Nam Khanh Tran

Worked in R&D projects as a research scientist and software engineer: i) eLabour: Developed Multihop Attention Networks for question answering. Implemented an exploratory search system using ElasticSearch and AngularJS. ii) ForgetIT: Developed Ranking algorithms to retrieve additional context for supporting interpretations of documents. iii) Alexandria: Developed a probabilistic model for context-aware entity recommendation using entity embeddings and entity graph

Worked in Amazon Machine Learning team, developed Deep Learning models to exploit syntactic and semantic information in tree-structured Long Short-Term Memory (LSTM) networks.

Worked in a Google Award Research project "Semantic spaces from text and images". Proposed and implemented an algorithm for combining textual and visual features in a semantic model.

Worked in an internal project "Automatic learning and recognizing patterns which link the word sense to the word usage". Implemented the Angluin Algorithm for learning corpus patterns using finite state automata