Past Events
PhD DefenseNeural Methods for Document Understanding |
|
||
Thursday, April 01, 2021, 12:30pm - 02:30pm |
|||
Speaker: Mohamed Abdellatif
Location : Remote via Zoom
Committee:
Prof. Ahmed Elgammal (Advisor)
Prof. Abdeslam Boularias
Prof. Gerard De Melo
Prof. Malihe Alikhani (external member)
Event Type: PhD Defense
Abstract: Document recognition involves both structured and unstructured data. This dissertation studies and proposes solutions to problems in both domains.Structured data often comes in tabular format. Many existing applications use structured tables as input such as spreadsheets, HTML, Excel, and CSV. However, there are a plethora of circumstances that can lead to generating images of tables (scanning invoices, sharing user manuals, downloading read-only publications, etc.). In such scenarios, we do not have access to the table structure. To be able to use the aforementioned applications in the given scenarios, we focus on the problem of structure recognition by graph modeling. We are investigating ways of inferring the hidden structures from tables’ images. By extracting table fields in the form of an undirected graph G and finding the corresponding Line graph L(G), AKA edge-to-vertex dual graph, we are able to model the structure-inference problem as a vertex classification using a Graph Convolutional Network (GCN) based model.We focus on text classification in three settings: English, multilingual and Arabic. We performed a comparative study with different models for text classification tasks using ULMFiT, Bert and XLM, across the spectrum of languages to understand their advantages and limitations.
: