Artículos, conferencias, monografías

Permanent URI for this collection

https://riunet7pre.upv.es/handle/10251/3829

Esta colección está formada por artículos, conferencias, comunicaciones y otras publicaciones elaborados por miembros de la Universitat Politècnica de València.

En el caso de documentos publicados, antes de incluirlos en RiuNet, el autor debe saber las condiciones en que ha cedido los derechos de explotación de su obra al editor. Aunque la mayoría de los editores permiten el archivo del pre-print (el borrador del texto a publicar antes de la revisión por pares) se recomienda consultar la base de datos Sherpa / Romeo para conocer la política sobre derechos de autor de las editoriales internacionales y la base de datos Dulcinea para conocer la política sobre derechos de autor de las editoriales españolas. En caso de duda el autor debe ponerse en contacto con la editorial para conocer su política respecto al archivo en repositorios institucionales.

Browse

Now showing 1 - 20 of 349

3D measurements in conventional X-ray imaging with RGB-D sensors
(Elsevier, 2017-04) Albiol Colomer, Francisco; Corbi, Alberto; Albiol Colomer, Alberto; Escuela Técnica Superior de Ingeniería de Telecomunicación; Dpto. de Comunicaciones; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Universitat de València; Ministerio de Economía y Competitividad; Ministerio de Industria, Energía y Turismo; Ministerio de Educación y Ciencia
[EN] A method for deriving 3D internal information in conventional X-ray settings is presented. It is based on the combination of a pair of radiographs from a patient and it avoids the use of X-ray-opaque fiducials and external reference structures. To achieve this goal, we augment an ordinary X-ray device with a consumer RGB-D camera. The patient' s rotation around the craniocaudal axis is tracked relative to this camera thanks to the depth information provided and the application of a modern surface-mapping algorithm. The measured spatial information is then translated to the reference frame of the X-ray imaging system. By using the intrinsic parameters of the diagnostic equipment, epipolar geometry, and X-ray images of the patient at different angles, 3D internal positions can be obtained. Both the RGB-D and Xray instruments are first geometrically calibrated to find their joint spatial transformation. The proposed method is applied to three rotating phantoms. The first two consist of an anthropomorphic head and a torso, which are filled with spherical lead bearings at precise locations. The third one is made of simple foam and has metal needles of several known lengths embedded in it. The results show that it is possible to resolve anatomical positions and lengths with a millimetric level of precision. With the proposed approach, internal 3D reconstructed coordinates and distances can be provided to the physician. It also contributes to reducing the invasiveness of ordinary X-ray environments and can replace other types of clinical explorations that are mainly aimed at measuring or geometrically relating elements that are present inside the patient's body.(C) 2017 IPEM. Published by Elsevier Ltd. All rights reserved.
A Bidirectional Recurrent Neural Language Model for Machine Translation
(Sociedad Española para el Procesamiento del Lenguaje Natural (SEPLN), 2015-09) Peris Abril, Álvaro; Casacuberta Nolla, Francisco; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Generalitat Valenciana
[EN] A language model based in continuous representations of words is presented, which has been applied to a statistical machine translation task. This model is implemented by means of a bidirectional recurrent neural network, which is able to take into account both the past and the future context of a word in order to perform predictions. Due to its high temporal cost at training time, for obtaining relevant training data an instance selection algorithm is used, which aims to capture useful information for translating a test set. Obtained results show that the neural model trained with the selected data outperforms the results obtained by an n-gram language model
A Comparison of Approaches for Measuring Cross-Lingual Similarity of Wikipedia Articles
(Springer Verlag (Germany), 2014) Barrón Cedeño, Luis Alberto; Paramita, Monica Lestari; Clough, Paul; Rosso, Paolo; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Ministerio de Economía y Competitividad; European Commission
Wikipedia has been used as a source of comparable texts for a range of tasks, such as Statistical Machine Translation and CrossLanguage Information Retrieval. Articles written in different languages on the same topic are often connected through inter-language-links. However, the extent to which these articles are similar is highly variable and this may impact on the use of Wikipedia as a comparable resource. In this paper we compare various language-independent methods for measuring cross-lingual similarity: character n-grams, cognateness, word count ratio, and an approach based on outlinks. These approaches are compared against a baseline utilising MT resources. Measures are also compared to human judgements of similarity using a manually created resource containing 700 pairs of Wikipedia articles (in 7 language pairs). Results indicate that a combination of language-independent models (char-ngrams, outlinks and word-count ratio) is highly effective for identifying cross-lingual similarity and performs comparably to language-dependent models (translation and monolingual analysis).
