Entity Extraction
   HOME

TheInfoList



OR:

Named-entity recognition (NER) (also known as (named) entity identification, entity chunking, and entity extraction) is a subtask of information extraction that seeks to locate and classify named entities mentioned in unstructured text into pre-defined categories such as person names (PER),
organizations An organization or organisation ( Commonwealth English; see spelling differences) is an entity—such as a company, or corporation or an institution (formal organization), or an association—comprising one or more people and having a par ...
(ORG), locations (LOC), geopolitical entities (GPE),
vehicles A vehicle () is a machine designed for self-propulsion, usually to transport people, cargo, or both. The term "vehicle" typically refers to land vehicles such as human-powered vehicles (e.g. bicycles, tricycles, velomobiles), animal-powered tr ...
(VEH), medical codes, time expressions, quantities, monetary values, percentages, etc. Most research on NER/NEE systems has been structured as taking an unannotated block of text, such as transducing: into an annotated block of text that highlights the names of entities: In this example, a person name consisting of one token, a two-token company name and a temporal expression have been detected and classified.


Problem


Definition

In the expression '' named entity'', the word ''named'' restricts the task to those entities for which one or many strings, such as words or phrases, stand (fairly) consistently for some referent. This is closely related to rigid designators, as defined by Kripke, although in practice NER deals with many names and referents that are not philosophically "rigid". For instance, the ''automotive company created by Henry Ford in 1903'' can be referred to as ''Ford'' or ''Ford Motor Company'', although "Ford" can refer to many other entities as well (see Ford). Rigid designators include proper names as well as terms for certain biological species and substances, but exclude pronouns (such as "it"; see coreference resolution), descriptions that pick out a referent by its properties (see also
De dicto and de re ''De dicto'' and ''de re'' are two phrases used to mark a distinction in intensional statements, associated with the intensional operators in many such statements. The distinction is used regularly in analytical metaphysics and in philosophy of ...
), and names for kinds of things as opposed to individuals (for example "Bank"). Full named-entity recognition is often broken down, conceptually and possibly also in implementations, as two distinct problems: detection of names, and
classification Classification is the activity of assigning objects to some pre-existing classes or categories. This is distinct from the task of establishing the classes themselves (for example through cluster analysis). Examples include diagnostic tests, identif ...
of the names by the type of entity they refer to (e.g. person, organization, or location). The first phase is typically simplified to a segmentation problem: names are defined to be contiguous spans of tokens, with no nesting, so that "Bank of America" is a single name, disregarding the fact that inside this name, the substring "America" is itself a name. This segmentation problem is formally similar to chunking. The second phase requires choosing an
ontology Ontology is the philosophical study of existence, being. It is traditionally understood as the subdiscipline of metaphysics focused on the most general features of reality. As one of the most fundamental concepts, being encompasses all of realit ...
by which to organize categories of things. Temporal expressions and some numerical expressions (e.g., money, percentages, etc.) may also be considered as named entities in the context of the NER task. While some instances of these types are good examples of rigid designators (e.g., the year 2001) there are also many invalid ones (e.g., I take my vacations in “June”). In the first case, the year ''2001'' refers to the ''2001st year of the Gregorian calendar''. In the second case, the month ''June'' may refer to the month of an undefined year (''past June'', ''next June'', ''every June'', etc.). It is arguable that the definition of ''named entity'' is loosened in such cases for practical reasons. The definition of the term ''named entity'' is therefore not strict and often has to be explained in the context in which it is used. Certain
hierarchies A hierarchy (from Greek: , from , 'president of sacred rites') is an arrangement of items (objects, names, values, categories, etc.) that are represented as being "above", "below", or "at the same level as" one another. Hierarchy is an importan ...
of named entity types have been proposed in the literature. BBN categories, proposed in 2002, are used for '' question answering'' and consists of 29 types and 64 subtypes. Sekine's extended hierarchy, proposed in 2002, is made of 200 subtypes.Sekine's Extended Named Entity Hierarchy
Nlp.cs.nyu.edu. Retrieved on 2013-07-21.
More recently, in 2011 Ritter used a hierarchy based on common Freebase entity types in ground-breaking experiments on NER over
social media Social media are interactive technologies that facilitate the Content creation, creation, information exchange, sharing and news aggregator, aggregation of Content (media), content (such as ideas, interests, and other forms of expression) amongs ...
text.


