Understanding temporal aspects in document classification

Abstract

Due to the increasing amount of information present on the Web, Automatic Document Classification (ADC) has become an important research topic. ADC usually follows a standard supervised learning strategy, where we first build a model using pre-classified documents and then use it to classify new unseen documents. One major challenge for ADC in many scenarios is that the characteristics of the documents and the classes to which they belong may change over time. However, most of the current techniques for ADC are applied without taking into account the temporal evolution of the collection of documents.In this work, we perform a detailed study of the temporal evolution in the ADC, introducing an analysis methodology. We discuss that temporal evolution may be explained by three factors: 1) class distribution; 2) term distribution; and 3) class similarity. We employ metrics and experimental strategies capable of isolating each of these factors in order to analyze them separately, using two very different document collections: the ACM Digital Library and the Medline medical collections. Moreover, we present some preliminary results of potential gains that could be obtained by varying the training set to find the ideal size that minimizes the time effects. We show that by using just 69% of the ACM database, we are able to have an accuracy of 89.76%, and with only 25% of the Medline, an accuracy of 87.57%, which means gains of up to 20% in accuracy with much smaller training sets.

Machine learning approach to recognize subject based sentiment values of reviews

“…Salles et al [8] have discussed the impact that temporal effects may have on Automatic Document Classification and how to minimize such impact. Previous researches have shown that temporal effects, such as the variation of the strength of term-class relationship over time, have a significant impact on Automatic Document Classification [9]. To deal with the effect of term-class relationship over time Salles et al [8] have introduced a temporal weighting function and have extended the classification algorithms to incorporate the temporal weighting function.…”

Effects of Evolutionary Linguistics in Text Classification

“…Mourao et al present an empirical analysis of temporal data on news classification [11]. The impact of empirical methods in information extraction is described by Cardie in [3].…”

Agents Behavior and Preferences Characterization in Civilization IV

“…A work that met these requirements is [Mourão et al 2008] where the authors characterized textual documents evolving through the time. They presented evidences of this evolution as metrics and experiments that confirmed it.…”

“…They presented evidences of this evolution as metrics and experiments that confirmed it. Another work in this same subject was [Salles 2011] that redid the analysis presented in [Mourão et al 2008] for an additional dataset. He also applied factorial projects techniques [Jain 1991] to identify the impact of variations in the classification algorithms.…”