INTRODUCTION
WHY SEARCH AND WHY GRAPH
Search is a key feature of most applications. Users search for
products, places, other users, documents, and more. And while
structured search was common for early applications, today,
driven by the ubiquity of internet search engines, full-text
search dominates the usage.
The world around us consists of connected information. Being
able to store and query those rich networks of data allows us
to support decisions, make recommendations, and predict
impacts. Graph databases enable both transactional as well as
analytical uses on top of our highly connected domains.
Bringing both together allows us to enhance search results
with graph-based capabilities like recommendation features
or concept search, and also to use advanced search results as
entry points to graph traversals.
OUR USE CASE: MULTI-FACETED SEARCH WITH
RECOMMENDATIONS
Each domain has specific expectations in terms of search
relevance and a different set of issues, constraints, and
requirements.
Our use case example is product search, used by any retailer
(Amazon, eBay, Target, etc.).
Text search and catalog navigation are not only the entry points
for users but they are also the main “salespeople”. Compared
to other search engines, the set of “items” to be searched is
more controlled and regulated.
For the search infrastructure, these aspects have to be taken
into account:
• Multiple data sources: Products and related information
come from various heterogeneous sources like product
suppliers, information providers, and sellers.
• Marketing strategy: New promotions, offers, and marketing
campaigns are created to promote the site or specific
products. All of them should affect results boosting.
• Personalization: In order to provide a better and more
customized user experience, clicks, purchases, search
queries, and other user signals must be captured,
processed, and used to personalize search results.
• Provider information: Product suppliers are the most
important. They provide information like quantity,
availability, delivery options, timing, and changes in the
product’s details.
All these requirements and data sources affect search results in
several ways. Designing a search infrastructure for e-commerce
vendors requires an entire ecosystem of data and related data
flows together with platforms to manage them.
THE VALUE OF SEARCH
Search is a conversation between a user and a search engine.
Search is ubiquitous in modern applications. It’s the fastest way
to find relevant information in vast amounts of data. The search
engine needs the ability to provide relevant results to the user’s
search terms, as well as to further refine and filter the search.
FACETING
Initial search results are often too broad and need to be filtered
or refined, e.g., by using facets. Facets are categories derived
from the search results that are useful for narrowing a search.
Each facet represents an attribute of the structured information
like category, price, color, location, etc., and comes with a count
of contained results.
252
BY DR. ALESSANDRO NEGRO, MICHAEL HUNGER, AND CHRISTOPHE WILLEMSEN
Introduction
The Value of Search
The Power of the Graph
A Graph-Centric Architecture for Search
Platforms
Single Knowledge Graph, Multiple Views...
and more!
CONTENTS
DZONE.COM
|
© DZONE, INC. VISIT DZONE.COM/REFCARDZ FOR MORE!
Graph-Powered Search:
Neo4j & Elasticsearch
BROUGHT TO YOU IN PARTNERSHIP WITH