereas an increase in Orb6 levels maintains polarized growth and delays mitosis by affecting the p34cdc2 mitotic kinase. Thus the Orb6 protein kinase coordinates maintenance of cell polarity during interphase with the onset of mitosis. orb6 interacts genetically with orb2, which encodes the Pak1/Shk1 protein kinase, a component of the Ras1 and Cdc42-dependent signaling pathway. Our results suggest that Orb6 may act downstream of Pak1/Shk1, forming part of a pathway coordinating cell morphogenesis with progression through the cell cycle.

Identification of Novel Temperature-sensitive Lethal Alleles in Essential β-Tubulin and Nonessential α2-Tubulin Genes as Fission Yeast Polarity Mutants

We have screened for temperature-sensitive (ts) fission yeast mutants with altered polarity (alp1–15). Genetic analysis indicates that alp2 is allelic to atb2 (one of two α-tubulin genes) and alp12 to nda3 (the single β-tubulin gene). atb2+ is nonessential, and the ts atb2 mutations we have isolated are dominant as expected. We sequenced two alleles of ts atb2 and one allele of ts nda3. In the ts atb2 mutants, the mutated residues (G246D and C356Y) are found at the longitudinal interface between α/β-heterodimers, whereas in ts nda3 the mutated residue (Y422H) is situated in the domain located on the outer surface of the microtubule. The ts nda3 mutant is highly sensitive to altered gene dosage of atb2+; overexpression of atb2+ lowers the restrictive temperature, and, conversely, deletion rescues ts. Phenotypic analysis shows that contrary to undergoing mitotic arrest with high viability via the spindle assembly checkpoint as expected, ts nda3 mutants execute cytokinesis and septation and lose viability. Therefore, it appears that the ts nda3 mutant becomes temperature lethal because of irreversible progression through the cell cycle in the absence of activating the spindle assembly checkpoint pathway.

The Rho GTPase Rho3 Has a Direct Role in Exocytosis That Is Distinct from Its Role in Actin Polarity

Budding yeast grow asymmetrically by the polarized delivery of proteins and lipids to specific sites on the plasma membrane. This requires the coordinated polarization of the actin cytoskeleton and the secretory apparatus. We identified Rho3 on the basis of its genetic interactions with several late-acting secretory genes. Mutational analysis of the Rho3 effector domain reveals three distinct functions in cell polarity: regulation of actin polarity, transport of exocytic vesicles from the mother cell to the bud, and docking and fusion of vesicles with the plasma membrane. We provide evidence that the vesicle delivery function of Rho3 is mediated by the unconventional myosin Myo2 and that the docking and fusion function is mediated by the exocyst component Exo70. These data suggest that Rho3 acts as a key regulator of cell polarity and exocytosis, coordinating several distinct events for delivery of proteins to specific sites on the cell surface.

Polarity of human replication protein A binding to DNA

Replication protein A (RPA), the nuclear single-stranded DNA binding protein is involved in DNA replication, nucleotide excision repair (NER) and homologous recombination. It is a stable heterotrimer consisting of subunits with molecular masses of 70, 32 and 14 kDa (p70, p32 and p14, respectively). Gapped DNA structures are common intermediates during DNA replication and NER. To analyze the interaction of RPA and its subunits with gapped DNA we designed structures containing 9 and 30 nucleotide gaps with a photoreactive arylazido group at the 3′-end of the upstream oligonucleotide or at the 5′-end of the downstream oligonucleotide. UV crosslinking and subsequent analysis showed that the p70 subunit mainly interacts with the 5′-end of DNA irrespective of DNA structure, while the subunit orientation towards the 3′-end of DNA in the gap structures strongly depends on the gap size. The results are compared with the data obtained previously with the primer–template systems containing 5′- or 3′-protruding DNA strands. Our results suggest a model of polar RPA binding to the gapped DNA.

Integrin-mediated Adhesion Regulates Cell Polarity and Membrane Protrusion through the Rho Family of GTPasesV⃞

Integrin-mediated adhesion is a critical regulator of cell migration. Here we demonstrate that integrin-mediated adhesion to high fibronectin concentrations induces a stop signal for cell migration by inhibiting cell polarization and protrusion. On fibronectin, the stop signal is generated through α5β1 integrin-mediated signaling to the Rho family of GTPases. Specifically, Cdc42 and Rac1 activation exhibits a biphasic dependence on fibronectin concentration that parallels optimum cell polarization and protrusion. In contrast, RhoA activity increases with increasing substratum concentration. We find that cross talk between Cdc42 and Rac1 is required for substratum-stimulated protrusion, whereas RhoA activity is inhibitory. We also show that Cdc42 activity is inhibited by Rac1 activation, suggesting that Rac1 activity may down-regulate Cdc42 activity and promote the formation of stabilized rather than transient protrusion. Furthermore, expression of RhoA down-regulates Cdc42 and Rac1 activity, providing a mechanism whereby RhoA may inhibit cell polarization and protrusion. These findings implicate adhesion-dependent signaling as a mechanism to stop cell migration by regulating cell polarity and protrusion via the Rho family of GTPases.

