Search

In this paper, we present an investigation of authorship identification on personal blogs or diaries, which are different from other types of text such as essays, emails, or articles based on the text properties. The investigation utilizes couple of intuitive feature sets and studies various parameters that affect the identification performance. Many studies manipulated the problem of authorship identification in manually collected corpora, but only few utilized real data from existing blogs. The complexity of the language model in personal blogs is motivating to identify the correspondent author. The main contribution of this work is at least three folds. Firstly, we utilize the LIWC and MRC feature sets together, which have been developed with Psychology background, for the first time for authorship identification on personal blogs. Secondly, we analyze the effect of various parameters, and feature sets, on the identification performance. This includes the number of authors in the data corpus, the post size or the word count, and the number of posts for each author. Finally, we study applying authorship identification over a limited set of users that have a common personality attributes. This analysis is motivated by the lack of standard or solid recommendations in literature for such task, especially in the domain of personal blogs. The results and evaluation show that the utilized features are compact while their performance is highly comparable with other larger feature sets. The analysis also confirmed the most effective parameters, their ranges in the data corpus, and the usefulness of the common users classifier in improving the performance, for the author identification task. [ABSTRACT FROM AUTHOR]

Our work explores the use of several text categorization techniques for classification of manufacturing quality defect and service shop data sets into fixed categories. Although our work was in the area of manufacturing quality the technique is applicable to free form, short text summaries of data that may be stored in a database, file, or document. We refer to these types of text as "technical text passages". These summaries may not follow standard grammar conventions; they commonly contain abbreviations, technical phrases, misspelled words and industry specific acronyms. Typical types of text to be classified include aircraft engine repair shop findings, industrial manufacturing quality problems and corrective actions, and standardization of attributes in a bill-of-materials. In this paper, we will present our results in using machine learning and rule based algorithms to categorize text. Our results show that the rules based approach is as good as several machine learning approaches. For example, using Support Vector Machine algorithms we were able to achieve 82% accuracy on validation set, using 1,645 training samples and 823 validation samples. Each category had 50 or more samples. Using rule-based approach we were able to achieve 80% accuracy. [ABSTRACT FROM AUTHOR]

This paper presents an energy back-propagation algorithm (EBP). Learning and convergence processes of the standard backpropagation algorithm (SBP) are based on the energy function. The energy function is used with the convergence process to extract the nearest image for the unknown tested image. The EBP algorithm shows considerably better performance in terms of time of learning, time of convergence, and size of input image compared to the SBP algorithm. [ABSTRACT FROM AUTHOR]