I am trying to migrate an application which was using AlchemyData News to search for News based on multiple entities in the query, for example
&q.enriched.url.entities.entity=|text=IBM,type=Company| &q.enriched.url.entities.entity=|text=Apple,type=Company|
The above query searches for IBM and Apple with entity type company
How do I do the same with Discovery News? The Migrating from AlchemyData News article on the one hand says:
Discovery and AlchemyData News use the same query syntax.
but also mentions:
The query structure and the structure of data returned is different between Watson Discovery News and AlchemyData News
The Alchemy query as is doesn't work.
Per Discovery documentation the the entity search query looks something like:
entities.text:IBM,entities.type:Company
However I am not sure how to specify multiple entities in this query, does anyone have an idea what the query could look like?
I think this is the query you are looking for: entities:(type:Company,(text:IBM|text:Apple))
Which says "Give me entities where the type is Company and the text is IBM or Apple"
(I am an IBM Watson Employee)
Related
OpenPhacts is a wonderful source of a number of RDFs related to drug discovery (https://www.openphacts.org/). The data is freely available (https://data.openphacts.org/commercial/2.1/)
Unfortunately there is no example of how to load it into a graph database such as neo4j or arangodb (http://support.openphacts.org/support/solutions/articles/168752-can-i-install-open-phacts-locally-)
Could someone provide a basic example on how to import one of the dozen RDFs into a graph database for a novice graph database user?
Thanks
Iain
Check out this post & tool by my colleague
https://jesusbarrasa.wordpress.com/2016/06/07/importing-rdf-data-into-neo4j/
It uses some sensible mapping and provides a Neo4j user defined procedure to do the hard work.
I'm implementing a sorting algorithm and I want to query the Insights for Twitter API in order to find if a topic is trendy in comparison to another topic.
As far as I've investigated, there is a Count API Operation which will return the number of Tweets based on a given query. Therefore I could query how many Tweets have the particular "keyword" in their body and then compare it to other different keywords to establish an order.
Is there is a more direct query to know if a keyword is trendy?
Your approach using the /api/v1/messages/count endpoint looks like the right one when working with the Insights for Twitter API.
I guess you could also use the posted query parameter to build trending charts over time for your different keywords.
My company wants to integrate our app with Quickbooks using IPP v2 API. We would like to retrieve all financial records related to a job. I have been playing around with the API using the API explorer. I see that I have to retrieve the data separately using entities like Bill, Invoice, Journal, etc. However, I didn't seem to find a way to filter the data returned in those entities by job.
Currently, I have to query the entire Bill table for example and to look for JobId under ReimbursableInfo myself. It's not feasible since I have to do the same thing for other tables.
Am I missing something from here?
In docs, JobID is not mentioned as a filterable attribute.
Ref - https://developer.intuit.com/docs/0025_quickbooksapi/0050_data_services/v2/0500_quickbooks_windows/0600_object_reference/bill#Retrieving_Bills_Using_a_Query_Filter
You need to do client side filtering.
For Invoice, you can use JobIdSet attribute.
Ref - https://developer.intuit.com/docs/0025_quickbooksapi/0050_data_services/v2/0500_quickbooks_windows/0600_object_reference/invoice
Thanks
I used Node js for my webapp and I wanted to create a new user with unique email ID and hence that required me to look up the index table but then I realised that the Node Driver doesn't have the index() function as the one in Java( graphDb.index() ). Any work around for such an issue, like using REST or some Cypher connetor
Neo4j just so happens to have a wonderful REST API that should work for your application. The docs for the api are detailed here:
http://docs.neo4j.org/chunked/milestone/rest-api.html.
It even has an entire section devoted to indexing.
I'm looking for an open source news feed engine to use in an app I'm developing.
The engine needs to be able to aggregate news (items) from multiple sources a user is following and also optionally group them by news source or news type. A scalable solution in Java or with Java interface would be great.
I have already developed a very simple one, but I would prefer to use a robust and reliable solution instead.
Do you have any suggestion?
I created a backend for this in Java builing on neo4j: It is independent on the number of users and news sources one follows but depends just linear on the amount of items you want to display
find an explaination how it works and benchmarks with social networks with up to 2 mio users at http://www.rene-pickhardt.de/graphity-an-efficient-graph-model-for-retrieving-the-top-k-news-feeds-for-users-in-social-networks/
There is also the source code available: https://github.com/renepickhardt/graphity-evaluation
Check out Rome
http://rometools.org/
In case you also use .NET, Argotic Syndication Framework is for sure the best
http://argotic.codeplex.com/
Yahoo pipes is a very good rss feed which lets you create your own feed aggregator with custom filters.
Note: Python version for Yahoo pipes Pipe2py.
Aonther offering from yahoo is Daper.
There are few more online tools for creating custom feeds which you might want to look at.
FeedDistiller is a free service for aggregating new feeds by subject,