Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling
Recording university lectures through lecture capture systems is increasingly common. However, a single continuous audio recording is often unhelpful for users, who may wish to navigate quickly to a particular part of a lecture, or locate a specific lecture within a set of recordings. A transcri...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Computer Science
2016
|
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613186715287552 |
|---|---|
| access_status_str | Open Access |
| author | Marquard, Stephen |
| author_browse | Marquard, Stephen |
| author_facet | Marquard, Stephen |
| author_sort | Marquard, Stephen |
| collection | Thesis |
| description | Recording university lectures through lecture capture systems is increasingly common.
However, a single continuous audio recording is often unhelpful for users, who may wish
to navigate quickly to a particular part of a lecture, or locate a specific lecture within a set
of recordings.
A transcript of the recording can enable faster navigation and searching. Automatic speech
recognition (ASR) technologies may be used to create automated transcripts, to avoid the
significant time and cost involved in manual transcription.
Low accuracy of ASR-generated transcripts may however limit their usefulness. In
particular, ASR systems optimized for general speech recognition may not recognize the
many technical or discipline-specific words occurring in university lectures. To improve
the usefulness of ASR transcripts for the purposes of information retrieval (search) and
navigating within recordings, the lexicon and language model used by the ASR engine may
be dynamically adapted for the topic of each lecture.
A prototype is presented which uses the English Wikipedia as a semantically dense, large
language corpus to generate a custom lexicon and language model for each lecture from a
small set of keywords. Two strategies for extracting a topic-specific subset of Wikipedia
articles are investigated: a naïve crawler which follows all article links from a set of seed
articles produced by a Wikipedia search from the initial keywords, and a refinement which
follows only links to articles sufficiently similar to the parent article. Pair-wise article
similarity is computed from a pre-computed vector space model of Wikipedia article term
scores generated using latent semantic indexing.
The CMU Sphinx4 ASR engine is used to generate transcripts from thirteen recorded
lectures from Open Yale Courses, using the English HUB4 language model as a reference
and the two topic-specific language models generated for each lecture from Wikipedia. |
| format | Thesis |
| id | oai:open.uct.ac.za:11427/21226 |
| institution | University of Cape Town (South Africa) |
| language | eng |
| last_indexed | 2026-06-10T12:32:08.355Z |
| license_str | Not specified — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from UCTD — University of Cape Town Open Access Repository |
| publishDate | 2016 |
| publishDateRange | 2016 |
| publishDateSort | 2016 |
| publisher | Computer Science |
| publisherStr | Computer Science |
| record_format | dspace |
| source_str | UCTD — University of Cape Town Open Access Repository |
| spelling | oai:open.uct.ac.za:11427/21226 Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling Marquard, Stephen Recording university lectures through lecture capture systems is increasingly common. However, a single continuous audio recording is often unhelpful for users, who may wish to navigate quickly to a particular part of a lecture, or locate a specific lecture within a set of recordings. A transcript of the recording can enable faster navigation and searching. Automatic speech recognition (ASR) technologies may be used to create automated transcripts, to avoid the significant time and cost involved in manual transcription. Low accuracy of ASR-generated transcripts may however limit their usefulness. In particular, ASR systems optimized for general speech recognition may not recognize the many technical or discipline-specific words occurring in university lectures. To improve the usefulness of ASR transcripts for the purposes of information retrieval (search) and navigating within recordings, the lexicon and language model used by the ASR engine may be dynamically adapted for the topic of each lecture. A prototype is presented which uses the English Wikipedia as a semantically dense, large language corpus to generate a custom lexicon and language model for each lecture from a small set of keywords. Two strategies for extracting a topic-specific subset of Wikipedia articles are investigated: a naïve crawler which follows all article links from a set of seed articles produced by a Wikipedia search from the initial keywords, and a refinement which follows only links to articles sufficiently similar to the parent article. Pair-wise article similarity is computed from a pre-computed vector space model of Wikipedia article term scores generated using latent semantic indexing. The CMU Sphinx4 ASR engine is used to generate transcripts from thirteen recorded lectures from Open Yale Courses, using the English HUB4 language model as a reference and the two topic-specific language models generated for each lecture from Wikipedia. 2016-08-13T18:55:00Z 2016-08-13T18:55:00Z 2012 2016-08-13T18:25:02Z Master Thesis Masters MPhil http://hdl.handle.net/11427/21226 eng application/pdf Computer Science Unknown University of Cape Town University of Cape Town |
| spellingShingle | Marquard, Stephen Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling |
| thesis_degree_str | Master's |
| title | Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling |
| title_full | Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling |
| title_fullStr | Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling |
| title_full_unstemmed | Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling |
| title_short | Improving Searchability of Automatically Transcribed Lectures Through Dynamic Language Modelling |
| title_sort | improving searchability of automatically transcribed lectures through dynamic language modelling |
| url | http://hdl.handle.net/11427/21226 |
| work_keys_str_mv | AT marquardstephen improvingsearchabilityofautomaticallytranscribedlecturesthroughdynamiclanguagemodelling |