Dwaipayan Roy

Generalized diagonal 2D FLDA for efficient face recognition

This paper presents a novel generalized diagonal two-dimensional Fishers linear discriminant (G-Dia2DFLD) analysis for face representation and recognition. The G-Dia2DFLD method is an extension of the existing DiaFLD method in two aspects. Firstly, the former seeks the maximum class separability by interlacing both the forward and backward diagonals of images simultaneously while the latter seeks optimal projection vectors either from forward or backward diagonal of images. Secondly, the DiaFLD method does not preserve continuity of image regions while generating the diagonal images; resulting partially diagonal images; whereas in G-Dia2DFLD method, this continuity is preserved by generating the diagonal images of the original images. The simulation results on the AT&T and AR databases demonstrate the superiority of the proposed G-Dia2DFLD method over the DiaFLD method and also some existing subspace methods.

Using Word Embeddings for Information Retrieval: How Collection and Term Normalization Choices Affect Performance

Neural word embedding approaches, due to their ability to capture semantic meanings of vocabulary terms, have recently gained attention of the information retrieval (IR) community and have shown promising results in improving ad hoc retrieval performance. It has been observed that these approaches are sensitive to various choices made during the learning of word embeddings and their usage, often leading to poor reproducibility. We study the effect of varying following two parameters, viz., i) the term normalization and ii) the choice of training collection, on ad hoc retrieval performance with word2vec and fastText embeddings. We present quantitative estimates of similarity of word vectors obtained under different settings, and use embeddings based query expansion task to understand the effects of these parameters on IR effectiveness.

An Improved Test Collection and Baselines for Bibliographic Citation Recommendation

The problem of recommending bibliographic citations to an author who is writing an article has been well-studied. However, different researchers have used different datasets to evaluate proposed techniques, and have sometimes reported contradictory findings regarding the relative effectiveness of various approaches. In addition, these datasets are problematic in one way or another (e.g., in terms of size or availability), precluding the possibility of adopting one (or some) of them as standard benchmarks. A recently created test collection that makes use of data from CiteSeerx is large, heterogenous, and publicly available, but has certain other limitations. In this paper, we propose a way to modify this test collection to address these limitations. We also use the improved test collection to establish a set of baseline results using elementary content-based techniques, as well as reference directed indexing.