On this page, we’ll provide you with the newest information and latest release notes!

“See what we’ve been up to!

On this page, we’ll provide you with the newest information about upcoming features and platform enhancements as soon as they’re released. 
Do you want to get informed? Sign-up for our Developer Newsletter here.

Last update: 1st August 2019


What’s next?

The following main topics we’ve on our schedule for the next few weeks/month. Don’t expect that these topics will be released in the given order. There might be changes, customer projects and unexpected issues which may affect this list. If you’ve suggestions, wishes, and needs, you’re invited to leave a feature request or send us your feedback to support@hyscore.io.

Some of these bigger topics will be split into tiny pieces to enhance the service step by step. 

Outlook (not prioritized):

  • Search API improvement:  adjustable search pattern accuracy (wide or very accurate)
  • Demographics analysis (Gender/Age) – cookie-less – based on content and context (alternative GDPR/DSVGO conform processing for users given no consent to track)
  • Ongoing: API Improvement: Optimization the determination of specific website types (e.g. eCommerce, Application / Services, etc.)
  • Ongoing: API Improvement: Optimization of the analysis related to special interest websites (e.g. legal and law websites, improve the disambiguation algorithms)
  • Crawler enhancements and performance updates (several)
  • Sentiment analysis enhancements for additional languages 
  • Development and setup of customer individual projects and endpoints.
  • Several additional partner and platform integrations such as AWS Marketplace (done), SAP Hybris Cloud Platform, etc.
  • New languages in preparation e.g. Greek
  • … more to follow.

In testing or BETA, will be released soon. Estimated time of Arrival (ETA):

  • hyScore|search API / Cockpit: User Interface (UI) to enable the search for a human user (e.g. for manual research) – additional product.
  • API Feature: Gender and age information based on the contextual analysis and environment of a website – no ETA yet. 

Latest releases

  • Release v3.4 (5th October 2019)

    • New: hyScore|segments
      Cockpit: User Interface (UI) –  “Segment Builder” to create custom segments based on keyword/entities and phrases.
  • Release v3.5 (10th September 2019)

    • New: hyScore|brandrisk, helps to identify critical content on journalistic websites, which are not optimally suited as an environment for an advertising campaign. The hyScore|brandrisk indicator identifies reports on catastrophes, accidents, and crimes.
  • Release v3.3 (18th July 2019)

    • New: added sentiment analysis in Spanish language 
  • API Release v3.2.02 (23th May 2019)

    • New: “Time-to-Read” extension for hyScore|analyze – average reading time for the given text/article in MM:SS.
    • Beta “Confidence”-Parameter for keyword extraction and named entity recognition, a value between “0.1” (low) and “1” (high) which influences the results of the output. v3/text endpoint only!
  • API Release v3.2.01 (5th February 2019)

    • New: Reset functionality for URI analysis 
    • Improvement: Some improvements to the entire backend system to handle huge loads.
  • API Release v3.2.01 (9th January 2019)

    • New: Search API improvement: New operator “and” for keywords and surface_forms of a keyword. Key/values: “operator”:”and”. The default value is “or”
  • API Release v3.2 (25th November 2018)

    • New: Sentiment API improvement (managed service): Keyword and Phrase sentiment analysis on specific topics for e.g. user feedback, questionnaires and more to analyze for example user experience.
    • Improvement: API Feature(managed service): Custom taxonomy/categorization can be trained on individual needs and use-cases 
  • API Release v3.1.51 (9th November 2018)

    • New: Search API improvement: exclusion criteria filters for keywords and categories. See: Search API documentation
    • Improvement:  API Endpoint v3/text has now two new parameters to send headlines and meta keywords to get considered in the NLP analysis.
    • Improvement: API Feature: Custom taxonomy/categorization can be trained on individual needs and use-cases (managed service).
  • API Release v3.1.5 (30th October 2018)

    • New: API Search Endpoint for Semantic and Simple Search to find the similar and comparable content of any provided url or text in our public URL index. See: Search API documentation
  • API Release v3.1.42 (2nd September 2018)

