We started to investigate recommendation system from scrach. We will document all our findings about how to build a literature recommendation system, how many approach could be used, what is the best approach for accounting domain.
Initial Research
- Investigation about the recommendation systems - #8
- Data preprocessing techniques - #12
- Topic modeling techniques - #10
- Similarity measurements - #11
- Estimate computation power that will be needed in the development - #13
- Think about finding a performance measurement for the software. i.e. how to calculate accuracy - #14
Data Preprocessing
- PDF Text Extraction Tools - #15
- Dataset Analyser
- First LDA model - #16
- Data Preprocessing
- Milestone_1 enhancement after removed unrelavent papers
- Data preprocessing steps comparison based on LDA Recommendation