Python - TED recommendation system
The project aimed to develop TED recommendation system via Python language in Jupyter Notebook. Following that, the system was based on topic modeling using gensim package. Models with different number of topics were trained, only the one produced the consistent coherence score was chosen (in this case, topics = 40). The results were evaluated based on how relevant the recommendations of your topic model are and provide reasons to support the judgments. The two videos were used in the evaluation section are:
- Kriti Sharma: How to keep human bias out of AI 2018-03-23-kriti_sharma_how_to_keep_human_biases_out_of_ai.txt
- Timothy Bartik: The Economic Case for Preschool 2012-09-14-timothy_bartik_the_economic_case_for_preschool.txt
The file would contain two files:
- Python - TED recommendation system - Report
- Python - TED recommendation system - Script