Harnessing the Power of News APIs for Risk Assessment and Management

The prevailing metaphor for a difficult search is “like trying to find a needle in a haystack.” And yet when it comes to identifying and investigating risk events using the amount of information produced and consumed each day, the problem is more difficult: It’s like finding a particular piece of straw in a haystack. It’s all hay, in enormous amounts—but only some of it is relevant to the task at hand.

Millions of news articles are published each day—in English, and countless other languages—making it an impossible task for a team of humans to keep up with them all. This can result in missed early warning signals that could lead to catastrophic consequences for governmental agencies and multinational corporations alike. Any entity with an interest in identifying and mitigating risk is at the mercy of the gargantuan scope of global news.

An unwieldy amount of data

The difficulties of working with such a large influx of news include the following:

• Accessing global news in a timely manner is complicated. It’s difficult to source, and difficult to process

• Manually searching and reading news content is impossible at that scale; any signal is drowned out by the volume of noise

• Translation is also an issue; there are millions of news articles published in native languages, requiring translation to English before they can be evaluated

• It’s difficult to integrate news data into workflows; the stories are written to be read, not aggregated

Despite these difficulties, global news is an incredibly powerful data set to uncover and identify critical insights in real time. To maximize their effectiveness, government departments must find a way of harnessing that data for use in their day-to-day operations.

Processing the flood of information

One of the most effective ways to access and process this deluge of information is to use an AI-powered News API. A News API is an interface designed to obtain machine-readable data automatically from different websites—in this case, news services.

Quantexa News Intelligence (QNI) is one such interface. It aggregates around 1.3 million news articles per day from over 90,000 publishers worldwide—everything from CNN and Le Monde to a host of niche blogs that go into the kind of granular detail that general-interest publications avoid.

Advanced AI, in particular natural language processing (NLP), is applied to every article as it is ingested, transforming the world's news from unstructured data (i.e., sentences like this one) into structured data (e.g., publisher_name: Reuters). With 26 datapoints added to every article, QNI's News API enables the most powerful search and filtering of global news on the market. This can enable government departments to streamline business processes and open up a whole new data set, rich with actionable insights.

Here are the benefits of using an AI-powered news API such as QNI.

• Instant access to a critical dataset. QNI handles all of this automatically, meaning there’s no more need to build or maintain programs to scrape data from the internet

• Reduction in repetitive tasks. QNI reads and tags 14 articles per second, on average, which would be impossible to do manually, and even a small subset would require huge amounts of worker hours.

• Expanded risk coverage. QNI coverage includes all countries, and automatically translates 16 languages into English—Chinese, Russian, Arabic, French, Spanish, Dutch, Turkish, and more

• Streamlined data workflow. API integration means the aggregated data is delivered through one easy-to-use interface in a machine-readable format, delivered directly to the workbench.

Say, for instance, your organization was interested in civil rights infringements. In a traditional search, you could carry out a keyword search of exactly "civil rights infringements." However, those keywords will have to appear in the news article exactly like that—otherwise an article will not appear in the results. This will mean you will miss out on relevant articles. Refining your keyword search to just "civil rights," however, will open the search too wide and return lots of irrelevant results. You could string together long, complex keyword searches with Boolean operators, but this is time-consuming, and prone to inaccuracies.

Filtering the signal from the noise

Using QNI's search capabilities makes finding relevant news articles significantly faster and more accurate. For example, you can use our industry-leading category taxonomy to narrow your search while still casting the net wide enough to catch all relevant articles. In this case, you could use the following categories: "civil rights law," "racial and ethnic discrimination," "civil rights movement," "job discrimination," and many more. To refine further, you could focus on specific industry tags (over 1,500 of them), or entities (over 5.6 million recognized). Or you can filter by language, publisher, or sentiment. And, of course, you can fine-tune the search using keywords.

You can choose to receive your results as stories—a format most like a traditional search engine. You can also receive information as a time series, allowing you to track changes in quantitative values contained in stories over time. Other options are to receive information as a cluster (a cluster of stories about a subject with the same metadata), or as a trend (indicating which elements of your search are rising and falling over time).

Ultimately, the QNI is a powerful, versatile tool that can provide you with the information you need, in a format that best suits your organization, and its processing tools. Sound interesting? You can try Quantexa News Intelligence for your own organization by signing up for a free 14-day trial.

Risk Management

5 Reasons to Make Your Respondent Bank Risk Assessment More Dynamic