Making a simple question into a complicated query
-
Upload
richard-boulton -
Category
Documents
-
view
790 -
download
0
Transcript of Making a simple question into a complicated query
![Page 1: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/1.jpg)
Making a simple question into a complicated query
Richard BoultonLemur Consulting Ltd
![Page 2: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/2.jpg)
Making a simple question into a complicated query
Richard BoultonLemur Consulting Ltd
![Page 3: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/3.jpg)
Making a simple question into a complicated query
Richard BoultonLemur Consulting Ltd
![Page 4: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/4.jpg)
Only 20 minutes until Lunch
![Page 5: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/5.jpg)
Where shall we have Lunch?
![Page 6: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/6.jpg)
“Lunch”
![Page 7: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/7.jpg)
Assertion
Complicated questions areeasier to answer well.
![Page 8: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/8.jpg)
Assertion
Complicated questions areeasier to answer well accurately.
![Page 9: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/9.jpg)
� Restaurant � Pizza restaurant near Covent Garden, fairly cheap.
![Page 10: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/10.jpg)
Time for a real example
http://mydeco.com/
Interior decoration site
![Page 11: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/11.jpg)
� Users type: “Sofa”
� We'd prefer them to ask questions like: “Red velvet, three seater, sofa, from a supplier who can deliver to central Cambridge at a weekend”.
� How can we move to this kind of search?
![Page 12: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/12.jpg)
Getting more from users
![Page 13: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/13.jpg)
Getting more from users
� Suggested search completions
![Page 14: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/14.jpg)
Getting more from users
� Facets
![Page 15: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/15.jpg)
Getting more from users
� Facets
� Which facets to display?
− Depends on the user.
� Which facet values are interesting?
− A particularly fun problem for continuous numeric values, like price.
� How many values should we display?
− Based on likelihood of any being useful?
![Page 16: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/16.jpg)
Getting more from users
� Personal data
− Using details about the user directly.
� e.g., Postcode
− Grouping users by similarity of interests
![Page 17: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/17.jpg)
Getting more from users
� Similarity search
− “More like this”
− Colour / image-based similarity
![Page 18: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/18.jpg)
Behind the scenes
� Applying our own bias.
− Perhaps we want to push some items
− Perhaps we want to avoid other items
− Perhaps some items go well together
− Behave like a shop assistant
− “Product Rank”
![Page 19: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/19.jpg)
Behind the scenes
� Categorisation
− User asks for “Sofa”.
− We search for “Products categorised as one of the sofa subcategories, based on the output of a machine learning system trained with some human judgements”.
![Page 20: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/20.jpg)
Behind the scenes
� Variety
− Don't display lots of very similar items
− Give the user a choice
− But don't display irrelevant junk, either!
− Need some way to measure variety
![Page 21: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/21.jpg)
Answering complicated questions
![Page 22: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/22.jpg)
Answering complicated questions
� Getting the best answer
− Good models
− Careful design
− Lots of tuning
![Page 23: Making a simple question into a complicated query](https://reader033.fdocuments.in/reader033/viewer/2022060121/559499cc1a28ab131f8b47cc/html5/thumbnails/23.jpg)
Answering complicated questions
� Getting an answer quickly
− Good algorithms
− Well matched data-structures
− Plenty of machines
− Plenty of RAM