A comparison of Covid-19 early detection between convolutional neural networks and radiologists
(SpringerOpen, 2022-07-28) Albiol Colomer, Alberto; Albiol, Francisco; Paredes Palacios, Roberto; Plasencia-Martínez, Juana María; Blanco Barrio, Ana; García Santos, José M.; Tortajada, Salvador; González Montaño, Victoria M.; Rodríguez Godoy, Clara E.; Fernández Gómez, Saray; Oliver-Garcia, Elena; de la Iglesia Vayá, María; Márquez Pérez, Francisca L.; Rayo Madrid, Juan I.; Escuela Técnica Superior de Ingeniería de Telecomunicación; Dpto. de Sistemas Informáticos y Computación; Dpto. de Comunicaciones; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Instituto de Salud Carlos III; Agència Valenciana de la Innovació; Universitat Politècnica de València; SGS INTERNATIONAL CERTIFICATION SERVICES IBERICA SA; DREUE ELECTRIC, SL; FERMAX ELECTRONICA S.A.U.
[EN] Background The role of chest radiography in COVID-19 disease has changed since the beginning of the pandemic from a diagnostic tool when microbiological resources were scarce to a different one focused on detecting and monitoring COVID-19 lung involvement. Using chest radiographs, early detection of the disease is still helpful in resource-poor environments. However, the sensitivity of a chest radiograph for diagnosing COVID-19 is modest, even for expert radiologists. In this paper, the performance of a deep learning algorithm on the first clinical encounter is evaluated and compared with a group of radiologists with different years of experience. Methods The algorithm uses an ensemble of four deep convolutional networks, Ensemble4Covid, trained to detect COVID-19 on frontal chest radiographs. The algorithm was tested using images from the first clinical encounter of positive and negative cases. Its performance was compared with five radiologists on a smaller test subset of patients. The algorithm's performance was also validated using the public dataset COVIDx. Results Compared to the consensus of five radiologists, the Ensemble4Covid model achieved an AUC of 0.85, whereas the radiologists achieved an AUC of 0.71. Compared with other state-of-the-art models, the performance of a single model of our ensemble achieved nonsignificant differences in the public dataset COVIDx. Conclusion The results show that the use of images from the first clinical encounter significantly drops the detection performance of COVID-19. The performance of our Ensemble4Covid under these challenging conditions is considerably higher compared to a consensus of five radiologists. Artificial intelligence can be used for the fast diagnosis of COVID-19.
A Decade of Shared Tasks in Digital Text Forensics at PAN
(Springer-Verlag, 2019) Potthast, Martin; Rosso, Paolo; Stamatatos, Efstathios; Stein, Benno; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Agencia Estatal de Investigación
[EN] Digital text forensics aims at examining the originality and credibility of information in electronic documents and, in this regard, to extract and analyze information about the authors of these documents. The research field has been substantially developed during the last decade. PAN is a series of shared tasks that started in 2009 and significantly contributed to attract the attention of the research community in well-defined digital text forensics tasks. Several benchmark datasets have been developed to assess the state-of-the-art performance in a wide range of tasks. In this paper, we present the evolution of both the examined tasks and the developed datasets during the last decade. We also briefly introduce the upcoming PAN 2019 shared tasks.
A Deep Source-Context Feature for Lexical Selection in Statistical Machine Translation
(Elsevier, 2016-05) Gupta, Parth Alokkumar; Costa-Jussa, Marta R; Rosso, Paolo; Banchs, Rafael; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Generalitat Valenciana; Ministerio de Economía y Competitividad; European Commission
This paper presents a methodology to address lexical disambiguation in a standard phrase-based statistical machine translation system. Similarity among source contexts is used to select appropriate translation units. The information is introduced as a novel feature of the phrase-based model and it is used to select the translation units extracted from the training sentence more similar to the sentence to translate. The similarity is computed through a deep autoencoder representation, which allows to obtain effective lowdimensional embedding of data and statistically significant BLEU score improvements on two different tasks (English-to-Spanish and English-to-Hindi). © 2016 Elsevier B.V. All rights reserved.
