Look up help in Wiktionary, the free dictionary.

Help is any form of assistance.

Help may also refer to:

In computing:

• Help (command), a command that invokes documentations and helping information
• Online help, documentation that accompanies computer application software
  • Context-sensitive help, a kind of online help that is obtained from a specific point in the state of the software, providing help for the situation that is associated with that state
• Help desk, a point of contact between product users and technical support

In films and television:

• Help! (film), a 1965 film directed by Richard Lester starring The Beatles
• Help (film), a 2010 Bollywood horror film
• Help (Australian TV series), a documentary series
• Help (UK TV series), a comedy series
• Help (Dutch TV series), a drama series
• The Help (TV series), a 2004 American series
• H.E.L.P., an American TV drama series
• The Help (film), a film adaptation of the Stockett novel
• "Help" (Buffy episode)
• "Help", an episode of The Protector
• "Help!", an episode of Samantha Who?

In literature:

• Help! (magazine)
• Help!! (manga)
• The Help, a historical novel by Kathryn Stockett

In music:

• Help! (album), the 1965 Beatles album which includes songs from the film
  • "Help!" (song), the number one single and theme song of the film
• The Help Album, a musical compilation album to benefit the War Child charity
• "Help" (song), a song by Lloyd Banks

Acronyms:

• Heat escape lessening position, a way to position oneself to reduce heat loss in cold water
• Higher Education Loan Programme, an Australian government loan scheme for students
• Hydrologic Evaluation of Landfill Performance, a model developed by the United States Environmental Protection Agency
• United States Senate HELP Committee or United States Senate Committee on Health, Education, Labor, and Pensions

See also[link]

• Emergency, a situation which poses a dangerous risk of health, life, property, or environment
  • Emergency telephone numbers around the world
  • Hue and cry, a shouted command to arrest a felon
• Distress signal, a message from a person or a craft such as a ship or airliner to inform others it is in difficulty
  • Mayday (distress signal)
  • SOS
• Assistance (disambiguation)
• Help me (disambiguation)
• HELLP Syndrome, a medical condition affecting pregnant women.
• All pages beginning with "Help"
• All pages with titles containing "Help"

This disambiguation page lists articles associated with the same title. If an internal link led you here, you may wish to change the link to point directly to the intended article.

In computational linguistics, word-sense disambiguation (WSD) is an open problem of natural language processing, which governs the process of identifying which sense of a word (i.e. meaning) is used in a sentence, when the word has multiple meanings (polysemy). The solution to this problem impacts other computer-related writing, such as discourse, improving relevance of search engines, anaphora resolution, coherence, inference et cetera.

Research has progressed steadily to the point where WSD systems achieve sufficiently high levels of accuracy on a variety of word types and ambiguities. A rich variety of techniques have been researched, from dictionary-based methods that use the knowledge encoded in lexical resources, to supervised machine learning methods in which a classifier is trained for each distinct word on a corpus of manually sense-annotated examples, to completely unsupervised methods that cluster occurrences of words, thereby inducing word senses. Among these, supervised learning approaches have been the most successful algorithms to date.

Current accuracy is difficult to state without a host of caveats. In English, accuracy at the coarse-grained (homograph) level is routinely above 90%, with some methods on particular homographs achieving over 96%. On finer-grained sense distinctions, top accuracies from 59.1% to 69.0% have been reported in recent evaluation exercises (SemEval-2007, Senseval-2), where the baseline accuracy of the simplest possible algorithm of always choosing the most frequent sense was 51.4% and 57%, respectively.

About[link]

A disambiguation process requires two strict things: a dictionary to specify the senses which are to be disambiguated and a corpus of language data to be disambiguated (in some methods, a training corpus of language examples is also required). WSD task has two variants: "lexical sample" and "all words" task. The former comprises disambiguating the occurrences of a small sample of target words which were previously selected, while in the latter all the words in a piece of running text need to be disambiguated. The latter is deemed a more realistic form of evaluation, but the corpus is more expensive to produce because human annotators have to read the definitions for each word in the sequence every time they need to make a tagging judgement, rather than once for a block of instances for the same target word.

To give a hint how all this works, consider two examples of the distinct senses that exist for the (written) word "bass":

1. a type of fish
2. tones of low frequency

and the sentences:

1. I went fishing for some sea bass.
2. The bass line of the song is too weak.

To a human, it is obvious that the first sentence is using the word "bass (fish)", as in the former sense above and in the second sentence, the word "bass (instrument)" is being used as in the latter sense below. Developing algorithms to replicate this human ability can often be a difficult task, as is further exemplified by the implicit equivocation between "bass (sound)" and "bass" (musical instrument).

History[link]

WSD was first formulated as a distinct computational task during the early days of machine translation in the 1940s, making it one of the oldest problems in computational linguistics. Warren Weaver, in his famous 1949 memorandum on translation,^[1] first introduced the problem in a computational context. Early researchers understood the significance and difficulty of WSD well. In fact, Bar-Hillel (1960) used the above example to argue^[2] that WSD could not be solved by "electronic computer" because of the need in general to model all world knowledge.

In the 1970s, WSD was a subtask of semantic interpretation systems developed within the field of artificial intelligence, but since WSD systems were largely rule-based and hand-coded they were prone to a knowledge acquisition bottleneck.

By the 1980s large-scale lexical resources, such as the Oxford Advanced Learner's Dictionary of Current English (OALD), became available: hand-coding was replaced with knowledge automatically extracted from these resources, but disambiguation was still knowledge-based or dictionary-based.

In the 1990s, the statistical revolution swept through computational linguistics, and WSD became a paradigm problem on which to apply supervised machine learning techniques.

The 2000s saw supervised techniques reach a plateau in accuracy, and so attention has shifted to coarser-grained senses, domain adaptation, semi-supervised and unsupervised corpus-based systems, combinations of different methods, and the return of knowledge-based systems via graph-based methods. Still, supervised systems continue to perform best.

