Bing Provider
The bing provider uses APIs from the Bing search engine to gather news, web and video search results.
| Provider name | bing |
|---|---|
| Type | Polling provider |
Table of Contents
Configuration
The bing provider supports three verticals: Web, News and Video. The configuration is different for each one.
News
| Field | Description | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| vertical | Set to News. | |||||||||||||||
| query | The query to search on Bing. The query can contain any valid query text that the Bing Engine supports. | |||||||||||||||
| market | The news market in which to find news. Example markets:
For the full list of available markets refer to the Bing API Schema Guide. | |||||||||||||||
| news_category | Limit search results to a category. This is optional. Available categories:
|
A maximum number of 15 results can be retrieved via the Bing API.
Video
| Field | Description |
|---|---|
| vertical | Set to Video. |
| query | The query to search on Bing. The query can contain any valid query text that the Bing Engine supports. |
| market | The video market in which to find results. Optional. |
Web
Web search results can also be added to Squirro with the bing provider. But unlike the News and Video verticals, Bing's Web vertical doesn't have support for ordering results by date. As a result Squirro fetches the best 50 results (based on Bing's result score; max 50 results are allowed by the Bing API for retrieval) and checks if that result set contains any previously unseen links. If so, those new links are added to Squirro as items.
| Field | Description |
|---|---|
| vertical | Set to Web. |
| query | The query to search on Bing. The query can contain any valid query text that the Bing Engine supports. |
| market | The web market in which to find results. Optional. |
Configuration Example
This is an example configuration for creating a news tracking configuration.
{
"vertical": "News",
"query": "Squirro",
"market": "en-US"
}
Using the Python SDK a subscription for this could be created with the following code snippet:
client = SquirroClient(None, None, cluster='https://next.squirro.net/')
client.authenticate(refresh_token='293d…a13b')
client.new_subscription(project_id, object_id='default', provider='bing',
config={'vertical': 'News', 'query': 'Squirro', 'market': 'en-US'})
Item Format
No special keywords are added to items created by this provider.
Item Filtering
For results in the News vertical the Bing provider fetches the story from its original site. For this, the story is retrieved and any ads, side columns, etc. are removed.
After this boilerplate removal the news story is then compared against the original query. If at this point it doesn't match the query anymore, the story is discarded. This additional step avoids Bing's many false positives for news search. This can result in fewer stories being delivered to Squirro than you would see in Bing's search engine directly.
To enable content fetching for web search, set the Content Augmentation enrichment's fetch_link_content and enable the Noise Removal removal enrichment (see Built-in enrichments for details).
Licensing
The Bing provider needs to be licensed separately. Because this is a polling provider, the cost have to be considered based on number of configurations and the retry interval.