Skip to content

Latest commit

 

History

History
40 lines (29 loc) · 2.52 KB

Readme.md

File metadata and controls

40 lines (29 loc) · 2.52 KB
Raw

Introduction

In this project, you will learn how to extract keywords or words that are more important than others in your sentence easily and implement them in an actual project. It is the plant keyword extraction project, the plant characterization word.

How to run project

Step 1: Working in statistics_words folder

  • Run the file convert json2txt to build the data corpus
  • Let start processing your corpus with pre_process file (pre: previously)
  • Now, run the statistics_words.js to see how often each word is repeated

Step 2: Switch to working in build_vecs folder, and don't forget copy your data has processed to using in there

Step 3: Let start training your Extract keywords model at train_extract_keywords model. As above, don't forget copy your data vecs, list keywords (positive) and list not keywords (negative) file from prev-steps.

  • Run create_data_train_extract.js to get your data train and data y_train (label of each word's vector)
  • Run the train.js to build your W (Weight for predicting)
  • Run the index.js to test your W (If you confusing where I build the file has the name contain "50d", it is also the data vecs has been built by another kind of NK-VEC that is still similar to your NK-VEC version, I just want to test new algorithm.). The data vector had been built by NK-VECTOR

Step 4: You can skip if you don't want to build the previous extract keywords for each plant. I create it for searching with const-databse (don't update if the user want to add a new plant)

  • Copy everyfile such as positive, negative files, the data vecs and your W to here
  • Run index.js

Step 5: You also skip if you skip step 4

  • Type your keyword in sentence variable
  • Node index.js to show your results of plant searching

Step 6: Run as API

  • Everything in here is similar to step 5.

Diagrams

Diagrams

Follow the flow of each diagram in this picture if you want to understand this project

P/s: The word statistics only give you an overview of your magnetic pocket. It will help you more conveniently recognize which words are keywords. It does NOT give you exactly where the keyword is!