Difficulties[link]

Differences between dictionaries[link]

One problem with word sense disambiguation is deciding what the senses are. In cases like the word bass above, at least some senses are obviously different. In other cases, however, the different senses can be closely related (one meaning being a metaphorical or metonymic extension of another), and in such cases division of words into senses becomes much more difficult. Different dictionaries and thesauruses will provide different divisions of words into senses. One solution some researchers have used is to choose a particular dictionary, and just use its set of senses. Generally, however, research results using broad distinctions in senses have been much better than those using narrow ones.^[3][4] However, given the lack of a full-fledged coarse-grained sense inventory, most researchers continue to work on fine-grained WSD.

Most research in the field of WSD is performed by using WordNet as a reference sense inventory for English. WordNet is a computational lexicon that encodes concepts as synonym sets (e.g. the concept of car is encoded as { car, auto, automobile, machine, motorcar }). Other resources used for disambiguation purposes include Roget's Thesaurus^[5] and Wikipedia.^[6]

Part-of-speech tagging[link]

In any real test, part-of-speech tagging and sense tagging are very closely related with each potentially making constraints to the other. And the question whether these tasks should be kept together or decoupled is still not unanimously resolved, but recently scientists incline to test these things separately (e.g. in the Senseval/SemEval competitions parts of speech are provided as input for the text to disambiguate).

It is instructive to compare the word sense disambiguation problem with the problem of part-of-speech tagging. Both involve disambiguating or tagging with words, be it with senses or parts of speech. However, algorithms used for one do not tend to work well for the other, mainly because the part of speech of a word is primarily determined by the immediately adjacent one to three words, whereas the sense of a word may be determined by words further away. The success rate for part-of-speech tagging algorithms is at present much higher than that for WSD, state-of-the art being around 95% accuracy or better, as compared to less than 75% accuracy in word sense disambiguation with supervised learning. These figures are typical for English, and may be very different from those for other languages.

Inter-judge variance[link]

Another problem is inter-judge variance. WSD systems are normally tested by having their results on a task compared against those of a human. However, while it is relatively easy to assign parts of speech to text, training people to tag senses is far more difficult.^[7] While users can memorize all of the possible parts of speech a word can take, it is often impossible for individuals to memorize all of the senses a word can take. Moreover, humans do not agree on the task at hand – give a list of senses and sentences, and humans will not always agree on which word belongs in which sense.^[8]

Thus, a computer cannot be expected to give better performance on such a task than a human (indeed, since the human serves as the standard, the computer being better than the human is incoherent),^{[citation needed]} so the human performance serves as an upper bound. Human performance, however, is much better on coarse-grained than fine-grained distinctions, so this again is why research on coarse-grained distinctions^[9][10] has been put to test in recent WSD evaluation exercises.^[3][4]

Common sense[link]

Some AI researchers like Douglas Lenat argue that one cannot parse meanings from words without some form of common sense ontology. For example, comparing these two sentences:

• "Jill and Mary are sisters." – (they are sisters of each other).
• "Jill and Mary are mothers." – (each is independently a mother).

To properly identify senses of words one must know common sense facts.^[11] Moreover, sometimes the common sense is needed to disambiguate such words like pronouns in case of having anaphoras or cataphoras in the text.

Sense inventory and algorithms' task-dependency[link]

A task-independent sense inventory is not a coherent concept^[12]: each task requires its own division of word meaning into senses relevant to the task. For example, the ambiguity of 'mouse' (animal or device) is not relevant in English-French machine translation, but is relevant in information retrieval. The opposite is true of 'river', which requires a choice in French (fleuve 'flows into the sea', or rivière 'flows into a river').

Also, completely different algorithms might be required by different applications. In machine translation, the problem takes the form of target word selection. Here, the "senses" are words in the target language, which often correspond to significant meaning distinctions in the source language ("bank" could translate to the French "banque" -- that is, 'financial bank' or "rive" -- that is, 'edge of river'). In information retrieval, a sense inventory is not necessarily required, because it is enough to know that a word is used in the same sense in the query and a retrieved document; what sense that is, is unimportant.

Discreteness of senses[link]

Finally, the very notion of "word sense" is slippery and controversial. Most people can agree in distinctions at the coarse-grained homograph level (e.g., pen as writing instrument or enclosure), but go down one level to fine-grained polysemy, and disagreements arise. For example, in Senseval-2, which used fine-grained sense distinctions, human annotators agreed in only 85% of word occurrences.^[13] Word meaning is in principle infinitely variable and context sensitive. It does not divide up easily into distinct or discrete sub-meanings.^[14] Lexicographers frequently discover in corpora loose and overlapping word meanings, and standard or conventional meanings extended, modulated, and exploited in a bewildering variety of ways. The art of lexicography is to generalize from the corpus to definitions that evoke and explain the full range of meaning of a word, making it seem like words are well-behaved semantically. However, it is not at all clear if these same meaning distinctions are applicable in computational applications, as the decisions of lexicographers are usually driven by other considerations. Recently, a task – named lexical substitution – has been proposed as a possible solution to the sense discreteness problem.^[15] The task consists of providing a substitute for a word in context that preserves the meaning of the original word (potentially, substitutes can be chosen from the full lexicon of the target language, thus overcoming discreteness).

Approaches and methods[link]

As in all natural language processing, there are two main approaches to WSD – deep approaches and shallow approaches.

