AP Classification Upgrades 2022

Introduction

AP Classification began in 2006 with the goal of funneling news content into newspaper sections. In this major upgrade, we’ve reinvented Classification for the digital news environment.

In the past, the taxonomy sought to describe the world of news content at multiple levels of granularity, giving users numerous entry points to discover content. For every term that matched the content, we output all of its broader categories to funnel content into broader sections. These well-meaning goals lead to far too many classification terms on each item and no clear indication of which were most relevant.

In this release, AP Classification is relaunched with a new, search- and SEO-friendly taxonomy designed to support digital news products. Each tag now includes a relevance score that can be used to filter tags or sort stories in your systems. Only the most salient and relevant tags are returned. We hope this new approach will make it easier for you to discover AP content and to use it in your own production systems.

AP Subject

• The set of subject terms has been reduced to a more salient collection of topics; redundant, ambiguous and low-frequency terms have been removed, and relevant, timely topics have been added, such as Activism, Diversity, Equity and inclusion, and Wellness.

• Only subjects that directly match the content are now returned. Broader topics are no longer returned unless they also match the content directly.

• There are now more top-level subjects; for example, Education and Religion. Other changes to top-level subjects include:

• Arts and entertainment is now Entertainment

• Environment and nature is now Climate and environment

• Government and politics is now Politics

• Social affairs has been removed (some of its children; for example, Education, are now top-level)

For more information, see Top-Level Subjects below.

• Most gendered sports have been removed. For example, Men’s golf and Women’s golf are deprecated, and both are now mapped to the Golf term.

• Many closely-related terms are combined into a single term. For example, the new term Sci-Fi and fantasy combines the content from Science fiction and fantasy movies with Science fiction and fantasy literature. In this case, content about SciFi TV shows, podcasts, and other formats not covered in the past are now included.

• Events that drop out of the news cycle and seem unlikely to return are reviewed and deprecated on a rolling basis.

Top-Level Subjects

AP Geography

• World regions (for example, Eastern Europe, Greater China, West Africa) have been removed.

• Only the following types of locations are now returned: cities, U.S. states, Canadian provinces, Australian states and territories, countries, continents, points of interest and natural features; for example, lakes and mountain ranges.

• Place names are often ambiguous. While we have removed hierarchical roll-up for all classification, we were concerned that the ambiguity of place names could create confusion. To resolve this issue, we have introduced a new parent term property for AP Geography matches only. All AP Geography matches that have a parent term now have this new property.

JSON Example

"place": [
  {
    "code": "7ab435e87ec110048cf8df092526b43e",
    "name": "Tokyo",
    "rels": ["direct"],
    "scheme": http://cv.ap.org/id/,
    "creator": "Machine",
    "parentids": ["661be2f07d5b10048250c076b8e3055c"],
    "parentnames": ["Japan"],
    "relevance": 50,
    "locationtype": {
      "code": "9d26a20b35f0484a891740f8189d4c7b",
      "name": "City"
    },
    "geometry_geojson": {
      "type": "Point",
      "coordinates": [
        139.5813,
        35.61488
      ]
    }
  }
]

NewsML-G2 Example

Location Type

AP Person

• Any person deceased for over two years and deemed not currently newsworthy is now deprecated.

• Sports figures:

• Any athlete or coach inactive for over two years is now deprecated. If a sports figure becomes active again or is newsworthy for any other reason (for example, such as in the case of O.J. Simpson), they will be reactivated by the AP.

• The AP now applies tags for the following:

• All US pro athletes and coaches on the rosters of the NFL, MLB, NHL, NBA, WNBA and MLS teams.

• All NCAA football and basketball athletes and coaches on the rosters of the Power Five Conference and Big East teams, plus a short list of priority players and coaches determined by AP’s sports coverage.

• A priority list of about 100 non-US pro soccer players and coaches determined by AP’s sports coverage.

• Select F1, NASCAR and IndyCar drivers based on AP's sports coverage.

• Select golfers, tennis players, cyclists and other professional athletes based on AP's sports coverage.

• The AP no longer applies tags for complete Olympic rosters for all national teams in favor of a list of priority athletes determined by AP’s sports coverage.

Person Type

AP Person main categories for named individuals include:

AP Organization

• US pro sports. The AP now applies tags for organization terms, including leagues, conferences, divisions and teams, only for NFL, MLB, NHL, NBA, WNBA and MLS.

• Olympic teams. AP’s terms for every country's team in the form "Italy Olympic team" have been deprecated.

• Non-US soccer.

• The AP now applies sports team and league terms only for English Premier League, Bundesliga, Serie A, Ligue 1 and La Liga. Relegated teams have been deprecated.

• National soccer team terms, in the form "Qatar National Soccer Team," now apply only to FIFA World Cup and FIFA Women’s World Cup content.

• NCAA football and men's basketball:

• The AP now applies tags for organization terms for teams of schools that compete in the Division I Football Bowl Subdivision (FBS), plus teams in the Big East, Atlantic 10 and West Coast Conference.

• The AP previously applied separate terms for NCAA men's basketball and football teams, excluding other sports. For example, there were separate terms for "Tennessee Volunteers men's basketball" and "Tennessee Volunteers football." There is now only one term: "Tennessee Volunteers." For cases in which a college had two terms, the URI for the football term is kept with the new name applied, and the basketball term is deprecated. For cases in which only a basketball term existed for the college, that URI is kept and renamed. The new terms apply to athletes representing the school in all varsity-level sports.

• US state governments. The AP previously applied tags for terms for each state government in the form "Nevada state government" and selected individual departments, such as "Nevada State Legislature." Departmental and functional leaf nodes have been deprecated, leaving just "Nevada state government."

• National military terms. The national military terms in the form “Australian armed forces” that AP previously maintained have been deprecated. AP will continue to tag US military branches, but not subordinate units.

AP Company

This change was made in March 2022: AP now tags only companies with primary shares trading on the following five stock exchanges: NYSE (New York Stock Exchange), NASDAQ, TYO (Tokyo Stock Exchange), TSE (Toronto Stock Exchange), XETRA (XETRA - Frankfurt); as well as other select privately-held or publicly-traded companies making the news. If a company trading on one of those five stock exchanges is also traded on other exchanges, the additional stock exchange codes are provided as well.