A hybrid approach for transliterated word-level language identification: CRF with post processing heuristics
(ACM, 2014-12-05) Banerjee, Somnath; Kuila, Alapan; Roy, Aniruddha; Naskar, Sudip Kumar; Rosso, Paolo; Bandyopadhyay, Sivaji; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Department of Electronics and Information Technology, Ministry of Communications and Information Technology, India; European Commission; Universitat de València; Ministerio de Economía y Competitividad
[EN] In this paper, we describe a hybrid approach for word-level language (WLL) identification of Bangla words written in Roman script and mixed with English words as part of our participation in the shared task on transliterated search at Forum for Information Retrieval Evaluation (FIRE) in 2014. A CRF based machine learning model and post-processing heuristics are employed for the WLL identification task. In addition to language identification, two transliteration systems were built to transliterate detected Bangla words written in Roman script into native Bangla script. The system demonstrated an overall token level language identification accuracy of 0.905. The token level Bangla and English language identification F-scores are 0.899, 0.920 respectively. The two transliteration systems achieved accuracies of 0.062 and 0.037. The word-level language identification system presented in this paper resulted in the best scores across almost all metrics among all the participating systems for the Bangla-English language pair.
A Hybrid Real-Time Vision-Based Person Detection Method
(Instituto de Telecomunicaciones y Aplicaciones Multimedia (ITEAM)", 2011) Oliver Moll, Javier; Albiol Colomer, Alberto; Morillas Gómez, Samuel; Peris Fajarnes, Guillermo; Escuela Técnica Superior de Ingeniería de Telecomunicación; Dpto. de Matemática Aplicada; Dpto. de Comunicaciones; Dpto. de Ingeniería Gráfica; Escuela Técnica Superior de Ingeniería Geodésica, Cartográfica y Topográfica; Instituto Universitario de Matemática Pura y Aplicada; Centro de Investigación en Tecnologías Gráficas; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; European Commission; Generalitat Valenciana
[EN] In this paper, we introduce a hybrid real-time method for vision-based pedestrian detection made up by the sequential combination of two basic methods applied in a coarse to fine fashion. The proposed method aims to achieve an improved balance between detection accuracy and computational load by taking advantage of the strengths of these basic techniques. Haar-like features combined with Boosting techniques, which have been demonstrated to provide rapid but not accurate enough results in human detection, are used in the first stage to provide a preliminary candidate selection in the scene. Then, feature extraction and classification methods, which present high accuracy rates at expenses of a higher computational cost, are applied over boosting candidates providing the final prediction. Experimental results show that the proposed method performs effectively and efficiently, which supports its suitability for real applications.
A Knowledge-Based Weighted KNN for Detecting Irony in Twitter
(Springer-Verlag, 2018) Hernandez-Farias, Delia Irazu; Montes Gomez, Manuel; Escalante, Hugo; Rosso, Paolo; Patti, Viviana; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Ministerio de Economía y Empresa; Università degli Studi di Torino; Ministerio de Economía y Competitividad; Consejo Nacional de Ciencia y Tecnología, México
[EN] In this work, we propose a variant of a well-known instancebased algorithm: WKNN. Our idea is to exploit task-dependent features in order to calculate the weight of the instances according to a novel paradigm: the Textual Attraction Force, that serves to quantify the degree of relatedness between documents. The proposed method was applied to a challenging text classification task: irony detection. We experimented with corpora in the state of the art. The obtained results show that despite being a simple approach, our method is competitive with respect to more advanced techniques.
A Low Dimensionality Representation for Language Variety Identification
(Springer-Verlag, 2018) Rangel-Pardo, Francisco Manuel; Franco-Salvador, Marc; Rosso, Paolo; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Ministerio de Economía y Empresa; Generalitat Valenciana
[EN] Language variety identification aims at labelling texts in a native language (e.g. Spanish, Portuguese, English) with its specific variation (e.g. Argentina, Chile, Mexico, Peru, Spain; Brazil, Portugal; UK, US). In this work we propose a low dimensionality representation (LDR) to address this task with five different varieties of Spanish: Argentina, Chile, Mexico, Peru and Spain. We compare our LDR method with common state-of-the-art representations and show an increase in accuracy of ~35%. Furthermore, we compare LDR with two reference distributed representation models. Experimental results show competitive performance while dramatically reducing the dimensionality¿and increasing the big data suitability¿to only 6 features per variety. Additionally, we analyse the behaviour of the employed machine learning algorithms and the most discriminating features. Finally, we employ an alternative dataset to test the robustness of our low dimensionality representation with another set of similar languages.