Deep approaches presume access to a comprehensive body of world knowledge. Knowledge, such as "you can go fishing for a type of fish, but not for low frequency sounds" and "songs have low frequency sounds as parts, but not types of fish", is then used to determine in which sense the word is used. These approaches are not very successful in practice, mainly because such a body of knowledge does not exist in a computer-readable format, outside of very limited domains.^[16] However, if such knowledge did exist, then deep approaches would be much more accurate than the shallow approaches.^{[citation needed]} Also, there is a long tradition in computational linguistics, of trying such approaches in terms of coded knowledge and in some cases, it is hard to say clearly whether the knowledge involved is linguistic or world knowledge. The first attempt was that by Margaret Masterman and her colleagues, at the Cambridge Language Research Unit in England, in the 1950s. This attempt used as data a punched-card version of Roget's Thesaurus and its numbered "heads", as an indicator of topics and looked for repetitions in text, using a set intersection algorithm. It was not very successful,^[17] but had strong relationships to later work, especially Yarowsky's machine learning optimisation of a thesaurus method in the 1990s.

Shallow approaches don't try to understand the text. They just consider the surrounding words, using information such as "if bass has words sea or fishing nearby, it probably is in the fish sense; if bass has the words music or song nearby, it is probably in the music sense." These rules can be automatically derived by the computer, using a training corpus of words tagged with their word senses. This approach, while theoretically not as powerful as deep approaches, gives superior results in practice, due to the computer's limited world knowledge. However, it can be confused by sentences like The dogs bark at the tree which contains the word bark near both tree and dogs.

There are four conventional approaches to WSD:

• Dictionary- and knowledge-based methods: These rely primarily on dictionaries, thesauri, and lexical knowledge bases, without using any corpus evidence.
• Supervised methods: These make use of sense-annotated corpora to train from.
• Semi-supervised or minimally supervised methods: These make use of a secondary source of knowledge such as a small annotated corpus as seed data in a bootstrapping process, or a word-aligned bilingual corpus.
• Unsupervised methods: These eschew (almost) completely external information and work directly from raw unannotated corpora. These methods are also known under the name of word sense discrimination.

Almost all these approaches normally work by defining a window of n content words around each word to be disambiguated in the corpus, and statistically analyzing those n surrounding words. Two shallow approaches used to train and then disambiguate are Naïve Bayes classifiers and decision trees. In recent research, kernel-based methods such as support vector machines have shown superior performance in supervised learning. Graph-based approaches have also gained much attention from the research community, and currently achieve performance close to the state of the art.

Dictionary- and knowledge-based methods[link]

The Lesk algorithm^[18] is the seminal dictionary-based method. It is based on the hypothesis that words used together in text are related to each other and that the relation can be observed in the definitions of the words and their senses. Two (or more) words are disambiguated by finding the pair of dictionary senses with the greatest word overlap in their dictionary definitions. For example, when disambiguating the words in "pine cone", the definitions of the appropriate senses both include the words evergreen and tree (at least in one dictionary).

An alternative to the use of the definitions is to consider general word-sense relatedness and to compute the semantic similarity of each pair of word senses based on a given lexical knowledge base such as WordNet. Graph-based methods reminiscent of spreading activation research of the early days of AI research have been applied with some success. More complex graph-based approaches have been shown to perform almost as well as supervised methods^[19] or even outperforming them on specific domains.^[3][20] Recently, it has been reported that simple graph connectivity measures, such as degree, perform state-of-the-art WSD in the presence of a sufficiently rich lexical knowledge base.^[21] Also, automatically transferring knowledge in the form of semantic relations from Wikipedia to WordNet has been shown to boost simple knowledge-based methods, enabling them to rival the best supervised systems and even outperform them in a domain-specific setting.^[22]

The use of selectional preferences (or selectional restrictions) is also useful, for example, knowing that one typically cooks food, one can disambiguate the word bass in "I am cooking basses" (i.e., it's not a musical instrument).

Supervised methods[link]

Supervised methods are based on the assumption that the context can provide enough evidence on its own to disambiguate words (hence, world knowledge and reasoning are deemed unnecessary). Probably every machine learning algorithm going has been applied to WSD, including associated techniques such as feature selection, parameter optimization, and ensemble learning. Support Vector Machines and memory-based learning have been shown to be the most successful approaches, to date, probably because they can cope with the high-dimensionality of the feature space. However, these supervised methods are subject to a new knowledge acquisition bottleneck since they rely on substantial amounts of manually sense-tagged corpora for training, which are laborious and expensive to create.

Semi-supervised methods[link]

Because of the lack of training data, many word sense disambiguation algorithms use semi-supervised learning, which allows both labeled and unlabeled data. The Yarowsky algorithm was an early example of such an algorithm.^[23] It uses the ‘One sense per collocation’ and the ‘One sense per discourse’ properties of human languages for word sense disambiguation. From observation, words tend to exhibit only one sense in most given discourse and in a given collocation.

The bootstrapping approach starts from a small amount of seed data for each word: either manually tagged training examples or a small number of surefire decision rules (e.g., 'play' in the context of 'bass' almost always indicates the musical instrument). The seeds are used to train an initial classifier, using any supervised method. This classifier is then used on the untagged portion of the corpus to extract a larger training set, in which only the most confident classifications are included. The process repeats, each new classifier being trained on a successively larger training corpus, until the whole corpus is consumed, or until a given maximum number of iterations is reached.

Other semi-supervised techniques use large quantities of untagged corpora to provide co-occurrence information that supplements the tagged corpora. These techniques have the potential to help in the adaptation of supervised models to different domains.

Also, an ambiguous word in one language is often translated into different words in a second language depending on the sense of the word. Word-aligned bilingual corpora have been used to infer cross-lingual sense distinctions, a kind of semi-supervised system.

Unsupervised methods[link]

Unsupervised learning is the greatest challenge for WSD researchers. The underlying assumption is that similar senses occur in similar contexts, and thus senses can be induced from text by clustering word occurrences using some measure of similarity of context,^[24] a task referred to as word sense induction or discrimination. Then, new occurrences of the word can be classified into the closest induced clusters/senses. Performance has been lower than other methods, above, but comparisons are difficult since senses induced must be mapped to a known dictionary of word senses. If a mapping to a set of dictionary senses is not desired, cluster-based evaluations (including measures of entropy and purity) can be performed. Alternatively, word sense induction methods can be tested and compared within an application. For instance, it has been shown that word sense induction improves Web search result clustering by increasing the quality of result clusters and the degree diversification of result lists.^[25] It is hoped that unsupervised learning will overcome the knowledge acquisition bottleneck because they are not dependent on manual effort.

