A practical jupyter notebook tutorial for Whoosh (hope this will help new users)

[ruow...@gmail.com]

Hi,

I am a research assistant in McGill University who are working for implementing different query parsers. Recently I find Whoosh on Google and enjoy the powerful predefined functions and extensions it gives to users after experimenting it with my data.

The documentation of Whoosh is clear and in detail, but there is no practical tutorial to implement Whoosh in a complete searching process. I really like the explanation for each step in building a search engine in Whoosh documentation, but a python tutorial start from the very begining to the end with the same documents might be more intuitional and easy to understand for new users.

That's why I write this python tutorial for Whoosh. This tutorial emphasizes on query parsing section since my research is mainly about it, but other parts like indexing documents are covered in my tutorial as well.

Hope this will help and I am willing to share this tutorial with other Whoosh users if you want to put it on Whoosh documentation website.

Best regards,

Ruowei  

[ruow...@gmail.com]

Sorry I just found there were some formatting problems in the file, please see the updated one I attached in this reply.

Regards,

Ruowei

在 2018年2月2日星期五 UTC+8上午5:14:45，ruow...@gmail.com写道：

[Chanakya]

Hello Ruowei,

Thank for the wonderful tutorial.

Can you help with language modal.

I am trying hard to implement my own scoring function based on unigram retrieval modal (probablistic modal)

I know how to do mathematically. But I think it will great, if we together can built the custom scoring function.

Kind regards

[Eashan 123]

Thanks a lot, for this wonderful tutorial.

[Andreas Balogh]

Hi Rouwei,

thanks for the Python notebook tutorial. An example usecase is actually what's missing for the otherwise excellent documentation.

Andreas

[David Lowry-Duda]

I've now set up a few simple whoosh-based systems. I was thinking about writing down the general process, since there isn't a lot of examples out there. But I think whoosh has been fairly straightforward and useful. Is this something that you would be interested in?