Extract trends from social media data
Budget: ₹600 – ₹1,500 INR
expected to identify trends from social media data; From all the
products available on Flipkart identify trending products, utilize all signals available (ex. posts, sessions,
check-ins, social graphs, media content, etc.). Output should also have photos, videos, gifs which can be
used on Flipkart app.Preferred tech: Open source
Bonus: Signal extraction from multiple social media channels (ex. FB, Instagram, Twitter, etc.)
Deliverable 1:
Identification of trends from social media
Identify trends on social media based on category. Can restrict to Fashion as a category for the project. Ex: Polka dots dresses are trending on twitter.
Ranking/scoring logic for trends extracted.
Outcome format:
: structured data according to flipkart category, sub category, vertical and product attributes Example: {category: Fashion, Sub-category: Women Western, vertical: Women dresses, trending attribute type: Pattern, trending attribute value: Polka Print, list of sample images and links from which the trend is derived}.
Deliverable 2:
Mapping trends with Flipkart products:
Create mapping of extracted trending keyword(s) with Flipkart category, sub category, vertical and product attribute(s), search page links. Example:{category: Fashion, Sub-category: Women Western, vertical: Women dresses, trending attribute type: Pattern, trending attribute value: Polka Print} Note: Use category, Subcategory combination from the Flipkart Website
From a trending keyword, creating a corresponding searchable term on Flipkart which will lead to matching products. Example: Tropical Tops keywords will not give right results directly on Flipkart but we can construct search query for it using some intelligence.
Points will be given based on similarity between sample images for trends and product results on Flipkart.
products available on Flipkart identify trending products, utilize all signals available (ex. posts, sessions,
check-ins, social graphs, media content, etc.). Output should also have photos, videos, gifs which can be
used on Flipkart app.Preferred tech: Open source
Bonus: Signal extraction from multiple social media channels (ex. FB, Instagram, Twitter, etc.)
Deliverable 1:
Identification of trends from social media
Identify trends on social media based on category. Can restrict to Fashion as a category for the project. Ex: Polka dots dresses are trending on twitter.
Ranking/scoring logic for trends extracted.
Outcome format:
: structured data according to flipkart category, sub category, vertical and product attributes Example: {category: Fashion, Sub-category: Women Western, vertical: Women dresses, trending attribute type: Pattern, trending attribute value: Polka Print, list of sample images and links from which the trend is derived}.
Deliverable 2:
Mapping trends with Flipkart products:
Create mapping of extracted trending keyword(s) with Flipkart category, sub category, vertical and product attribute(s), search page links. Example:{category: Fashion, Sub-category: Women Western, vertical: Women dresses, trending attribute type: Pattern, trending attribute value: Polka Print} Note: Use category, Subcategory combination from the Flipkart Website
From a trending keyword, creating a corresponding searchable term on Flipkart which will lead to matching products. Example: Tropical Tops keywords will not give right results directly on Flipkart but we can construct search query for it using some intelligence.
Points will be given based on similarity between sample images for trends and product results on Flipkart.