Other approaches[link]

Other approaches may vary differently in their methods:

• Identification of dominant word senses;^[26][27][28]
• Domain-driven disambiguation;^[29][30]
• WSD using Cross-Lingual Evidence.^[31][32]

Local impediments and summary[link]

The knowledge acquisition bottleneck is perhaps the major impediment to solving the WSD problem. Unsupervised methods rely on knowledge about word senses, which is barely formulated in dictionaries and lexical databases. Supervised methods depend crucially on the existence of manually annotated examples for every word sense, a requisite that can so far be met only for a handful of words for testing purposes, as it is done in the Senseval exercises.

Therefore, one of the most promising trends in WSD research is using the largest corpus ever accessible, the World Wide Web, to acquire lexical information automatically.^[33] WSD has been traditionally understood as an intermediate language engineering technology which could improve applications such as information retrieval (IR). In this case, however, the reverse is also true: Web search engines implement simple and robust IR techniques that can be successfully used when mining the Web for information to be employed in WSD. Therefore, the lack of training data provoked appearing some new algorithms and techniques described here:

External knowledge sources[link]

Knowledge is a fundamental component of WSD. Knowledge sources provide data which are essential to associate senses with words. They can vary from corpora of texts, either unlabeled or annotated with word senses, to machine-readable dictionaries, thesauri, glossaries, ontologies, etc. They can be^[34][35] classified as follows:

1. Structured:
   • Thesauri
   • Machine-readable dictionaries (MRDs)
   • Ontologies
2. Unstructured:
   • Corpora: raw corpora and sense-annotated corpora
   • Collocation resources
   • Other resources (such as word frequency lists, stoplists, domain labels,^[36] etc.)

Evaluation[link]

Comparing and evaluating different WSD systems is extremely difficult, because of the different test sets, sense inventories, and knowledge resources adopted. Before the organization of specific evaluation campaigns most systems were assessed on in-house, often small-scale, data sets. In order to test one's algorithm, developers should spend their time to annotate all word occurrences. And comparing methods even on the same corpus is not eligible if there is different sense inventories.

In order to define common evaluation datasets and procedures, public evaluation campaigns have been organized. Senseval (now renamed SemEval) is an international word sense disambiguation competition, held every three years since 1998: Senseval-1 (1998), Senseval-2 (2001), Senseval-3 (2004), and its successor, SemEval (2007). The objective of the competition is to organize different lectures, preparing and hand-annotating corpus for testing systems, perform a comparative evaluation of WSD systems in several kinds of tasks, including all-words and lexical sample WSD for different languages, and, more recently, new tasks such as semantic role labeling, gloss WSD, lexical substitution, etc. The systems submitted for evaluation to these competitions usually integrate different techniques and often combine supervised and knowledge-based methods (especially for avoiding bad performance in lack of training examples).

Task Design Choices[link]

Sense Inventories. During the first Senseval workshop the HECTOR sense inventory was adopted. The reason for adopting a previously unknown sense inventory was mainly to avoid the use of popular fine-grained word senses (such as WordNet), which could make the experiments unfair or biased. However, given the lack of coverage of such inventories, since the second Senseval workshop the WordNet sense inventory has been adopted.

A set of testing words. Comparison of methods can be divided in 2 groups by amount of words to test. The difference consists in the amount of analysis and processing:

• all-words task implies disambiguating all the words of the text
• lexical sample consists in disambiguating some previously chosen target words.

It is assumed that the former one is more realistic evaluation, although with very laborious testing of results. Initially only the latter was used in evaluation but later the former was included.

Lexical sample organizers had to choose samples on which the systems were to be tested. A criticism of earlier forays into lexical-sample WSD evaluation is that the lexical sample had been chosen according to the whim of the experimenter (or, to coincide with earlier experimenters' selections). For English Senseval, a sampling frame was devised in which words were classified according to their frequency (in the BNC) and their polysemy level (in WordNet). Also, inclusion POS-tagging problem was a matter of discussion and it was decided that samples should be words with known part of speech and some indeterminants (for ex. 15 noun tasks, 13 verb tasks, 8 adjectives, and 5 indeterminates).

Baselines. For comparison purposes, known, yet simple, algorithms named baselines are used. These include different variants of Lesk algorithm or most frequent sense algorithm.

Sense inventory. WSD exercises require a dictionary, to specify the word senses which are to be disambiguated, and a corpus of language data to be disambiguated. WordNet is the most popular example of sense inventory. The reason for adopting the HECTOR database during Senseval-1 was that the WordNet inventory was already publicly available.^[37]

Evaluation measures. During the evaluation of WSD systems two main performance measures are used:

• Precision: the fraction of system assignments made that are correct
• Recall: the fraction of total word instances correctly assigned by a system

If a system makes an assignment for every word, then precision and recall are the same, and can be called accuracy. This model has been extended to take into account systems that return a set of senses with weights for each occurrence.

Software[link]

• WordNet::SenseRelate, is a project that includes free, open source systems for word sense disambiguation and lexical sample sense disambiguation.
• UKB: Graph Base WSD, is a collection of programs for performing graph-based Word Sense Disambiguation and lexical similarity/relatedness using a pre-existing Lexical Knowledge Base (LKB).

