AcadOverFlow is a platform for professional and enthusiast programmers of IIIT-H where technical questions can be posted/queried by any student, so that fellow peers can post responses to it or previously asked similar questions can be listed along with the answers.
The dataset for the required problem statement is quite big. So it need to be run on ada servers or the machines which have RAM > 12GB. Otherwise MemoryError would be shown.
Before running the flask server, there are some ML files that are need to be downloaded and kept in their specific folders and also some modules need to be installed:
(i) For downloading the ML related files, go to ML_Module/models and download the data shown upon clicking the links mentioned in the following files(and save at the same location with the same name):
(a) SO_word2vec_embeddings.bin
(b) Tag_predictor.h5
(c) title_embeddings.csv.zip, unzip it in the same location.
(d) tokenizer.pickle
Also download csv_files.zip, unzip it in SE_Project/AcadOverflow/app/session_data. After unzipping it, do the following:
$ cd csv_files
$ git checkout Preprocessed_users.csv
(ii) Install the following modules:
pip3 install -U gensim
pip3 uninstall zipp => Do this only if shows some error, no need to uninstall if zipp is not installed.
pip3 install inflect
pip3 install nltk
pip3 install -U spacy
pip3 install smart-open
python3 -m spacy download en_core_web_sm
To run the module, Go to AcadOverflow folder and run the following command:
python3 run.py
Webserver will run on http://127.0.0.1:8889/ . The website can be accessed using this address. The user can register and then login into the AcadOverFlow Webapp. :)