    • Improvement: API Improvement: Force specific language and analysis per TLD (in testing)
    • Improvement: Automatically block (all known to hyScore) adserver URLs on the first request (in testing)
  • hyScore UI: Cockpit v1.1 (1th September 2018)

    • Improvement: Create/Edit/Delete User in hyScore.io Cockpit, so that you as an account owner is able to manage users for your account.
    • Improvement: User rights management for an account owner.
  • API Release v3.1.41 (13th August 2018)

    • Improvement: API v3: We’ve added two new status codes & messages to the hyScore.io API response:
      • Incomplete (Code: 9) : Analysis in progress 
      • Error (Code: 98) : Blockey by hyScore
  • API Release v3.1.3 (15th June 2018)

    • New: Sentiment Analysis support for French (FR) added
    • Improvement: further algorithm optimization for better results.
  • API available on AWS Marketplace (21st May 2018)

    • New: The hyScore API is now available at AWS Marketplace
  • API Release v3.1.2 (2nd May 2018)

    • New: Language Support Thai (TH) added
    • New: Language Support Vietnamese (VN) added
    • Improvement: algorithm optimization for better results.
  • API Release v3.1.1 (20th April 2018)

    • New: API endpoint https://api.hyscore.io/v3/text for TEXT analysis only
    • Improvement: algorithm optimization for better results.
    • Update: API Version 3 documentation 
  • API Release v3 (26th March 2018)

    • New: API endpoint https://api.hyscore.io/v3/url for website analysis
    • Improvement: algorithm optimization for better results.
    • Improvement: new structured JSON response.
    • Improvement: enhanced customData parameter into key value store.
    • Improvement: image will be always requested
    • Improvement: invented message in JSON response
    • Improvement: beside the metaKeywords we provide the website description
    • Feature: provide article headlines in JSON result incl. and in descending order.
    • Update: API Version 3 documentation 
  • API Release v2.5.2 (9th February 2018)

    • Improvement: Crawler technology improved.
    • Improvement: More detailed status messages in the JSON response.
    • Improvement: invented status codes in the status messages (see documentation)
    • Improvement: A more granular reporting based upon the new status messages.
    • Feature: added two languages – Norwegian (NO) and Finish (FI)
  • API Release v2.5.1 (29th November 2017)

    • Improvement: Algorithm optimization for better-weighted keyword results.
    • Improvement: surface_forms of a keyword is now presented in a [ list ] in the JSON response.

API Release v2.5 (9th November 2017)

  • Improvement: New version of the hyScore.io contextual data crawler. For more information take a look at https://hyscore.io/crawler
  • Feature: “surface_form” – Keyword. Beside the normalized, weighted keywords we’ll provide the keyword in the form a user would read it on-page so that you can decide for yourself if you want to use the normalized or “surface_form” of a keyword.
  • API Release v2.4 (14th October 2017)

    • Improvement: The API allows now allows you to specify which part of the website should be used for the text analysis/keyword extraction.
    • Improvement: Common system optimization to improve performance and scalability.
    • Improvement: Some minor modification to the keyword extraction algorithm which will provide better results.

API Release v2.3 (10th October 2017)

  • Improvement: API provides now more detailed “status” and “message” in the API api.hyscore.io/v2/ JSON result instead of permanently showing only the message: “No data yet!”. See API documentation in the chapter ” for further information.
  • Removed: bulk – Endpoint; you should use API endpoint “api.hyscore.io/v2/” only!

API Release v2.2 (28th September 2017)

  • Feature: Sentiment detection and analysis enhancement – available in German (DE) and EN (English). New result type:
    “sentiment”: 0.30833333333333335
    Sentiment has a value between +1 (perfect), 0 (neutral) and -1 (poor, negative). A tiny demo of the functionality itself you can demo here.
  • Improvement: In case a Target URL is not reachable (DNS Error, multi-chain-redirects,
    etc.) you’ll get new type of response:
    “error”: “Target not reachable. Check URL” }
    That means the URL or website you want to analyze is not available or reachable. That’s counted as a valid result too. Please keep in mind that you’ve to pre-check the URLs you send to our API service if you want to make the analysis more efficient.
  • Update of IAB Taxonomy mapping (29th August 2017)

