Searching for similar music tracks 🎹🎻🎸 David Pilato | @dadoonet
Slide 1
Slide 2
Elasticsearch You Know, for Vector Search
Slide 3
Example: 1-dimensional vector Character Vector [ 1 ] ] Realistic
[ Embeddings represent your data Cartoon 1
Slide 4
represent different data aspects Human Character Vector [ 1, 1 Realistic Cartoon ] ] Machine
[ Multiple dimensions 1, 0
Slide 5
is grouped together Human Character Vector [ 1.0, 1.0 Realistic Cartoon 1.0, 0.0 [ 1.0, 0.8 ] ] ]
Machine
[ Similar data
Slide 6
Vector search ranks objects by similarity (~relevance) to the query Human Rank Query 1 Realistic Cartoon 2 3 4 5 Machine Result
Slide 7
Similarity: cosine (cosine) Human q cos(θ) = d1 d2 Realistic θ q⃗ × d ⃗ | q⃗ | × | d |⃗ _score = 1 + cos(θ) 2
Slide 8
Similarity: cosine (cosine) 1+1 _score = =1 2 1+0 _score = = 0.5 2 1−1 _score = =0 2
Slide 9
Similarity: Dot Product (dot_product) q⃗ × d ⃗ = | q⃗ | × cos(θ) × | d |⃗ q d θ | q⃗ | × co s (θ ) 1 + dot_ product(q, d) scorefloat = 2 0.5 + dot product(q, d) _scorebyte = 32768 × dims
Slide 10
Similarity: Euclidean distance (l2_norm) y 2 n i (x ∑ 1 i= − y i) q l2_normq,d = y1 d x1 y2 x2 n ∑ i=1 (xi − yi) 1 _score = 1 + (l2_normq,d )2 x 2
Slide 11
How do you index vectors ?
Slide 12
Data Ingestion and Embedding Generation POST /_doc { “_id”:”product-1234”, “product_name”:”Summer Dress”, “description”:”Our best-selling…”, “Price”: 118, “color”:”blue”, “fabric”:”cotton”, “fabric”:”cotton” } “desc_embedding”:[0.452,0.3242,…], “desc_embedding”:[0.452,0.3242,…] } “img_embedding”:[0.012,0.0,…] } Source data POST /_doc
Slide 13
How do you search vectors ?
Slide 14
Vector Query GET product-catalog/_search { “knn”: { “field”: “desc_embbeding”, “k”: 5, “num_candidates”: 50, “query_vector”: [0.123, 0.244,…], “filter”: { “term”: { “department”: “women” } } } }, “size”: 10
Slide 15
https://djdadoo.pilato.fr/
Slide 16
https://github.com/dadoonet/music-search/
Slide 17
Searching for similar music tracks 🎹🎻🎸 David Pilato | @dadoonet
