AMI-Topic-Segmentation-Processor

Download and Process AMI Corpus for Topic Segmentation

AMI Meeting Corpus

• Link
• Number of meetings after preprocessing : 139

Requirements

Tested on Python3.8

pip install -r requirements.txt

How to Use

Download and collect topics from each meeting

python download_and_process.py

• Original AMI Corpus will be downloaded into : data/ami_public_manual_1.6.2.
• Meeting transcription after processing: data/transcripts.

Each meeting with topics can be found in data/transcripts/{meeting_id}.json. In {meeting_id}.json are list of topics :

[                                                    
 {
  "topic_idx": "top.4",
  "topic_type": "other",
  "other_description": "introduction of participants and their roles",
  "sentences": [
   {
    "filename": "ES2002a.B.words.xml",
    "start_word": 0,
    "end_word": 71,
    "text": "Okay . Right . [vocalsound] Um well this is the kick-off meeting for our our project . Um [vocalsound] and um this is just what we're gonna be doing over the next twenty five minutes . Um so first of all , just to kind of make sure that we all know each other , I'm Laura and I'm the project manager . [vocalsound] Do you want to introduce yourself again ?"
   }, 
   ...
   
  ], # dialogue in topic
  "sub_topics":[] # sub topics are also list of topics
 }, # topic end
 ...
 
]

Acknowledgement

If you use the AMI Meeting Corpus, please add the citation:

@INPROCEEDINGS{Mccowan05theami,
    author = {I. Mccowan and G. Lathoud and M. Lincoln and A. Lisowska and W. Post and D. Reidsma and P. Wellner},
    title = {The AMI Meeting Corpus},
    booktitle = {In: Proceedings Measuring Behavior 2005, 5th International Conference on Methods and Techniques in Behavioral Research. L.P.J.J. Noldus, F. Grieco, L.W.S. Loijens and P.H. Zimmerman (Eds.), Wageningen: Noldus Information Technology},
    year = {2005}
}