See also[link]

• Ambiguity
• Lesk algorithm
• Lexical substitution
• Part-of-speech tagging
• Polysemy
• Semeval
• Syntactic ambiguity
• Word sense
• Word sense induction

Notes[link]

Works cited[link]

• Agirre, E.; Lopez de Lacalle, A.; Soroa, A. (2009) "Knowledge-based WSD on Specific Domains: Performing better than Generic Supervised WSD" Proc. of IJCAI http://www.ijcai.org/papers09/Papers/IJCAI09-251.pdf 
• Agirre, E.; M. Stevenson. 2006. Knowledge sources for WSD. In Word Sense Disambiguation: Algorithms and Applications, E. Agirre and P. Edmonds, Eds. Springer, New York, NY.
• Bar-Hillel, Y. (1964). Language and information. Reading, MA: Addison-Wesley. 
• Buitelaar, P.; B. Magnini, C. Strapparava and P. Vossen. 2006. Domain-specific WSD. In Word Sense Disambiguation: Algorithms and Applications, E. Agirre and P. Edmonds, Eds. Springer, New York, NY.
• Chan, Y. S.; H. T. Ng. 2005. Scaling up word sense disambiguation via parallel texts. In Proceedings of the 20th National Conference on Artificial Intelligence (AAAI, Pittsburgh, PA).
• Edmonds, P. 2000. Designing a task for SENSEVAL-2. Tech. note. University of Brighton, Brighton. U.K.
• Fellbaum, Christiane (1997) "Analysis of a handwriting task" Proc. of ANLP-97 Workshop on Tagging Text with Lexical Semantics: Why, What, and How? Washington D.C., USA 
• Gliozzo, A.; B. Magnini and C. Strapparava. 2004. Unsupervised domain relevance estimation for word sense disambiguation. In Proceedings of the 2004 Conference on Empirical Methods in Natural Language Processing (EMNLP, Barcelona, Spain).
• Ide, N.; T. Erjavec, D. Tufis. 2002. Sense discrimination with parallel corpora. In Proceedings of ACL Workshop on Word Sense Disambiguation: Recent Successes and Future Directions (Philadelphia, PA).
• Kilgarriff, A. 1997. I don't believe in word senses. Comput. Human. 31(2), pp. 91–113.
• Kilgarriff, A.; G. Grefenstette. 2003. Introduction to the special issue on the Web as corpus. Computational Linguistics 29(3), pp. 333–347
• Kilgarriff, Adam; Joseph Rosenzweig, English Senseval: Report and Results May–June, 2000, University of Brighton
• Lapata, M.; and F. Keller. 2007. An information retrieval approach to sense ranking. In Proceedings of the Human Language Technology Conference of the North American Chapter of the Association for Computational Linguistics (HLT-NAACL, Rochester, NY).
• Lenat, D. "Computers versus Common Sense". http://www.youtube.com/watch?v=KSrUHGaUE_c. Retrieved 2008-12-10.  (GoogleTachTalks on YouTube)
• Lenat, D.; R. V. Guha. 1989. Building Large Knowledge-Based Systems, Addison-Wesley
• Lesk; M. 1986. Automatic sense disambiguation using machine readable dictionaries: How to tell a pine cone from an ice cream cone. In Proc. of SIGDOC-86: 5th International Conference on Systems Documentation, Toronto, Canada.
• Litkowski, K. C. 2005. Computational lexicons and dictionaries. In Encyclopaedia of Language and Linguistics (2nd ed.), K. R. Brown, Ed. Elsevier Publishers, Oxford, U.K.
• Magnini, B; G. Cavaglià. 2000. Integrating subject field codes into WordNet. In Proceedings of the 2nd Conference on Language Resources and Evaluation (LREC, Athens, Greece).
• McCarthy, D.; R. Koeling, J. Weeds, J. Carroll. 2007. Unsupervised acquisition of predominant word senses. Computational Linguistics 33(4): 553–590.
• McCarthy, D.; R. Navigli. 2009. The English Lexical Substitution Task, Language Resources and Evaluation, 43(2), Springer.
• Mihalcea, R. 2007. Using Wikipedia for Automatic Word Sense Disambiguation. In Proc. of the North American Chapter of the Association for Computational Linguistics (NAACL 2007), Rochester, April 2007.
• Mohammad, S; G. Hirst. 2006. Determining word sense dominance using a thesaurus. In Proceedings of the 11th Conference on European chapter of the Association for Computational Linguistics (EACL, Trento, Italy).
• Navigli, R. 2006. Meaningful Clustering of Senses Helps Boost Word Sense Disambiguation Performance. Proc. of the 44th Annual Meeting of the Association for Computational Linguistics joint with the 21st International Conference on Computational Linguistics (COLING-ACL 2006), Sydney, Australia.
• Navigli, R.; G. Crisafulli. Inducing Word Senses to Improve Web Search Result Clustering. Proc. of the 2010 Conference on Empirical Methods in Natural Language Processing (EMNLP 2010), MIT Stata Center, Massachusetts, USA.
• Navigli, R.; M. Lapata. An Experimental Study of Graph Connectivity for Unsupervised Word Sense Disambiguation. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 32(4), IEEE Press, 2010.
• Navigli, R.; K. Litkowski, O. Hargraves. 2007. SemEval-2007 Task 07: Coarse-Grained English All-Words Task. Proc. of Semeval-2007 Workshop (SemEval), in the 45th Annual Meeting of the Association for Computational Linguistics (ACL 2007), Prague, Czech Republic.
• Navigli, R.;P. Velardi. 2005. Structural Semantic Interconnections: a Knowledge-Based Approach to Word Sense Disambiguation. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 27(7).
• Palmer, M.; O. Babko-Malaya and H. T. Dang. 2004. Different sense granularities for different applications. In Proceedings of the 2nd Workshop on Scalable Natural Language Understanding Systems in HLT/NAACL (Boston, MA).
• Ponzetto, S. P.; R. Navigli. Knowledge-rich Word Sense Disambiguation rivaling supervised systems. In Proc. of the 48th Annual Meeting of the Association for Computational Linguistics (ACL), 2010.
• Pradhan, S.; E. Loper, D. Dligach, M. Palmer. 2007. SemEval-2007 Task 17: English lexical sample, SRL and all words. Proc. of Semeval-2007 Workshop (SEMEVAL), in the 45th Annual Meeting of the Association for Computational Linguistics (ACL 2007), Prague, Czech Republic.
• Schütze, H. 1998. Automatic word sense discrimination. Computational Linguistics, 24(1): 97–123.
• Snow, R.; S. Prakash, D. Jurafsky, A. Y. Ng. 2007. Learning to Merge Word Senses, Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning (EMNLP-CoNLL).
• Snyder, B.; M. Palmer. 2004. The English all-words task. In Proc. of the 3rd International Workshop on the Evaluation of Systems for the Semantic Analysis of Text (Senseval-3), Barcelona, Spain.
• Weaver, Warren (1949). "Translation". In Locke, W.N.; Booth, A.D.. Machine Translation of Languages: Fourteen Essays. Cambridge, MA: MIT Press. http://www.hutchinsweb.me.uk/MTNI-22-1999.pdf. 
• Wilks, Y.; B. Slator, L. Guthrie. 1996. Electric Words: dictionaries, computers and meanings. Cambridge, MA: MIT Press.
• Yarowsky, D. Word-sense disambiguation using statistical models of Roget's categories trained on large corpora. In Proc. of the 14th conference on Computational linguistics (COLING), 1992.
• Yarowsky, D. 1995. Unsupervised word sense disambiguation rivaling supervised methods. In Proc. of the 33rd Annual Meeting of the Association for Computational Linguistics.