Difficulties

NER involves ambiguities. The same name can refer to different entities of the same type. For example, "JFK" can refer to the former president or his son. This is basically a reference resolution problem. The same name can refer to completely different types. "JFK" might refer to the airport in New York. "IRA" can refer to
Individual Retirement Account An individual retirement account (IRA) in the United States is a form of pension provided by many financial institutions that provides tax advantages for retirement savings. It is a trust that holds investment assets purchased with a taxpayer's ...
or
International Reading Association The International Literacy Association (ILA), formerly the International Reading Association (IRA), is an international global advocacy and member professional organization that was created in 1956 to improve reading instruction, facilitate dial ...
. This can be caused by
metonymy Metonymy () is a figure of speech in which a concept is referred to by the name of something associated with that thing or concept. For example, the word " suit" may refer to a person from groups commonly wearing business attire, such as sales ...
. For example, "The White House" can refer to an
organization An organization or organisation (English in the Commonwealth of Nations, Commonwealth English; American and British English spelling differences#-ise, -ize (-isation, -ization), see spelling differences) is an legal entity, entity—such as ...
instead of a
location In geography, location or place is used to denote a region (point, line, or area) on Earth's surface. The term ''location'' generally implies a higher degree of certainty than ''place'', the latter often indicating an entity with an ambiguous bou ...
.


Formal evaluation

To evaluate the quality of an NER system's output, several measures have been defined. The usual measures are called precision, recall, and F1 score. However, several issues remain in just how to calculate those values. These statistical measures work reasonably well for the obvious cases of finding or missing a real entity exactly; and for finding a non-entity. However, NER can fail in many other ways, many of which are arguably "partially correct", and should not be counted as complete success or failures. For example, identifying a real entity, but: * with fewer tokens than desired (for example, missing the last token of "John Smith, M.D.") * with more tokens than desired (for example, including the first word of "The University of MD") * partitioning adjacent entities differently (for example, treating "Smith, Jones Robinson" as 2 vs. 3 entities) * assigning it a completely wrong type (for example, calling a personal name an organization) * assigning it a related but inexact type (for example, "substance" vs. "drug", or "school" vs. "organization") * correctly identifying an entity, when what the user wanted was a smaller- or larger-scope entity (for example, identifying "James Madison" as a personal name, when it's part of "James Madison University"). Some NER systems impose the restriction that entities may never overlap or nest, which means that in some cases one must make arbitrary or task-specific choices. One overly simple method of measuring accuracy is merely to count what fraction of all tokens in the text were correctly or incorrectly identified as part of entity references (or as being entities of the correct type). This suffers from at least two problems: first, the vast majority of tokens in real-world text are not part of entity names, so the baseline accuracy (always predict "not an entity") is extravagantly high, typically >90%; and second, mispredicting the full span of an entity name is not properly penalized (finding only a person's first name when his last name follows might be scored as ½ accuracy). In academic conferences such as CoNLL, a variant of the F1 score has been defined as follows: * Precision is the number of predicted entity name spans that line up ''exactly'' with spans in the
gold standard A gold standard is a backed currency, monetary system in which the standard economics, economic unit of account is based on a fixed quantity of gold. The gold standard was the basis for the international monetary system from the 1870s to the ...
evaluation data. I.e. when sub>Person Hans sub>Person Blickis predicted but sub>Person Hans Blickwas required, precision for the predicted name is zero. Precision is then averaged over all predicted entity names. * Recall is similarly the number of names in the gold standard that appear at exactly the same location in the predictions. * F1 score is the
harmonic mean In mathematics, the harmonic mean is a kind of average, one of the Pythagorean means. It is the most appropriate average for ratios and rate (mathematics), rates such as speeds, and is normally only used for positive arguments. The harmonic mean ...
of these two. It follows from the above definition that any prediction that misses a single token, includes a spurious token, or has the wrong class, is a hard error and does not contribute positively to either precision or recall. Thus, this measure may be said to be pessimistic: it can be the case that many "errors" are close to correct, and might be adequate for a given purpose. For example, one system might always omit titles such as "Ms." or "Ph.D.", but be compared to a system or ground-truth data that expects titles to be included. In that case, every such name is treated as an error. Because of such issues, it is important actually to examine the kinds of errors, and decide how important they are given one's goals and requirements. Evaluation models based on a token-by-token matching have been proposed. Such models may be given partial credit for overlapping matches (such as using the Intersection over Union criterion). They allow a finer grained evaluation and comparison of extraction systems.