A multidimensional approach for detecting irony in Twitter
(Springer Netherlands, 2013-03) Reyes Pérez, Antonio; Rosso, Paolo; Veale, Tony; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; European Commission; Ministerio de Ciencia e Innovación; Consejo Nacional de Ciencia y Tecnología, México
Irony is a pervasive aspect of many online texts, one made all the more difficult by the absence of face-to-face contact and vocal intonation. As our media increasingly become more social, the problem of irony detection will become even more pressing. We describe here a set of textual features for recognizing irony at a linguistic level, especially in short texts created via social media such as Twitter postings or ¿tweets¿. Our experiments concern four freely available data sets that were retrieved from Twitter using content words (e.g. ¿Toyota¿) and user-generated tags (e.g. ¿#irony¿). We construct a new model of irony detection that is assessed along two dimensions: representativeness and relevance. Initial results are largely positive, and provide valuable insights into the figurative issues facing tasks such as sentiment analysis, assessment of online reputations, or decision making.
A Multilevel Approach to Sentiment Analysis of Figurative Language in Twitter
(Springer-Verlag, 2018) Gopal Patra, Braja; Mazumda, Soumadeep; Das, Dipankar; Rosso, Paolo; Bandyopadhyay, Sivaji; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Ministerio de Economía y Empresa; Generalitat Valenciana
[EN] Commendable amount of work has been attempted in the field of Sentiment Analysis or Opinion Mining from natural language texts and Twitter texts. One of the main goals in such tasks is to assign polarities (positive or negative) to a piece of text. But, at the same time, one of the important as well as difficult issues is how to assign the degree of positivity or negativity to certain texts. The answer becomes more complex when we perform a similar task on figurative language texts collected from Twitter. Figurative language devices such as irony and sarcasm contain an intentional secondary or extended meaning hidden within the expressions. In this paper we present a novel approach to identify the degree of the sentiment (fine grained in an 11-point scale) for the figurative language texts. We used several semantic features such as sentiment and intensifiers as well as we introduced sentiment abruptness, which measures the variation of sentiment from positive to negative or vice versa. We trained our systems at multiple levels to achieve the maximum cosine similarity of 0.823 and minimum mean square error of 2.170.
A Neural, Interactive-predictive System for Multimodal Sequence to Sequence Tasks
(Association for Computational Linguistics, 2019-08-02) Peris, Álvaro; Casacuberta Nolla, Francisco; Centro Propio de Investigación Pattern Recognition and Human Language Technology; GENERALITAT VALENCIANA; MINISTERIO DE ECONOMIA Y EMPRESA
[EN] We present a demonstration of a neural interactive-predictive system for tackling mul- timodal sequence to sequence tasks. The system generates text predictions to different sequence to sequence tasks: machine translation, image and video captioning. These predictions are revised by a human agent, who introduces corrections in the form of characters. The system reacts to each correction, providing alternative hypotheses, compelling with the feed-back provided by the user. The final objective is to reduce the human effort required during this correction process. This system is implemented following a client server architecture. For accessing the system, we developed a website, which communicates with the neural model, hosted in a local server. From this website, the different tasks can be tackled following the interactive-predictive framework. We opensource all the code developed for building this system.
A new corpus for the evaluation of arabic intrinsic plagiarism detection
(Springer Verlag (Germany), 2013) Bensalem, Imene; Rosso, Paolo; Chikhi, Salim; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Agencia Española de Cooperación Internacional para el Desarrollo; Ministerio de Asuntos Exteriores y Cooperación
The present paper introduces the first corpus for the evaluation of Arabic intrinsic plagiarism detection. The corpus consists of 1024 artificial suspicious documents in which 2833 plagiarism cases have been inserted automatically from source documents
A New Weighted k-Nearest Neighbor Algorithm Based on Newton¿s Gravitational Force
(Springer-Verlag, 2019) Aguilera, Juan; González, Luis C.; Montes-y-Gómez, Manuel; Rosso, Paolo; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Consejo Nacional de Ciencia y Tecnología, México; Ministerio de Economía y Competitividad
