Saurav Pandey

Qontent Explorer

• Constructed an automated knowledge graph pipeline to extract meaningful information from the documents in the for of triplet
• Used Neo4j for Querying the Knowledge Graph. Used Cypher Query to query the Knowledge Graph in Neo4j
• We implemented Question answering system to find response to a query that has no defined context.
• Created a DeepSearch functionality to search for the most relevant paragrapha and triplets to the query
• For storing the unstructured data, we used the BigQuery on Google Cloud Platform.

• With a dataset of customer reviews, we developed an NLP model for sentiment analysis, attaining an accuracy of 85% .
• A named entity recognition system was put in place to extract important entities from legal documents, increasing the accuracy of document classification.
• Analyzed the performance of several algorithms on a big corpus of news stories while conducting research on text summarizing methods.
• Developed a chatbot utilizing NLP and machine learning in collaboration with a team, increasing customer service by automating answers to frequently asked queries.