Approaches

NER systems have been created that use linguistic
grammar In linguistics, grammar is the set of rules for how a natural language is structured, as demonstrated by its speakers or writers. Grammar rules may concern the use of clauses, phrases, and words. The term may also refer to the study of such rul ...
-based techniques as well as
statistical model A statistical model is a mathematical model that embodies a set of statistical assumptions concerning the generation of Sample (statistics), sample data (and similar data from a larger Statistical population, population). A statistical model repre ...
s such as
machine learning Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of Computational statistics, statistical algorithms that can learn from data and generalise to unseen data, and thus perform Task ( ...
. State of the art systems may incorporate multiple approaches. *
GATE A gate or gateway is a point of entry to or from a space enclosed by walls. The word is derived from Proto-Germanic language, Proto-Germanic ''*gatan'', meaning an opening or passageway. Synonyms include yett (which comes from the same root w ...
supports NER across many languages and domains out of the box, usable via a graphical interface and a
Java Java is one of the Greater Sunda Islands in Indonesia. It is bordered by the Indian Ocean to the south and the Java Sea (a part of Pacific Ocean) to the north. With a population of 156.9 million people (including Madura) in mid 2024, proje ...
API. * OpenNLP includes rule-based and statistical named-entity recognition. * SpaCy features fast statistical NER as well as an open-source named-entity visualizer. Hand-crafted grammar-based systems typically obtain better precision, but at the cost of lower recall and months of work by experienced computational linguists. Statistical NER systems typically require a large amount of manually
annotated An annotation is extra information associated with a particular point in a document or other piece of information. It can be a note that includes a comment or explanation. Annotations are sometimes presented in the margin of book pages. For anno ...
training data. Semisupervised approaches have been suggested to avoid part of the annotation effort. In the statistical learning era, NER was usually performed by learning a simple linear regression model on engineered features, then decoded by a bidirectional Viterbi algorithm. Some commonly used features include: * Lexical items: The token itself be labeled. * Stemmed lexical items. * Shape: The orthographic pattern of the target word. For example, all lowercase, all uppercase, initial uppercase, mixed case, uppercase followed by a period (often indicating a middle name), contains hyphen, etc. *
Affixes In linguistics, an affix is a morpheme that is attached to a word stem to form a new word or word form. The main two categories are derivational and inflectional affixes. Derivational affixes, such as ''un-'', ''-ation'', ''anti-'', ''pre-'' et ...
of the target word and surrounding words. *
Part of speech In grammar, a part of speech or part-of-speech ( abbreviated as POS or PoS, also known as word class or grammatical category) is a category of words (or, more generally, of lexical items) that have similar grammatical properties. Words that are ...
of the word. * Whether the word appears in one or more named entity lists (gazetteers). * Words and/or n-grams occurring in the surrounding context. A
gazetteer A gazetteer is a geographical dictionary or wikt:directory, directory used in conjunction with a map or atlas.Aurousseau, 61. It typically contains information concerning the geographical makeup, social statistics and physical features of a co ...
is a list of names and their types, such as "General Electric". It can be used to augment any system for NER. They had been often used in the era of statistical machine learning. Many different classifier types have been used to perform machine-learned NER, with
conditional random field Conditional random fields (CRFs) are a class of statistical modeling methods often applied in pattern recognition and machine learning and used for structured prediction. Whereas a classifier predicts a label for a single sample without consi ...
s being a typical choice.
Transformers ''Transformers'' is a media franchise produced by American toy company Hasbro and Japanese toy company Tomy, Takara Tomy. It primarily follows the heroic Autobots and the villainous Decepticons, two Extraterrestrials in fiction, alien robot fac ...
features token classification using deep learning models.


