120 likes | 137 Views
Implementing a dialogue-driven system for university intranet search, enhancing user experience with natural language interactions and domain-specific knowledge extraction.
E N D
Suma Adindla School of Computer Science & Electronic Engineering sadind@essex.ac.uk Dialogue – Driven Intranet Search 8th LANGUAGE & COMPUTATION DAY 2009 LAC 09
Motivating Example Imagine you could interact with a university intranet search engine as follows: User : Head of department System : Which department are you looking for? User : Computer science System : The head of the computer science department is Dr. Maria Fasli. Her contact details are as follows ... Do you want any further information? User : How do I get to her office? System : The quickest route from the University information point is as follows ... 2 LAC 09 2
Problem Overview • Short queries in most cases (often just keywords rather than • questions) • Unstructured data to start with • Explicit domain knowledge • Wide range of possible user queries LAC 09
Table 1. Research Context 30 October 2009 LAC 09 4
Possible Solution • Impose a dialogue manager to assist a user in the navigation process • Turn document collection into usable knowledge source • Employs automatically extracted domain knowledge LAC 09
Research Questions to Address • 1. How can a dialogue system be incorporated into the search process to provide the user with a more natural language interface? • 2. Can such a dialogue system on an intranet provide the user with more relevant information and offer a better user experience than a standard search engine? LAC 09
System Architecture Dialogue Manager request Search Engine User response suggestions Domain Model clarifications Query Analysis 20 December 2019 7 LAC 09
Online query processing Document Processing Identifying Triplets Extracting Entities and Facts Domain Knowledge Dialogue Manager Query Analysis (Question type) Pre- processing Query mapping Process User 30 October 2009 LAC 09 8 Match a user query against the knowledge base Imposed dialogue system helps a user in the navigation of results
Extraction process Documents Sentence Detection Parts of Speech Tagging Parsing Named Entity Recognition Reference Resolution 20 December 2019 LAC 09 9
Number of interaction steps (dialogue length) Time taken to process a user query Precision and recall (success rate of retrieved results) User satisfaction Evaluation Parameters LAC 09
Future challenges • Extracting useful knowledge from the document collection • Knowledge representation • Interactive dialogue manager with some domain knowledge • Natural language interface design LAC 09
Thank you 30 October 2009 LAC 09 12