Otis Gospondnetic,Sematext @ Lucene Revolution 2011
-
Upload
lucid-imagination -
Category
Documents
-
view
219 -
download
0
description
Transcript of Otis Gospondnetic,Sematext @ Lucene Revolution 2011
![Page 1: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/1.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.1
Search Analytics
What? Why? How?
Otis Gospodnetić – Sematext International
![Page 2: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/2.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.2
About Otis Gospodnetić
• Member: Apache Lucene, Solr, Nutch, Mahout
• Author: Lucene in Action 1 & 2
• Entrepreneur: Sematext, Simpy
![Page 3: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/3.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.3
About Sematext
Consulting, development, support:
Search (Lucene, Solr, Elastic Search...) Big Data (Hadoop, HBase, Voldemort...) Web Crawling (Nutch) Machine Learning (Mahout)
![Page 4: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/4.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.4
Agenda Intro: Otis & Sematext - DONE What Why Reports
![Page 5: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/5.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.5
What is Search Analytics? Input: queries and clicks Output: reports – over time Next: actions
The means, not the goal Ongoing, not one-off
![Page 6: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/6.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.6
Search Analytics and SEO Not the same SA can help with SEO
![Page 7: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/7.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.7
Search vs. Web Analytics User intent and information needs vs. inferring Hand in hand Ideally you can relate data from both or even
unify it
![Page 8: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/8.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.8
Why Search Analytics? Measure and monitor everything Supports (re)design, navigation choices Helps with content acquisition & enhancement Improve search experience Mula
![Page 9: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/9.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.9
Report Groups Failures vs. non-failures Actionable vs. non-actionable
![Page 10: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/10.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.10
Failures
Be aware of failures, but don't be one.
Zero hits Low query CTR High search exit rate Irrelevant results Over N refinements
![Page 11: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/11.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.11
Report: Zero Hit Queries Popular queries; percentage, not raw count Misspellings? Synonyms? No matching content? Need (different) tagging? Bad analysis? Multilingual issue?
![Page 12: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/12.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.12
Report: Zero Hit Queries (cont.) Use Query Spellchecker (aka DYM) Using AutoComplete Using DYM ReSearcher Designing No Results page
![Page 13: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/13.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.13
Report: High Exit Rate Queries Disappointed, frustrated users Major revenue loss, no second chance Marriage with Web Analytics Relevance bad? Default ordering bad? Titles of hits need adjusting? Search terms highlighting looking bad? Bad thumbnails? Need thumbnails?
![Page 14: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/14.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.14
Report: Irrelevant Result Queries Manual: top N hits of top N queries Judge relevance of each hit and assign score Per-query score: sum scores of top N hits Cumulative top N query score: sum per-query
scores
Automated: Mean Reciprocal Rank (MRR)
![Page 15: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/15.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.15
Report: Total Queries Search vs. navigation/browsing Search vs. overall site usage Related report: % of visits with search Segment: new users vs. return users, etc.
Questions: do you count paging? Facet selection? Re-sorting?
![Page 16: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/16.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.16
Report: Total Distinct Queries What's distinct? Car vs. Cars # Total Queries / # Distinct Queries = Avg. # Tied to performance and query cache utilization Extension: Total distinct words in queries
![Page 17: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/17.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.17
Report: Words Per Query Informative, slowly changing, not terribly
actionable Can affect search box size Use AutoComplete if queries are long
![Page 18: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/18.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.18
Report: Top Queries User intent and information needs Ensure good results Calculate MRR for top N queries Calculate MRR for each top N query Compare to global MRR
![Page 19: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/19.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.19
Report: Top Queries (cont.) New top queries – new trend? New demand? Best Bets (aka Query Elevation in Solr) Expose before search is needed Seasonality – hour of day, day of the week, etc.
Adjust content presentation and availability (e.g. week vs. weekend, business vs. personal)
Anticipate demand in the next cycle
![Page 20: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/20.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.20
Clickstream Analysis Query analysis is not a complete story:
Queries Clicks (Trans)action
![Page 21: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/21.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.21
Query and Hit Valuation Query: by popularity (count) Query: by CTR Query: by subsequent (trans)action count/pct.
Hit: by click count Hit: by subsequent (trans)action count/pct.
![Page 22: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/22.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.22
Query and Hit Valuation (cont.) Maximize:
pop(query) + ctr(query) + action(query)clicks(hit) + action(hit)
Failures:high pop(q), yet low ctr(q)
high pop(q), high ctr(q), yet low action(q)
Integration with backend required
![Page 23: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/23.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.23
Report: Low CTR Queries Popular queries; percentage, not raw count Relevance bad? Default ordering bad? Titles of hits need adjusting? Search terms highlighting looking bad? Bad thumbnails? Need thumbnails?
![Page 24: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/24.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.24
Report: Queries with Most Clicks i.e. Queries with Highest CTR Informative? Yes Actionable? Somewhat: expose relevant
content outside of search
![Page 25: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/25.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.25
Search Session Search activity aimed at satisfying a specific
information need in a some limited amount of time.
i.e. it's very fuzzy
![Page 26: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/26.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.26
Interesting Search Sessions More than N queries in M minutes Sessions that end in a failure Sessions for specific type of info (e.g. person
name, product name, event)
![Page 27: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/27.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.27
Segmentation Searches that resulted in conversion vs. not Search metrics for New vs. returning visitors English vs. French vs. Spanish vs. … Chrome vs. IE ...
![Page 28: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/28.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.28
More SA Reports/Questions % of queries from DYM vs. AC vs. typed Most common queries per clicked hit Which hits are generally popular? Which hits are trending up? Are there docs that are never ever clicked on? Average number of queries per session Breakdown of queries by number of hits
![Page 29: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/29.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.29
More SA Reports/Questions Breakdown of queries by latency Frequently used facets or sort criteria Avg number of clicks per query Time spent on site before/after searching Search initiation pages How deep into SERPs are people drilling? Are too many clicks on pages other than 1st? ...
![Page 30: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/30.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.30
Data Collection
![Page 31: Otis Gospondnetic,Sematext @ Lucene Revolution 2011](https://reader034.fdocuments.net/reader034/viewer/2022051704/568c510b1a28ab4916b10c58/html5/thumbnails/31.jpg)
Copyright 2011 Sematext Int'l. All rights reserved.31
• sematext.com• blog.sematext.com• @sematext• @otisg• [email protected]
Contact