External links and suggested reading[link]

Look up disambiguation in Wiktionary, the free dictionary.

• Computational Linguistics Special Issue on Word Sense Disambiguation (1998)
• Evaluation Exercises for Word Sense Disambiguation The de-facto standard benchmarks for WSD systems.
• Roberto Navigli. Word Sense Disambiguation: A Survey, ACM Computing Surveys, 41(2), 2009, pp. 1–69. An up-to-date state of the art of the field.
• Word Sense Disambiguation as defined in Scholarpedia
• Word Sense Disambiguation: The State of the Art (PDF) A comprehensive overview By Prof. Nancy Ide & Jean Véronis (1998).
• Word Sense Disambiguation Tutorial, by Rada Mihalcea and Ted Pedersen (2005).
• Word Sense Disambiguation: Algorithms and Applications, edited by Eneko Agirre and Philip Edmonds (2006), Springer. Covers the entire field with chapters contributed by leading researchers. www.wsdbook.org site of the book
• Bar-Hillel, Yehoshua. 1964. Language and Information. New York: Addison-Wesley.
• Edmonds, Philip & Adam Kilgarriff. 2002. Introduction to the special issue on evaluating word sense disambiguation systems. Journal of Natural Language Engineering, 8(4):279-291.
• Edmonds, Philip. 2005. Lexical disambiguation. The Elsevier Encyclopedia of Language and Linguistics, 2nd Ed., ed. by Keith Brown, 607-23. Oxford: Elsevier.
• Ide, Nancy & Jean Véronis. 1998. Word sense disambiguation: The state of the art. Computational Linguistics, 24(1):1-40.
• Jurafsky, Daniel & James H. Martin. 2000. Speech and Language Processing. New Jersey, USA: Prentice Hall.
• Litkowski, K. C. 2005. Computational lexicons and dictionaries. In Encyclopaedia of Language and Linguistics (2nd ed.), K. R. Brown, Ed. Elsevier Publishers, Oxford, U.K., 753–761.
• Manning, Christopher D. & Hinrich Schütze. 1999. Foundations of Statistical Natural Language Processing. Cambridge, MA: MIT Press. http://nlp.stanford.edu/fsnlp/
• Mihalcea, Rada. 2007. Word sense disambiguation. Encyclopedia of Machine Learning. Springer-Verlag.
• Resnik, Philip and David Yarowsky. 2000. Distinguishing systems and distinguishing senses: New evaluation methods for word sense disambiguation, Natural Language Engineering, 5(2):113-133. http://www.cs.jhu.edu/~yarowsky/pubs/nle00.ps
• Yarowsky, David. 2001. Word sense disambiguation. Handbook of Natural Language Processing, ed. by Dale et al., 629-654. New York: Marcel Dekker.

Junior Wells
Wells performing at the New Orleans Jazz & Heritage Festival, 1996
Background information
Birth name	Amos Wells Blakemore Jr.
Born	(1934-12-09)December 9, 1934
Origin	Memphis, Tennessee or West Memphis, Arkansas
Died	January 15, 1998(1998-01-15) (aged 63) Chicago, Illinois, United States
Genres	Chicago blues
Occupations	Musician, Songwriter
Instruments	Vocals, harmonica
Years active	1950s – 1997
Labels	States, Chief, Profile, Delmark, Vanguard, Telarc
Associated acts	The Aces Junior Wells Chicago Blues Band Buddy Guy

Junior Wells (December 9, 1934 – January 15, 1998), born Amos Wells Blakemore Jr.,^[1] was an American Chicago blues vocalist, harmonica player, and recording artist. Wells, who was best known for his performances and recordings with Muddy Waters, Earl Hooker, and Buddy Guy, also performed with Bonnie Raitt, The Rolling Stones, and Van Morrison.^[1]

Life and career[link]

Junior Wells was born in Memphis, Tennessee, United States,^[1] and raised in West Memphis, Arkansas, though other sources report that his birth was in West Memphis.^[2] Initially taught by his cousin, Junior Parker, and Sonny Boy Williamson II, Wells learned how to play the harmonica by the age of seven with surprising skill. He moved to Chicago in 1948 with his mother after her divorce and began sitting in with local musicians at house parties and taverns.^[3] Wild and rebellious but needing an outlet for his talents, he began performing with The Aces (guitarist brothers Dave and Louis Myers and drummer Fred Below) and developed a more modern amplified harmonica style influenced by Little Walter.^[3] In 1952, he made his first recordings, when he replaced Little Walter in Muddy Waters' band and appeared on one of Muddy's sessions for Chess Records in 1952.^[3] His first recordings as a band leader were made in the following year for States Records.^[4] In the later 1950s and early 1960s he also recorded singles for Chief Records and its Profile Records subsidiary, including "Messin' with the Kid", "Come on in This House", and "It Hurts Me Too", which would remain in his repertoire throughout his career. His 1960 Profile single "Little by Little" (written by Chief owner and producer Mel London) reached #23 in the Billboard R&B chart, making it the first of two Wells' singles to enter the chart.^[5]

