An end-to-end deep generative approach with meta-learning optimization for zero-shot object classification

Zero-shot object classification aims to recognize the object of unseen classes whose supervised data are unavailable in the training stage. Recent zero-shot learning (ZSL) methods usually propose to generate new supervised data for unseen classes by designing various deep generative networks. In this paper, we propose an end-to-end deep generative ZSL approach that trains the data generation module and object classification module jointly, rather than separately as in the majority of existing generation-based ZSL methods. Due to the ZSL assumption that unseen data are unavailable in the training stage, the distribution of generated unseen data will shift to the distribution of seen data, and subsequently causes the projection domain shift problem. Therefore, we further design a novel meta-learning optimization model to improve the proposed generation-based ZSL approach, where the parameters initialization and the parameters update algorithm are meta-learned to assist model convergence. We evaluate the proposed approach on five standard ZSL datasets. The average accuracy increased by the proposed jointly training strategy is 2.7% and 23.0% for the standard ZSL task and generalized ZSL task respectively, and the meta-learning optimization further improves the accuracy by 5.0% and 2.1% on two ZSL tasks respectively. Experimental results demonstrate that the proposed approach has significant superiority in various ZSL tasks.     

Beyond vector space model for hierarchical Arabic text classification: A Markov chain approach

The vector space model (VSM) is a textual representation method that is widely used in documents classification. However, it remains to be a space-challenging problem. One attempt to alleviate the space problem is by using dimensionality reduction techniques, however, such techniques have deficiencies such as losing some important information. In this paper, we propose a novel text classification method that neither uses VSM nor dimensionality reduction techniques. The proposed method is a space efficient method that utilizes the first order Markov model for hierarchical Arabic text classification. For each category and sub-category, a Markov chain model is prepared based on the neighboring characters sequences. The prepared models are then used for scoring documents for classification purposes. For evaluation, we used a hierarchical Arabic text data collection that contains 11,191 documents that belong to eight topics distributed into 3-levels. The experimental results show that the Markov chains based method significantly outperforms the baseline system that employs the latent semantic indexing (LSI) method. That is, the proposed method enhances the F1-measure by 3.47%. The novelty of this work lies on the idea of decomposing words into sequences of characters, which found to be a promising approach in terms of space and accuracy. Based on our best knowledge, this is the first attempt to conduct research for hierarchical Arabic text classification with such relatively large data collection.     

A text classification approach to detect psychological stress combining a lexicon-based feature framework with distributional representations

Nowadays, stress has become a growing problem for society due to its high impact on individuals but also on health care systems and companies. In order to overcome this problem, early detection of stress is a key factor. Previous studies have shown the effectiveness of text analysis in the detection of sentiment, emotion, and mental illness. However, existing solutions for stress detection from text are focused on a specific corpus. There is still a lack of well-validated methods that provide good results in different datasets. We aim to advance state of the art by proposing a method to detect stress in textual data and evaluating it using multiple public English datasets. The proposed approach combines lexicon-based features with distributional representations to enhance classification performance. To help organize features for stress detection in text, we propose a lexicon-based feature framework that exploits affective, syntactic, social, and topic-related features. Also, three different word embedding techniques are studied for exploiting distributional representation. Our approach has been implemented with three machine learning models that have been evaluated in terms of performance through several experiments. This evaluation has been conducted using three public English datasets and provides a baseline for other researchers. The obtained results identify the combination of FastText embeddings with a selection of lexicon-based features as the best-performing model, achieving F-scores above 80%.     

A machine learning based framework to identify unseen classes in open-world text classification

Classical supervised machine learning (ML) follows the assumptions of closed-world learning. However, this assumption does not work in an open-world dynamic environment. Therefore, the automated systems must be able to discover and identify unseen instances. Open-world ML can deal with unseen instances and classes through a two-step process: (1) discover and classify unseen instances and (2) identify novel classes discovered in step (1). Most existing research on open-world machine learning (OWML) only focuses on step 1. However, performing step 2 is required to build intelligent systems. The proposed framework comprises three different but interconnected modules that discover and identify unseen classes. Our in-depth performance evaluation establishes that the proposed framework improves open accuracy by up to 8% compared to the state-of-the-art models.     

A new approach to assessing the technological complexity of different categories of R&D (with examples from Singapore)

This paper attempts to develop a set of generic questions the answers to which allow any given R&D activity to be correctly classified in one or another conventional R&D box—basic, applied, development, etc. Such systematic pigeon-holing allows latecomer countries to gauge the distance of their R&D activity from the world frontier, and how policies may be designed to overcome the hurdle that divides applied research and developmental research.