You are on page 1of 9

Automated Meeting

Summarizer
Milan Jijo,Pious Tony C, Vishnu P
(Group 17)
March 17th, 2023

1
Contents
1 Introduction 3

2 Overall Desccription 3

3 System Design 3
3.1 Data Preprocessing Module . . . . . . . . . . . . . . . . . . . . . 3
3.2 Natural Language Processing . . . . . . . . . . . . . . . . . . . . 3

4 Component-level Design 4
4.1 Audio Input Component . . . . . . . . . . . . . . . . . . . . . . . 4
4.2 Audio Input Component . . . . . . . . . . . . . . . . . . . . . . . 4
4.3 Natural Language Processing Component . . . . . . . . . . . . . 4
4.4 User Interface Component . . . . . . . . . . . . . . . . . . . . . . 5

5 Interface Design 5
5.1 Login Screen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.2 Upload Audio Screen . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.3 Meeting Summary Screen . . . . . . . . . . . . . . . . . . . . . . 5
5.4 Settings Screen . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.5 Navigation Bar . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

6 Data Flow 5

7 UML Diagrams 7
7.1 Use Case Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 7
7.2 DFD Level 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
7.3 DFD Level 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2
1 Introduction
The auto meeting summarizer is a software system designed to automatically
summarize meetings in order to save time and increase productivity. The sys-
tem will use speech recognition technology to transcribe meeting audio, natural
language processing techniques to extract key information from the transcript,
and machine learning algorithms to generate a summary of the meeting.

2 Overall Desccription
System Overview The Auto Meeting Summarizer system consists of the follow-
ing modules:
• Data Preprocessing Module: This module will be responsible for convert-
ing audio recordings of meetings into text format, removing noise and
irrelevant information, and identifying the speaker.
• Natural Language Processing Module: This module will extract key phrases
and sentences from the meeting transcript, identify the main topics dis-
cussed, and generate a summary of the meeting.
• User Interface Module: This module will provide a user-friendly inter-
face for users to upload audio recordings, view meeting summaries, and
download meeting transcripts and summaries.

3 System Design
The Auto Meeting Summarizer system will be designed using a client-server
architecture, where the client is the user interface module and the server consists
of the data preprocessing and natural language processing modules. The system
will be implemented using Python programming language.

3.1 Data Preprocessing Module


1. Convert audio recordings of meetings into text format using automatic
speech recognition (ASR) technology.
2. Remove irrelevant information such as background noise, fillers, and in-
terruptions.
3. Segment the transcript into smaller sections to facilitate NLP processing.

3.2 Natural Language Processing


1. Extract key phrases and sentences from the meeting transcript using tech-
niques such as named entity recognition (NER), part-of-speech (POS) tag-
ging, and dependency parsing.

3
2. Identify the main topics discussed during the meeting using topic modeling
techniques such as latent Dirichlet allocation (LDA).
3. Generate a summary of the meeting by selecting the most important sen-
tences based on their relevance to the identified topics.

4 Component-level Design
4.1 Audio Input Component
The audio input component is responsible for capturing audio from microphones
or recording devices.
1. Microphone or recording device:This subcomponent will capture au-
dio input.

4.2 Audio Input Component


The speech recognition component is responsible for converting audio input into
text.
1. Audio input module: This subcomponent will receive audio input from
the audio input component.
2. Speech recognition engine: This subcomponent will convert audio in-
put into text using speech recognition algorithms.

4.3 Natural Language Processing Component


The natural language processing component is responsible for extracting key
information from the text generated by the speech recognition component

1. Text input module: This subcomponent will receive text input from
the speech recognition component.
2. Topic modeling module:This subcomponent will identify the main top-
ics discussed in the meeting.

3. Action item detection module: This subcomponent will identify ac-


tion items assigned during the meeting.
4. Decision detection module: Decision detection module: This sub-
component will identify decisions made during the meeting.
5. Summary generation module: This subcomponent will generate a
summary of the meeting based on the extracted information.

4
4.4 User Interface Component
The user interface component is responsible for providing a web or mobile in-
terface for users to upload audio files and access meeting summaries
1. File upload module: This subcomponent will allow users to upload
audio files.
2. Meeting summary display module: This subcomponent will display
the meeting summary generated by the machine learning component.

5 Interface Design
5.1 Login Screen
The login screen will allow users to enter their login credentials or create a new
account.

5.2 Upload Audio Screen


The upload audio screen will allow users to upload an audio file of the meeting
they want to summarize. Users will be able to select the file from their device
and upload it to the system.

5.3 Meeting Summary Screen


The meeting summary screen will display the summary of the meeting generated
by the system. The summary will include the main topics discussed in the
meeting, action items assigned during the meeting, decisions made during the
meeting, and the overall sentiment of the meeting. Users will be able to view
and download the summary.

5.4 Settings Screen


The settings screen will allow users to customize the system settings, such as
the language used for speech recognition and the output format of the meeting
summary.

5.5 Navigation Bar


The navigation bar will allow users to switch between different screens, including
the upload audio screen, meeting summary screen, settings screen etc..

6 Data Flow
• Audio input is captured by microphones or recording devices.

5
• The audio input is processed by the speech recognition component to
produce text.
• This text is preprocessed to remove optimise it for NLP like removing stop
words etc

• The text is analyzed by the natural language processing component to


extract key information.
• The extracted information is used by the machine learning component to
generate a summary of the meeting.
• The summary is displayed in the user interface.

6
7 UML Diagrams
7.1 Use Case Diagram

7
7.2 DFD Level 0

8
7.3 DFD Level 1

You might also like