Machine learning for WebSensors: algorithms and applications

• Grant number: 14/08996-0
• Support Opportunities: Regular Research Grants
• Start date: August 01, 2014
• End date: July 31, 2016
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Solange Oliveira Rezende
• Grantee: Solange Oliveira Rezende
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated researchers: Bruno Magalhães Nogueira ; Gustavo Enrique de Almeida Prado Alves Batista ; Ricardo Marcondes Marcacini ; Veronica Oliveira de Carvalho

Abstract

The popularization of textual content published in web platforms has motivated the development of methods for automatic knowledge extraction from texts. Particularly, a new range of applications and studies have been proposed to use the web as a powerful "social sensor". This allows to identify and monitor events published in news portals and social networks as epidemics detection, sentiment analysis, and political and economic indicators. Currently, the construction of a websensor is a complex task, since it depends of domain specialists to define the parameters of the sensors, i.e., search queries, filters and monitoring textual content from web. Moreover, for some problems there is no comprehension about the phenomenons to monitor, which limits the application of websensors. In this research project we investigate the use of machine learning methods to support the building of websensors. The basic idea is to use a sample of textual document from a problem and apply semi/non supervised learning methods to extract patterns from texts and support the generation of websensors. Thus, we hope to reduce the dependency of specialist domains to define parameters for websensor. Besides, each websensor represents a phenomenon related to a problem and it can be monitored during the time to be used as support to decision making. (AU)

Scientific publications

(8)

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

MARCACINI, RICARDO MARCONDES; ROSSI, RAFAEL GERALDELI; MATSUNO, IVONE PENQUE; REZENDE, SOLANGE OLIVEIRA. Cross-domain aspect extraction for sentiment analysis: A transductive learning approach. DECISION SUPPORT SYSTEMS, v. 114, p. 70-80, OCT 2018. (14/08996-0)

ROSSI, RAFAEL GERALDELI; LOPES, ALNEU DE ANDRADE; REZENDE, SOLANGE OLIVEIRA. Using bipartite heterogeneous networks to speed up inductive semi-supervised learning and improve automatic text categorization. KNOWLEDGE-BASED SYSTEMS, v. 132, p. 94-118, SEP 15 2017. (11/12823-6, 14/08996-0, 15/14228-9)

MARCACINI, RICARDO M.; CARNEVALI, JULIO C.; DOMINGOS, JOAO; IEEE. On Combining Websensors and DTW Distance for kNN Time Series Forecasting. 2016 23RD INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), v. N/A, p. 5-pg., 2016-01-01. (14/08996-0)

CORREA, GERALDO N.; MARCACINI, RICARDO M.; HRUSCHKA, EDUARDO R.; REZENDE, SOLANGE O.. Interactive textual feature selection for consensus clustering. PATTERN RECOGNITION LETTERS, v. 52, p. 25-31, JAN 15 2015. (14/08996-0)

MANZATO, MARCELO G.; DOMINGUES, MARCOS A.; FORTES, ARTHUR C.; SUNDERMANN, CAMILA V.; D'ADDIO, RAFAEL M.; CONRADO, MERLEY S.; REZENDE, SOLANGE O.; PIMENTEL, MARIA G. C.. Mining unstructured content for recommender systems: an ensemble approach. INFORMATION RETRIEVAL JOURNAL, v. 19, n. 4, p. 378-415, AUG 2016. (13/22547-1, 13/10756-5, 12/13830-9, 14/08996-0, 13/16039-3)

SOUZA, VINICIUS M. A.; ROSSI, RAFAEL G.; BATISTA, GUSTAVO E. A. P. A.; REZENDE, SOLANGE O.. Unsupervised active learning techniques for labeling training sets: An experimental evaluation on sequential data. Intelligent Data Analysis, v. 21, n. 5, p. 1061+, 2017. (14/08996-0, 11/12823-6, 11/17698-5)

ROSSI, RAFAEL GERALDELI; LOPES, ALNEU DE ANDRADE; REZENDE, SOLANGE OLIVEIRA. Optimization and label propagation in bipartite heterogeneous networks to improve transductive classification of texts. INFORMATION PROCESSING & MANAGEMENT, v. 52, n. 2, p. 217-257, MAR 2016. (11/12823-6, 11/22749-8, 14/08996-0)

SINOARA, ROBERTA A.; ROSSI, RAFAEL G.; REZENDE, SOLANGE O.; IEEE. Semantic Role-based Representations in Text Classification. 2016 23RD INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), v. N/A, p. 6-pg., 2016-01-01. (16/07620-2, 14/08996-0, 11/12823-6, 13/14757-6)