Junior Wells worked with guitarist Buddy Guy in the 1960s, and featured Guy on guitar when he recorded his first album, Hoodoo Man Blues for Delmark Records.^[3][6] Wells and Guy supported the Rolling Stones on numerous occasions in the 1970s.^[6] Although his albums South Side Blues Jam (1971) and On Tap (1975) proved he had not lost his aptitude for Chicago blues, his 1980s and 1990s discs were inconsistent.^[6] However, 1996's Come On in This House was an intriguing set of classic blues songs with a rotating cast of slide guitarists, among them Alvin Youngblood Hart, Corey Harris, Sonny Landreth^[6] and Derek Trucks. Wells made an appearance in the film Blues Brothers 2000, the sequel to The Blues Brothers, which was released in 1998.^[3]

Wells continued performing until he was diagnosed with cancer in the summer of 1997.^[3] That fall, he suffered a heart attack while undergoing treatment, sending him into a coma.^[3] Wells died in Chicago, after succumbing to lymphoma on January 15, 1998, and was interred in the Oak Woods Cemetery, Chicago.^[1]

Wells was mentioned in the Steppenwolf song, "Tighten Up Your Wig", in which the lyrics explicitly state that they copied the music from Junior Wells' tune, "Messing with the Kid".

Album discography[link]

{† Albums that feature Buddy Guy)

See also[link]

• Chicago Blues Festival

References[link]

External links[link]

• Centerstage Chicago Obituary
• Junior Wells at Find a Grave
• Elvis Pelvis Obituary
• Island Net Obituary
• Junior Wells at Allmusic

George Stephenson
Engineer and Inventor
Born	(1781-06-09)9 June 1781 Wylam, Northumberland, England
Died	12 August 1848(1848-08-12) (aged 67) Tapton House, Chesterfield, Derbyshire, England
Resting place	Holy Trinity Church, Chesterfield
Nationality	English
Citizenship	British

George Stephenson (9 June 1781 – 12 August 1848) was an English civil engineer and mechanical engineer who built the first public railway line in the world to use steam locomotives. Renowned as being the "Father of Railways", the Victorians considered him a great example of diligent application and thirst for improvement, with self-help advocate Samuel Smiles particularly praising his achievements. His rail gauge of 4 feet 8+¹⁄₂ inches (1,435 mm), sometimes called "Stephenson gauge", is the world's standard gauge.

Early life[link]

George Stephenson

George Stephenson was born in Wylam, Northumberland, 9.3 miles (15.0 km) west of Newcastle upon Tyne. He was the second child of Robert and Mabel,^[1] neither of whom could read or write. Robert was the fireman for Wylam Colliery pumping engine, earning a very low wage, so that there was no money for schooling. At 17, Stephenson became an engineman at Water Row Pit, Newburn. George realised the value of education and paid to study at night school to learn reading, writing and arithmetic—he was illiterate till the age of 18. In 1801 he began work at Black Callerton colliery as a 'brakesman', controlling the winding gear of the pit. In 1802 he married Frances (Fanny) Henderson and moved to Willington Quay, east of Newcastle. There he worked as a brakesman while they lived in one room of a cottage. George made shoes and mended clocks to supplement his income.

Dial Cottage, West Moor, Killingworth

In 1803 their son Robert was born, and in 1804 they moved to West Moor, near Killingworth while George worked as a brakesman at Killingworth pit. His wife gave birth to a daughter, who died after a few weeks, and in 1806 Fanny died of consumption (tuberculosis). George then decided to find work in Scotland, and he left Robert with a local woman while he went to work in Montrose. After a few months he returned, probably because his father was blinded in a mining accident. George moved back into his cottage at West Moor and his unmarried sister Eleanor moved in to look after Robert. In 1811 the pumping engine at High Pit, Killingworth was not working properly and Stephenson offered to fix it. He did so with such success that he was soon promoted to enginewright for the neighbouring collieries at Killingworth, responsible for maintaining and repairing all of the colliery engines. He soon became an expert in steam-driven machinery.^[2]

The miners' safety lamp[link]

Stephenson's safety lamp shown with Davy's lamp on the left

In 1815, aware of the explosions often caused in mines by naked flames, Stephenson began to experiment with a safety lamp that would burn without causing an explosion. At the same time, Cornishman Sir Humphry Davy, the eminent scientist was also looking at the problem. Despite his lack of any scientific knowledge, Stephenson, by trial and error, devised a lamp in which the air entered via tiny holes. Stephenson demonstrated the lamp himself to two witnesses by taking it down Killingworth colliery and holding it directly in front of a fissure from which fire damp was issuing. This was a month before Davy presented his design to the Royal Society. The two designs differed in that, the Davy's lamp was surrounded by a screen of gauze, whereas Stephenson's lamp was contained in a glass cylinder. For his invention Davy was awarded £2,000, whilst Stephenson was accused of stealing the idea from Davy. A local committee of enquiry exonerated Stephenson, proved that he had been working separately and awarded him £1,000 but Davy and his supporters refused to accept this. They could not see how an uneducated man such as Stephenson could come up with the solution that he had. In 1833 a House of Commons committee found that Stephenson had equal claim to having invented the safety lamp. Davy went to his grave believing that Stephenson had stolen his idea. The Stephenson lamp was used exclusively in the North East, whereas the Davy lamp was used everywhere else. The experience with Davy gave Stephenson a life-long distrust of London-based, theoretical, scientific experts.^[2]

There is a theory that it was Stephenson who indirectly gave the name of Geordies to the people of Tyneside. By this theory, the name of the Geordie lamp attached to the pit men themselves. By 1866 any native of Tyneside could be called a Geordie.^[3]

Early locomotives[link]

Cornishman Richard Trevithick is credited with the first realistic design of the steam locomotive in 1802. Later, he visited Tyneside and built an engine there for a mine-owner. Several local men were inspired by this, and designed engines of their own.