[EN] The kNN algorithm has three main advantages that make it appealing to the community: it is easy to understand, it regularly offers competitive performance and its structure can be easily tuning to adapting to the needs of researchers to achieve better results. One of the variations is weighting the instances based on their distance. In this paper we propose a weighting based on the Newton's gravitational force, so that a mass (or relevance) has to be assigned to each instance. We evaluated this idea in the kNN context over 13 benchmark data sets used for binary and multi-class classification experiments. Results in F1 score, statistically validated, suggest that our proposal outperforms the original version of kNN and is statistically competitive with the distance weighted kNN version as well.
A Novel Low Processing Time System for Criminal Activities Detection Applied to Command and Control Citizen Security Centers
(MDPI AG, 2019-12) Suarez-Paez, Julio; Salcedo-Gonzalez, Mayra; Climente Alarcón, Alfonso; Esteve Domingo, Manuel; Gómez Adrian, Jon Ander; Palau Salvador, Carlos Enrique; Pérez Llopis, Israel; Escuela Técnica Superior de Ingeniería de Telecomunicación; Dpto. de Sistemas Informáticos y Computación; Dpto. de Comunicaciones; Escuela Técnica Superior de Ingeniería Informática; Grupo de Sistemas y Aplicaciones de Tiempo Real Distribuido. SATRD; Centro Propio de Investigación Pattern Recognition and Human Language Technology; European Commission
[EN] This paper shows a Novel Low Processing Time System focused on criminal activities detection based on real-time video analysis applied to Command and Control Citizen Security Centers. This system was applied to the detection and classification of criminal events in a real-time video surveillance subsystem in the Command and Control Citizen Security Center of the Colombian National Police. It was developed using a novel application of Deep Learning, specifically a Faster Region-Based Convolutional Network (R-CNN) for the detection of criminal activities treated as "objects" to be detected in real-time video. In order to maximize the system efficiency and reduce the processing time of each video frame, the pretrained CNN (Convolutional Neural Network) model AlexNet was used and the fine training was carried out with a dataset built for this project, formed by objects commonly used in criminal activities such as short firearms and bladed weapons. In addition, the system was trained for street theft detection. The system can generate alarms when detecting street theft, short firearms and bladed weapons, improving situational awareness and facilitating strategic decision making in the Command and Control Citizen Security Center of the Colombian National Police.
A Novel Method of Spatiotemporal Dynamic Geo-Visualization of Criminal Data, Applied to Command and Control Centers for Public Safety
(MDPI AG, 2020-03-10) Salcedo-González, Mayra Liliana; Suarez-Paez, Julio Ernesto; Esteve Domingo, Manuel; Gómez Adrian, Jon Ander; Palau Salvador, Carlos Enrique; Escuela Técnica Superior de Ingeniería de Telecomunicación; Dpto. de Sistemas Informáticos y Computación; Dpto. de Comunicaciones; Escuela Técnica Superior de Ingeniería Informática; Grupo de Sistemas y Aplicaciones de Tiempo Real Distribuido. SATRD; Centro Propio de Investigación Pattern Recognition and Human Language Technology; European Commission
[EN] This article shows a novel geo-visualization method of dynamic spatiotemporal data that allows mobility and concentration of criminal activity to be study. The method was developed using, only and significantly, real data of Santiago de Cali (Colombia), collected by the Colombian National Police (PONAL). This method constitutes a tool that allows criminal influx to be analyzed by concentration, zone, time slot and date. In addition to the field experience of police commanders, it allows patterns of criminal activity to be detected, thereby enabling a better distribution and management of police resources allocated to crime deterrence, prevention and control. Additionally, it may be applied to the concepts of safe city and smart city of the PONAL within the architecture of Command and Control System (C2S) of Command and Control Centers for Public Safety. Furthermore, it contributes to a better situational awareness and improves the future projection, agility, efficiency and decision-making processes of police officers, which are all essential for fulfillment of police missions against crime. Finally, this was developed using an open source software, it can be adapted to any other city, be used with real-time data and be implemented, if necessary, with the geographic software of any other C2S.
A Particle Swarm Optimizer to Cluster Parallel Spanish-English Short-text Corpora
(CEUR Workshop Proceedings, 2011) Ingaramo, Diego Alejandro; Errecalde, Marcelo Luis; Cagnina, Leticia; Rosso, Paolo; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology
Short-texts clustering is currently an important research area because of its applicability to web information retrieval, text summarization and text mining. These texts are often available in different languages and parallel multilingual corpora. Some previous works have demonstrated the effectiveness of a discrete Particle Swarm Optimizer algorithm, named CLUDIPSO, for clustering monolingual corpora containing very short documents. In all the considered cases, CLUDIPSO outperformed different algorithms representative of the state-of-the-art in the area. This paper presents a preliminary study showing the performance of CLUDIPSO on parallel Spanish-English corpora. The idea is to analyze how this bilingual information can be incorporated in the CLUDIPSO algorithm and to what extent this information can improve the clustering results. In order to adapt CLUDIPSO to a bilingual environment, some alternatives are proposed and evaluated. The results were compared considering CLUDIPSO in both environments, bilingual and monolingual. The experimental work shows that bilingual information allows to obtain just comparable results to those obtained with monolingual corpora. More work is required to make an effective use of this kind of information.
A phonetic-based approach to query-by-example spoken term detection
(Springer Verlag (Germany), 2013) Hurtado Oliver, Lluis Felip; Calvo Lance, Marcos; Gómez Adrian, Jon Ander; García Granada, Fernando; Sanchís Arnal, Emilio; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Geodésica, Cartográfica y Topográfica; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Instituto Universitario Valenciano de Investigación en Inteligencia Artificial; Universitat Politècnica de València; Ministerio de Ciencia e Innovación; Ministerio de Educación, Cultura y Deporte
Query-by-Example Spoken Term Detection (QbE-STD) tasks are usually addressed by representing speech signals as a sequence of feature vectors by means of a parametrization step, and then using a pattern matching technique to find the candidate detections. In this paper, we propose a phoneme-based approach in which the acoustic frames are first converted into vectors representing the a posteriori probabilities for every phoneme. This strategy is specially useful when the language of the task is a priori known. Then, we show how this representation can be used for QbE-STD using both a Segmental Dynamic Time Warping algorithm and a graph-based method. The proposed approach has been evaluated with a QbE-STD task in Spanish, and the results show that it can be an adequate strategy for tackling this kind of problems
A resource-light method for cross-lingual semantic textual similarity
(Elsevier, 2018-03-01) Glavas, Goran; Franco-Salvador, Marc; Ponzetto, Simone Paolo; Rosso, Paolo; Dpto. de Sistemas Informáticos y Computación; Escuela Técnica Superior de Ingeniería Informática; Centro Propio de Investigación Pattern Recognition and Human Language Technology; Ministerio de Economía y Competitividad; Deutscher Akademischer Austauschdienst; Ministerium für Wissenschaft, Forschung und Kunst Baden-Württemberg
[EN] Recognizing semantically similar sentences or paragraphs across languages is beneficial for many tasks, ranging from cross-lingual information retrieval and plagiarism detection to machine translation. Recently proposed methods for predicting cross-lingual semantic similarity of short texts, however, make use of tools and resources (e.g., machine translation systems, syntactic parsers or named entity recognition) that for many languages (or language pairs) do not exist. In contrast, we propose an unsupervised and a very resource-light approach for measuring semantic similarity between texts in different languages. To operate in the bilingual (or multilingual) space, we project continuous word vectors (i.e., word embeddings) from one language to the vector space of the other language via the linear translation model. We then align words according to the similarity of their vectors in the bilingual embedding space and investigate different unsupervised measures of semantic similarity exploiting bilingual embeddings and word alignments. Requiring only a limited-size set of word translation pairs between the languages, the proposed approach is applicable to virtually any pair of languages for which there exists a sufficiently large corpus, required to learn monolingual word embeddings. Experimental results on three different datasets for measuring semantic textual similarity show that our simple resource-light approach reaches performance close to that of supervised and resource-intensive methods, displaying stability across different language pairs. Furthermore, we evaluate the proposed method on two extrinsic tasks, namely extraction of parallel sentences from comparable corpora and cross-lingual plagiarism detection, and show that it yields performance comparable to those of complex resource-intensive state-of-the-art models for the respective tasks. (C) 2017 Published by Elsevier B.V.

Browse

Browsing Artículos, conferencias, monografías by UPV Entity "Centro Propio de Investigación Pattern Recognition and Human Language Technology"

Results Per Page

Sort Options