Voice Information Retrieval In Collaborative Information Seeking

TL;DR

This paper explores voice information retrieval within collaborative information seeking, proposing a system that transcribes spoken queries using ASR with MFCC and HMM, achieving 81.25% accuracy, to support group-based search activities.

Contribution

It introduces a voice-based CIS system integrating ASR techniques for transcription, addressing the lack of collaborative voice search support in existing solutions.

Findings

Achieved 81.25% transcription accuracy

Integrated ASR with CIS for collaborative search support

Demonstrated system effectiveness through simulation

Abstract

Voice information retrieval is a technique that provides Information Retrieval System with the capacity to transcribe spoken queries and use the text output for information search. CIS is a field of research that involves studying the situation, motivations, and methods for people working in a collaborative group for information seeking projects, as well as building a system for supporting such activities. Humans find it easier to communicate and express ideas via speech. Existing voice search like Google and other mainstream voice search does not support collaborative search. The spoken speeches passed through the ASR for feature extraction using MFCC and HMM, Viterbi algorithm precisely for pattern matching. The result of the ASR is then passed as input into CIS System, results is then filtered to have an aggregate result. The result from the simulation shows that our model was able to achieve 81.25% transcription accuracy.