History

Early work in NER systems in the 1990s was aimed primarily at extraction from journalistic articles. Attention then turned to processing of military dispatches and reports. Later stages of the automatic content extraction (ACE) evaluation also included several types of informal text styles, such as
weblog A blog (a Clipping (morphology), truncation of "weblog") is an informational website consisting of discrete, often informal diary-style text entries also known as posts. Posts are typically displayed in Reverse chronology, reverse chronologic ...
s and text transcripts from conversational telephone speech conversations. Since about 1998, there has been a great deal of interest in entity identification in the
molecular biology Molecular biology is a branch of biology that seeks to understand the molecule, molecular basis of biological activity in and between Cell (biology), cells, including biomolecule, biomolecular synthesis, modification, mechanisms, and interactio ...
,
bioinformatics Bioinformatics () is an interdisciplinary field of science that develops methods and Bioinformatics software, software tools for understanding biological data, especially when the data sets are large and complex. Bioinformatics uses biology, ...
, and medical
natural language processing Natural language processing (NLP) is a subfield of computer science and especially artificial intelligence. It is primarily concerned with providing computers with the ability to process data encoded in natural language and is thus closely related ...
communities. The most common entity of interest in that domain has been names of
gene In biology, the word gene has two meanings. The Mendelian gene is a basic unit of heredity. The molecular gene is a sequence of nucleotides in DNA that is transcribed to produce a functional RNA. There are two types of molecular genes: protei ...
s and gene products. There has been also considerable interest in the recognition of chemical entities and drugs in the context of the CHEMDNER competition, with 27 teams participating in this task. In 2001, research indicated that even state-of-the-art NER systems were brittle, meaning that NER systems developed for one domain did not typically perform well on other domains. Considerable effort is involved in tuning NER systems to perform well in a new domain; this is true for both rule-based and trainable statistical systems. As of 2007, state-of-the-art NER systems for English produce near-human performance. For example, the best system entering MUC-7 scored 93.39% of
F-measure In statistics, statistical analysis of binary classification and information retrieval systems, the F-score or F-measure is a measure of predictive performance. It is calculated from the Precision (information retrieval), precision and Recall (in ...
while human annotators scored 97.60% and 96.95%.


Current challenges

Despite high F1 numbers reported on the MUC-7 dataset, the problem of named-entity recognition is far from being solved. The main efforts are directed to reducing the annotations labor by employing
semi-supervised learning Weak supervision (also known as semi-supervised learning) is a paradigm in machine learning, the relevance and notability of which increased with the advent of large language models due to large amount of data required to train them. It is charact ...
, robust performance across domains and scaling up to fine-grained entity types. In recent years, many projects have turned to
crowdsourcing Crowdsourcing involves a large group of dispersed participants contributing or producing goods or services—including ideas, votes, micro-tasks, and finances—for payment or as volunteers. Contemporary crowdsourcing often involves digit ...
, which is a promising solution to obtain high-quality aggregate human judgments for supervised and semi-supervised machine learning approaches to NER. Another challenging task is devising models to deal with linguistically complex contexts such as Twitter and search queries. There are some researchers who did some comparisons about the NER performances from different statistical models such as HMM (
hidden Markov model A hidden Markov model (HMM) is a Markov model in which the observations are dependent on a latent (or ''hidden'') Markov process (referred to as X). An HMM requires that there be an observable process Y whose outcomes depend on the outcomes of X ...
), ME ( maximum entropy), and CRF (
conditional random field Conditional random fields (CRFs) are a class of statistical modeling methods often applied in pattern recognition and machine learning and used for structured prediction. Whereas a classifier predicts a label for a single sample without consi ...
s), and feature sets. And some researchers recently proposed graph-based semi-supervised learning model for language specific NER tasks. A recently emerging task of identifying "important expressions" in text and cross-linking them to Wikipedia can be seen as an instance of extremely fine-grained named-entity recognition, where the types are the actual Wikipedia pages describing the (potentially ambiguous) concepts. Below is an example output of a Wikification system: Michael Jordan is a professor at Berkeley Another field that has seen progress but remains challenging is the application of NER to
Twitter Twitter, officially known as X since 2023, is an American microblogging and social networking service. It is one of the world's largest social media platforms and one of the most-visited websites. Users can share short text messages, image ...
and other microblogs, considered "noisy" due to non-standard orthography, shortness and informality of texts. NER challenges in English Tweets have been organized by research communities to compare performances of various approaches, such as bidirectional LSTMs, Learning-to-Search, or CRFs.


See also

*
Controlled vocabulary A controlled vocabulary provides a way to organize knowledge for subsequent retrieval. Controlled vocabularies are used in subject indexing schemes, subject headings, thesauri, taxonomies and other knowledge organization systems. Controlled v ...
* Coreference resolution *
Entity linking In natural language processing, Entity Linking, also referred to as named-entity disambiguation (NED), named-entity recognition and disambiguation (NERD), named-entity normalization (NEN), or Concept Recognition, is the task of assigning a unique ...
(aka named entity normalization, entity disambiguation) * Information extraction *
Knowledge extraction Knowledge extraction is the creation of knowledge from structured ( relational databases, XML) and unstructured (text, documents, images) sources. The resulting knowledge needs to be in a machine-readable and machine-interpretable format and must ...
*
Onomastics Onomastics (or onomatology in older texts) is the study of proper names, including their etymology, history, and use. An ''alethonym'' ('true name') or an ''orthonym'' ('real name') is the proper name of the object in question, the object of onom ...
* Record linkage * Smart tag (Microsoft)


References

* * {{Authority control Computational linguistics Tasks of natural language processing