Predictive models are all pervasive with usage in search engines, recommender systems, health, legal and financial domains. But for the most part they are used as black boxes which output a prediction, score or rankings without understanding partially or even completely how different features influence the model prediction.

Machine learning models are progressively becoming complex and training datasets are getting larger by the data. Embeddings models are trained over Web scale collections of text and graphs, language models are learnt over millions or billions of sentences.

My main focus in information retrieval is currently investigating neural models for ranking models for document retrieval and open-domain question answering. I also work extensively on temporal information retrieval and enrichment of textual knowledge sources like Wikipedia.