Fission Yeast Aip3p (spAip3p) Is Required for an Alternative Actin-directed Polarity Program

Aip3p is an actin-interacting protein that regulates cell polarity in budding yeast. The Schizosaccharomyces pombe-sequencing project recently led to the identification of a homologue of Aip3p that we have named spAip3p. Our results confirm that spAip3p is a true functional homologue of Aip3p. When expressed in budding yeast, spAip3p localizes similarly to Aip3p during the cell cycle and complements the cell polarity defects of an aip3Δ strain. Two-hybrid analysis shows that spAip3p interacts with actin similarly to Aip3p. In fission yeast, spAip3p localizes to both cell ends during interphase and later organizes into two rings at the site of cytokinesis. spAip3p localization to cell ends is dependent on microtubule cytoskeleton, its localization to the cell middle is dependent on actin cytoskeleton, and both patterns of localization require an operative secretory pathway. Overexpression of spAip3p disrupts the actin cytoskeleton and cell polarity, leading to morphologically aberrant cells. Fission yeast, which normally rely on the microtubule cytoskeleton to establish their polarity axis, can use the actin cytoskeleton in the absence of microtubule function to establish a new polarity axis, leading to the formation of branched cells. spAip3p localizes to, and is required for, branch formation, confirming its role in actin-directed polarized cell growth in both Schizosaccharomyces pombe and Saccharomyces cerevisiae.

The physiological significance of β-actin mRNA localization in determining cell polarity and directional motility

β-actin mRNA is localized near the leading edge in several cell types, where actin polymerization is actively promoting forward protrusion. The localization of the β-actin mRNA near the leading edge is facilitated by a short sequence in the 3′ untranslated region, the “zip code.” Localization of the mRNA at this region is important physiologically. Treatment of chicken embryo fibroblasts with antisense oligonucleotides complementary to the localization sequence (zip code) in the 3′ untranslated region leads to delocalization of β-actin mRNA, alteration of cell phenotype, and a decrease in cell motility. To determine the components of this process responsible for the change in cell behavior after β-actin mRNA delocalization, the Dynamic Image Analysis System was used to quantify movement of cells in the presence of sense and antisense oligonucleotides to the zip code. It was found that net path length and average speed of antisense-treated cells were significantly lower than in sense-treated cells. Total path length and the velocity of protrusion of antisense-treated cells were not affected compared with those of control cells. These results suggest that a decrease in persistence of direction of movement and not in velocity results from treatment of cells with zip code-directed antisense oligonucleotides. To test this, direct analysis of directionality was performed on antisense-treated cells and showed a decrease in directionality (net path/total path) and persistence of movement. Less directional movement of antisense-treated cells correlated with a unpolarized and discontinuous distribution of free barbed ends of actin filaments and of β-actin protein. These results indicate that delocalization of β-actin mRNA results in delocalization of nucleation sites and β-actin protein from the leading edge followed by loss of cell polarity and directional movement.

Polarity and permeation profiles in lipid membranes

The isotropic 14N-hyperfine coupling constant, a\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} \begin{equation*}{\mathrm{_{o}^{N}}}\end{equation*}\end{document}, of nitroxide spin labels is dependent on the local environmental polarity. The dependence of a\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} \begin{equation*}{\mathrm{_{o}^{N}}}\end{equation*}\end{document} in fluid phospholipid bilayer membranes on the C-atom position, n, of the nitroxide in the sn-2 chain of a spin-labeled diacyl glycerophospholipid therefore determines the transmembrane polarity profile. The polarity variation in phospholipid membranes, with and without equimolar cholesterol, is characterized by a sigmoidal, trough-like profile of the form {1 + exp [(n − no)/λ]}−1, where n = no is the point of maximum gradient, or polarity midpoint, beyond which the free energy of permeation decreases linearly with n, on a characteristic length-scale, λ. Integration over this profile yields a corresponding expression for the permeability barrier to polar solutes. For fluid membranes without cholesterol, no ≈ 8 and λ ≈ 0.5–1 CH2 units, and the permeability barrier introduces an additional diffusive resistance that is equivalent to increasing the effective membrane thickness by 35–80%, depending on the lipid. For membranes containing equimolar cholesterol, no ≈ 9–10, and the total change in polarity is greater than for membranes without cholesterol, increasing the permeability barrier by a factor of 2, whereas the decay length remains similar. The permeation of oxygen into fluid lipid membranes (determined by spin-label relaxation enhancements) displays a profile similar to that of the transmembrane polarity but of opposite sense. For fluid membranes without cholesterol no ≈ 8 and λ ≈ 1 CH2 units, also for oxygen. The permeation profile for polar paramagnetic ion complexes is closer to a single exponential decay, i.e., no lies outside the acyl-chain region of the membrane. These results are relevant not only to the permeation of water and polar solutes into membranes and their permeabilities, but also to depth determinations of site-specifically spin-labeled protein residues by using paramagnetic relaxation agents.

Ras Regulates the Polarity of the Yeast Actin Cytoskeleton through the Stress Response Pathway

Polarized growth in yeast requires cooperation between the polarized actin cytoskeleton and delivery of post-Golgi secretory vesicles. We have previously reported that loss of the major tropomyosin isoform, Tpm1p, results in cells sensitive to perturbations in cell polarity. To identify components that bridge these processes, we sought mutations with both a conditional defect in secretion and a partial defect in polarity. Thus, we set up a genetic screen for mutations that conferred a conditional growth defect, showed synthetic lethality with tpm1Δ, and simultaneously became denser at the restrictive temperature, a hallmark of secretion-defective cells. Of the 10 complementation groups recovered, the group with the largest number of independent isolates was functionally null alleles of RAS2. Consistent with this, ras2Δ and tpm1Δ are synthetically lethal at 35°C. We show that ras2Δ confers temperature-sensitive growth and temperature-dependent depolarization of the actin cytoskeleton. Furthermore, we show that at elevated temperatures ras2Δ cells are partially defective in endocytosis and show a delocalization of two key polarity markers, Myo2p and Cdc42p. However, the conditional enhanced density phenotype of ras2Δ cells is not a defect in secretion. All the phenotypes of ras2Δ cells can be fully suppressed by expression of yeast RAS1 or RAS2 genes, human Ha-ras, or the double disruption of the stress response genes msn2Δmsn4Δ. Although the best characterized pathway of Ras function in yeast involves activation of the cAMP-dependent protein kinase A pathway, activation of the protein kinase A pathway does not fully suppress the actin polarity defects, suggesting that there is an additional pathway from Ras2p to Msn2/4p. Thus, Ras2p regulates cytoskeletal polarity in yeast under conditions of mild temperature stress through the stress response pathway.

Polarity of microtubule assemblies during neuronal cell migration.

The active migration of neurons from their sites of origin to their final destinations requires the unidirectional translocation of the nuclei and somatic cytoplasm within the growing leading processes. To explore the cellular machinery underlying this translocation, we determined the polarity of microtubules situated within the leading and trailing processes of migrating cerebellar granule cells in situ. Our analysis reveals that the newly assembled positive ends of the microtubules in the leading process uniformly face the growing tip, while their disintegrating negative ends face the nucleus. In the trailing process, by contrast, microtubule arrays are of mixed polarity. We suggest that the dynamics of slow polymerization in combination with fast disintegration of oriented microtubules create "push" and "pull" forces that contribute to the piston-like saltatory displacement of the nucleus and cytoplasm within the membrane cylinder of the leading process of the migrating neuron.

Independent determination of symmetry and polarity in the Drosophila eye.

In each facet of the Drosophila compound eye, a cluster of photoreceptor cells assumes an asymmetric trapezoidal pattern. These clusters have opposite orientations above and below an equator, showing global dorsoventral mirror symmetry. However, in the mutant spiny legs, the polarization of each cluster appears to be random, so that no equator is evident. The apparent lack of an equator suggests that spiny legs+ may be involved in the establishment of global dorsoventral identity that might be essential for proper polarization of the photoreceptor clusters. Alternatively, a global dorsoventral pattern could be present, but spiny legs+ may be required for local polarization of individual clusters. Using an enhancer trap strain in which white+ gene expression is restricted to the dorsal field, we show that white+ expression in spiny legs correctly respects dorsoventral position even in facets with inappropriate polarizations; the dorsoventral boundary is indeed present, whereas the mechanism for polarization is perturbed. It is suggested that the boundary is established before the action of spiny legs+ by an independent mechanism.