Finding related content in the Signals Feed automatically

What is it? 

With the help of our Natural Language Processing (NLP) engine, you can automatically detect new content, coming from news, scientific publications, and patents. This powerful solution allows you to save time in the research phase of your innovation process. To see an overview of ITONICS Signals, watch the video below or continue reading the article.

Learn more about:

How does it work?

Saving content from the Signals Feed

Every content element in your application can contain a Signals feed if activated by your Application Owner. Each feed accesses information from our huge data lake, comprising 3000+ top-traffic-ranked news sites, manually curated RSS feeds, and open-access research and patent providers. The incoming content pieces are categorized accordingly under the tabs news, (scientific) publications, and patents.

Each feed itself is generated by using the textual information (title, summary, tags) of the corresponding content element and comparing it with the pieces on the data lake.

To see the Signals feed, go to an element’s detail page and select the Signals tab located in the tab section.

Signals_Tab.png

The resulting content list is displayed on the same page below.

A content widget includes the type of the content piece, i.e., whether it is originating from a news site, scientific publication site, RSS feed, or patent source. It further includes the date of publication by the external source, the source(-URL) itself, and the country of publication in the first line.

Screenshot 2023-09-06 at 15.52.21.png

A content piece itself is displayed by the title, a summary of the content, tags, and companies mentioned in the original source. After clicking on the title, the full content piece will be opened in a new browser tab.

If a content piece is of interest to you, you can:

  1. Flag the Signal from the three dots menu on a content widget: The flagged content will be displayed for all users and allow you and your colleagues to filter for it and keep track of such flagged signals.
  2. Save it as a content element in your application from the three dots menu on a content widget: Clicking the button "Save as..." will open the content create form in a new tab where several fields are pre-filled with the content from the original source, such as the title, abstract, and tags. Once created, the content will be permanently available in your application and you can progress it further, i.e., relating, rating, or commenting on it.

    If the content from the signals feed should be saved in another content category, please get in touch with your Application Owner who will reach out to an ITONICS Customer Innovation Success Manager.

Saving companies from the Signals Feed

If licensed, the Signals Feed can also display companies that have been mentioned most frequently within the incoming content snippets. This will be shown in a dedicated company widget right to the signals feed. You can use this functionality to identify interesting players, competitors, or potential partners in the market that seem to be investing in a certain technology. 

Screenshot 2023-09-06 at 15.56.22.png

The widget also allows you to quickly narrow down the signals feed by clicking on the filter icon next to a company's name. If of interest, it is also possible to save a company profile (fueled by information from Crunchbase) directly in your application by clicking the "+"-icon next to a company's name. 

You find more detailed information about the company widget in this article.

Filtering for content in the Signals Feed

Although a comprehensive list of Signals allows you to analyze the situation from many angles, the option to Filter by specific criteria can help you focus on your needs.

Screenshot_2021-02-12_at_14.05.07.png

You can filter the list of Signals by certain criteria:

  • Use the Search field to type a specific keyword or phrase
  • Select dates (From and To) to focus on the Signals that were published during a specific time (for example, you can choose those that were published just this month)
  • Select a Source to define specific websites you want to focus on
  • Pick a Company that needs to be mentioned on the Signals
  • Select a Country of publication
  • Select the Language(s) of publication
  • Either Show All or Show Flagged Only
  • Sort Signals, for example by their Relevancy

For the Source, Company and Country filters, you can use the Search field in that dropdown to search the options. Once you have found what you are looking for, select entries by checking the boxes for all the options you want to see results for.

The selection will automatically appear. Click Reset Filters to see all results again.

Advanced search for content in the Signals Feed

You can use more complex operators within the search input to precisely filter down the results. Some examples:

  1. Use double quotation marks to make sure the results contain the exact phrase: “machine learning”
  2. The plus sign (+) between two words or phrases will ensure that both are found in the results: “machine learning” + “neural network”
  3. The pipe sign (|) between two words or phrases acts as a logical OR operator: “machine learning” | “artificial intelligence”
  4. The minus sign (-) to exclude words or phrases: -course
  5. Use parentheses to group logical units
  6. Use * as a wildcard to replace any number of characters at the end of a word: “neural network*” will also return results with neural networks or neural networking

These operators can also be combined, for example to find all documents that include the specific phrases “machine learning” or “artificial intelligence” as well as “neural network” but do not contain any mention of courses:

(“machine learning” | “artificial intelligence”) + “neural network*” + -course

 

To recap, here is the full list of available operators:

Operator

Use

+

Logical AND

|

Logical OR (default if nothing is specific between words or phrases)

*

wildcard

(...)

grouping

“...”

exact match

-

negation

How to get good results in the Signals Feed

While the results of the Signals feed are usually relevant for technological topics, accuracy and relevance can be an issue for societal trends. Here are tips on how to improve the quality of the feed:

Dos: Don’ts:
  • Add text that is specific to the underlying topic, for instance, machine learning. The better you describe machine learning, the more relevant results will show.
  • Balance the number of keywords: 
    • 1-3 most relevant keywords in the title
    • 10-25 keywords in the abstract
    • Add tags to your element
  • Less can be more. Adding a lot of text to the summary can be counterproductive and “water down” the search. This can lead to worse quality terms being used for the search.
  • Do not add information that is not topic-specific For instance, adding that you are looking for “companies” will not give you better company results.
  • Do not mix languages as this might confuse the search.

Sources on the Signals Feed

In the Signals feed, we display patents, scientific publications, and news from a multitude of sources.
News (updated several times daily)

  • English news from the 2500 top-trafficked news sites.
  • German news from a top-trafficked news sites.
  • Additionally, news from several hundred manually curated sites across different industries.

RSS (updated several times daily)

  • Several hundred manually curated RSS feeds across different industries.
  • Content curated from RSS feeds is automatically assigned to either the News, (Scientific-) Publications, or Patent category based on the respective type of the signal.

Patents (updated weekly)

  • United States Patent and Trademark Office (USPTO)
  • European Patent Office (EPO)

Open Access Scientific Publications (monthly)

  • OpenAlex

To get a detailed overview of the signals feed, you can also use the search option and check the dropdown field 'sources'

signals_source_list.png

💡 If you have suggestions for sources (especially for news sites and RSS feeds), feel free to contact our support at support@itonics-innovation.com!

Was this article helpful?
3 out of 3 found this helpful