Lab Report 01.10.2024

📅 Date 01.10.2024

✔️ Participant List

• Mehmet Efe AKCA
• Nazire ATA
• Enes BASER
• Cagatay COLAK
• Aslı GOK
• Tarık Can Özden
• Ahmet Enes Semerci
• Atakan YASAR

📑 Work Done During the Lab (Deliverables)

1. Discussed the requirements in the Requirements Specification document 🔗 Document Link

2. Discussed the selection of W3C standard of the project

3. Elicited requirements with client, received following deliverables (not exhaustive):

• • Implement a recommendation algorithm

4. Discussed recommendation algorithms that can be used:
   • Vectorization of the users & posts with programming languages to find cosine similarity:
     • For each user (0 for languages not followed and 1 for languages followed),
     • For each posted question (0 if language not in tags and 1 if language in tags)
   • Vectorization of the posts with doc2vec embeddings:
     • Users' vectors can be the average of the vectors of liked posts.
     • Posts' vectors can be calculated by concatenating title and descriptions with doc2vec embeddings.
   • Graph based recommendations using Graph Databases.

📌 Action Items

1. Created Customer Milestone-1

References

• 🔗 pgvector
• 🔗 TF-IDF
• 🔗 Grahp DB