Found 182 bookmarks
Newest
Harsh Thakkar on Twitter
Harsh Thakkar on Twitter
For the interested -- you can find an empirical study of the sparql-gremlin mapping used by the plugin here: https://t.co/QGIzufzDKo— Harsh Thakkar (@Harsh9t) January 11, 2019
·twitter.com·
Harsh Thakkar on Twitter
Andrei Kashcha on Twitter
Andrei Kashcha on Twitter
https://t.co/7T0EOs6yG7 - Made this tiny tool to discover related subreddits.The graph is created based on jaccard similarity between two subreddits. Jaccard similarity is constructed from set of shared users.Source code https://t.co/J9r1jl1JjR pic.twitter.com/4hcg7mI4sg— Andrei Kashcha (@anvaka) January 10, 2019
·twitter.com·
Andrei Kashcha on Twitter
TinkerPop on Twitter: "3.4.0 Highlights - The Gremlin Recipes have been expanded with a new Anti-Patterns section that describes common pitfalls around Gremlin readability, maintainability and performance. https://t.co/NwUNY7zFqY #graphdb… https://t.co/mY
TinkerPop on Twitter: "3.4.0 Highlights - The Gremlin Recipes have been expanded with a new Anti-Patterns section that describes common pitfalls around Gremlin readability, maintainability and performance. https://t.co/NwUNY7zFqY #graphdb… https://t.co/mY
“3.4.0 Highlights - The Gremlin Recipes have been expanded with a new Anti-Patterns section that describes common pitfalls around Gremlin readability, maintainability and performance. https://t.co/NwUNY7zFqY #graphdb”
·twitter.com·
TinkerPop on Twitter: "3.4.0 Highlights - The Gremlin Recipes have been expanded with a new Anti-Patterns section that describes common pitfalls around Gremlin readability, maintainability and performance. https://t.co/NwUNY7zFqY #graphdb… https://t.co/mY
Nicolas Torzec on Twitter: "Q: which product taxonomies are used in the Shopping / Ad industries? Google's Product Taxonomy is a de facto standard but it lacks freshness, coverage and/or finesse in some areas. I'm also looking at product taxonomies from A
Nicolas Torzec on Twitter: "Q: which product taxonomies are used in the Shopping / Ad industries? Google's Product Taxonomy is a de facto standard but it lacks freshness, coverage and/or finesse in some areas. I'm also looking at product taxonomies from A
“Q: which product taxonomies are used in the Shopping / Ad industries? Google's Product Taxonomy is a de facto standard but it lacks freshness, coverage and/or finesse in some areas. I'm also looking at product taxonomies from Amazon, Ebay, Walmart, Target, Groupon. What else?”
·twitter.com·
Nicolas Torzec on Twitter: "Q: which product taxonomies are used in the Shopping / Ad industries? Google's Product Taxonomy is a de facto standard but it lacks freshness, coverage and/or finesse in some areas. I'm also looking at product taxonomies from A
Bryan J. Brown on Twitter: "So can any SemWeb gurus explain to me what the difference is between SHACL and ShEx? If SHACL is an official W3C recommendation, why use ShEx? Why is effort being split on this? Not a criticism at all, but a genuine question as
Bryan J. Brown on Twitter: "So can any SemWeb gurus explain to me what the difference is between SHACL and ShEx? If SHACL is an official W3C recommendation, why use ShEx? Why is effort being split on this? Not a criticism at all, but a genuine question as
So can any SemWeb gurus explain to me what the difference is between SHACL and ShEx? If SHACL is an official W3C recommendation, why use ShEx? Why is effort being split on this? Not a criticism at all, but a genuine question as I'm new to all this and only familiar with SHACL.— Bryan J. Brown (@bryjbrown) November 26, 2018
·twitter.com·
Bryan J. Brown on Twitter: "So can any SemWeb gurus explain to me what the difference is between SHACL and ShEx? If SHACL is an official W3C recommendation, why use ShEx? Why is effort being split on this? Not a criticism at all, but a genuine question as
Martynas Jusevicius on Twitter: "Am I the only one who thinks there's nothing really wrong with the #RDF stack? To me it's such a powerful technology that 20 years in we're still exploring what it is capable of. To do that, we need to throw out the old so
Martynas Jusevicius on Twitter: "Am I the only one who thinks there's nothing really wrong with the #RDF stack? To me it's such a powerful technology that 20 years in we're still exploring what it is capable of. To do that, we need to throw out the old so
Am I the only one who thinks there's nothing really wrong with the #RDF stack?To me it's such a powerful technology that 20 years in we're still exploring what it is capable of. To do that, we need to throw out the old software design paradigms and think graph- and Web-native.— Martynas Jusevicius (@namedgraph) November 27, 2018
·twitter.com·
Martynas Jusevicius on Twitter: "Am I the only one who thinks there's nothing really wrong with the #RDF stack? To me it's such a powerful technology that 20 years in we're still exploring what it is capable of. To do that, we need to throw out the old so
Harsh Thakkar on Twitter
Harsh Thakkar on Twitter
At long last! With this, we aim to bridge the gap between semantic web and graph database communities. Happy graph-querying! https://t.co/8al1uLi1aC— Harsh Thakkar (@Harsh9t) August 17, 2018
·twitter.com·
Harsh Thakkar on Twitter
On "Benchmarking RedisGraph 1.0"
On "Benchmarking RedisGraph 1.0"
Recently RedisGraph published a blog [1], comparing their performance to that of TigerGraph’s, following the tests [2] in TigerGraph’s benchmark report [3], which requires solid performance on 3-hop, 6-hop, and even 10-hop queries. Multi-hop queries on large data sets are the future of graph analytics....
·tigergraph.com·
On "Benchmarking RedisGraph 1.0"