    The Latest IAB hyScore category mapping file (*.zip-file) contains the latest CSV-File with hyScore’s category mapping (IAB category mapping). Last Update: 29th August 2017

API Release v2.1 (22nd August 2017)

  • Improvement: Several improvements to the hyScore contextual extraction and website categorization algorithms.
  • Feature: Language support Chinese (CN)
  • Feature: Language support Czech (CZ)
  • Feature: Language support Polish (PL)
  • Feature: Language support Hindi (IN)

API Release v2.0 (27th July 2017)

  • Improvement: Several improvements to the hyScore contextual extraction algorithms.
  • Feature: By using API v2 we now show multiple categories and their weight (max. weight: 50), including the related IAB mapping (category/code) in the JSON results.
  • Feature: Image URL extraction; show URL of the main article image if available. Either chosen automatically or via the <i>imagedivid</i> parameter. Instead of letting the system automatically choose the appropriate image, you can enter the id of the image div that should be used.

Documentation Update (3rd July 2017)

  • Update: API documentation, added a couple of status messages and extend some descriptions.
  • Update: FAQ, added some additional frequently asked questions and answers to the list, such as Service, Reports, Payment, etc.

API Release v1.1.6 (28th June 2017)

  • Improvement: IAB category mapping update and enhancement. Several improvements to the hyScore categorization algorithms.
  • Feature: text length (INT) provides the number of chars in the analyzed text. Another indicator related to the size of the text and the quality of the analysis. Fewer chars, e.g. a picture gallery with one sentence might lead to low keyword scores and are less valid.

API Release v1.1.5 (27th June 2017)

  • Improvement: IAB category mapping update and enhancement. Several improvements to the hyScore categorization algorithms.

API Release v1.1.4 (20th June 2017)

  • Improvement: Extension of more feasible status messages to enhance the JSON response and provide some more information about possible issues to analyze an url
  • Improvement: Enhancement of the performance of the main system. An issue with a higher latency combined with a qualified category response is solved and much faster as before.

API Release v1.1.3 (31th May 2017)

  • Improvement: Show the official IAB content taxonomy (category/code) in the JSON response for the requested URL/Website.
  • Improvement: Further optimization of algorithms and performance of the main system.

API Release v1.1.2 (17th May 2017)

  • Bugfix: “numberOfKeywords” parameter is now working properly and shows the number of scored & weighted keywords you requested (default setting = 5).
  • Improvement: “getMeta” parameter; Type = boolean (Default=False). Switch specifying (getMeta=True) if the system should return in addition the extracted meta keywords of the requested URL if available and set by the website publisher.
  • API Release v1.1.1 (15th May 2017)

    • Improvement: In the JSON response you’ll get any provided meta keywords (if provided by the website/URL) in addition to the scored and weighted keywords. Info: we rolled back the changes of release v1.02 – start pages of a TLD show now both types.

API Release v1.1.0 (11th May 2017)

  • Improvement: Switch to part-serverless architecture.
    Parts of the API were switched to serverless to decrease reaction time and improve scaling.
  • Improvement: Optimization of the API response times to an average of < 100ms.
  • New I/O: Added customData field to increase integration flexibility. Added getText switch to activate/deactivate extracted text.
  • Improvement: Optimization of channel categorization / websites.

API Release v1.0.2 (20th April 2017)

  • Improvement: specific handling of start- and overview pages
    If the hyScore API receives a starting-page URL instead of a content URL, it will only return the site category, any provided meta keywords, the site language, url, tld and uuid. Due to the thematically diverse, fluctuating nature of these sites properly weighted keywords and text content are omitted.

API Release v1.0.1 (18th April 2017)

  • Improvement:  deliver “always” an API response within < 200ms
    If an analysis of a requested website URL or text takes longer than > 200ms the API tells you “No Data yet” in the request body of the JSON response. If the analysis is finished and we’ve determined the scored & weighted keywords, category, etc., the next or one of the next requests will return a proper result of the same URL in the JSON response file. This improvement gives you the security that the API always answers in a particular time window.

Started public release list (18th April 2017)