Locomotive constructed in 1816 by Stephenson for the Killingworth Colliery

Stephenson designed his first locomotive in 1814, a travelling engine designed for hauling coal on the Killingworth wagonway, and named Blücher after the Prussian general Gebhard Leberecht von Blücher. It was constructed in the colliery workshop behind Stephenson's home, Dial Cottage, on Great Lime Road. This locomotive could haul 30 tons of coal up a hill at 4 mph (6.4 km/h), and was the first successful flanged-wheel adhesion locomotive: its traction depended only on the contact between its flanged wheels and the rail. Altogether, Stephenson is said to have produced 16 locomotives at Killingworth,^[2] although it has never proved possible to produce a convincing list of all 16. Of those that have been identified most were built for use at Killingworth itself or for the Hetton colliery railway. A six-wheeled locomotive was built for the Kilmarnock and Troon Railway in 1817 but it was soon withdrawn from service because of damage to the cast iron rails.^[4] A further locomotive was supplied to Scott's Pit railroad at Llansamlet, near Swansea in 1819 but it too was soon withdrawn, apparently because it was under-boilered and also because of damage to the track.^[5]

The new engines were too heavy to be run on wooden rails, and iron rails were in their infancy, with cast iron exhibiting excessive brittleness. Together with William Losh, Stephenson improved the design of cast iron rails to reduce breakage; these were briefly made by Losh, Wilson and Bell at their Walker ironworks. According to Rolt, he also managed to solve the problem caused by the weight of the engine upon these primitive rails. He experimented with a 'steam spring' (to 'cushion' the weight using steam pressure), but soon followed the new practice of 'distributing' weight by utilising a number of wheels. For the Stockton and Darlington Railway, however, Stephenson would use only wrought iron rails, notwithstanding the financial loss he would suffer from not using his own, patented design (see below).^[6]

Stephenson was hired to build an 8-mile (13-km) railway from Hetton colliery to Sunderland in 1820. The finished result used a combination of gravity on downward inclines and locomotives for level and upward stretches. It was the first railway using no animal power.

Other locomotives include:

• 1817-1848 The Duke for the Kilmarnock and Troon Railway

Stockton and Darlington Railway[link]

The No. 1 engine, called Locomotion, for the Stockton & Darlington Railway

In 1821, a parliamentary bill was passed to allow the building of the Stockton and Darlington Railway (S&DR). The 25-mile (40 km) railway was intended to connect various collieries situated near Bishop Auckland to the River Tees at Stockton, passing through Darlington on the way. The original plan was to use horses to draw coal carts on metal rails, but after company director Edward Pease met Stephenson he agreed to change the plans. Stephenson surveyed the line in 1821, assisted by his eighteen-year-old son Robert. That same year construction of the line began.^[2]

The Experiment - the first railway carriage

A manufacturer was now needed to provide the locomotives for the new line. As it turned out, Pease and Stephenson jointly established a company in Newcastle to manufacture locomotives. The company was set up as Robert Stephenson and Company, and George's son Robert was the managing director. A fourth partner was Michael Longridge of Bedlington Ironworks.^[2] On an early trade card, Robert Stephenson & Co were described as "Engineers, Millwrights & Machinists, Brass & Iron Founders".^[7] In September 1825 the works at Forth Street, Newcastle completed the first locomotive for the new railway: originally named Active, it was soon renamed Locomotion. It was followed by "Hope", "Diligence" and "Black Diamond". The Stockton and Darlington Railway opened on 27 September 1825. Driven by Stephenson, Locomotion hauled an 80-ton load of coal and flour nine miles (15 km) in two hours, reaching a speed of 24 miles per hour (39 km/h) on one stretch. The first purpose-built passenger car, dubbed Experiment, was attached, and carried dignitaries on the opening journey. It was the first time passenger traffic had been run on a steam locomotive railway.^[2]

The rails used for the new line were wrought-iron ones, produced by John Birkinshaw at the Bedlington Ironworks. Wrought-iron rails could be produced in much longer lengths than the cast-iron ones and were much less liable to crack under the weight of heavy locomotives. William Losh of Walker Ironworks had thought that he had an agreement with Stephenson to use his cast-iron rails, and Stephenson's decision caused a permanent rift between the two men. The gauge that Stephenson chose for the line was 4 feet 8+¹⁄₂ inches (1,435 mm), and this subsequently came to be adopted as the standard gauge for railways, not only in Britain, but also throughout the world.^[2]

Liverpool and Manchester Railway[link]

Statue of George Stephenson at the National Railway Museum, York

First passenger railway, L&MR

Stephenson had ascertained by experiments at Killingworth that half of the power of the locomotive was consumed by a gradient as little as 1 in 260.^[8] He came to the conclusion that railways should be kept as level as possible. He used this knowledge while working on the Bolton and Leigh Railway, and the Liverpool and Manchester Railway (L&MR), executing a series of difficult cuts, embankments and stone viaducts to smooth the route the railways took. Defective surveying of the original route of the L&MR caused by the hostility of some of the affected landowners meant that Stephenson encountered difficulty during Parliamentary scrutiny of the original bill, especially under cross-examination by Edward Hall Alderson. The Bill was rejected. A revised bill with a new alignment was submitted and passed in a subsequent session. The revised alignment presented a considerable problem: the crossing of Chat Moss, an apparently bottomless peat bog, which Stephenson eventually overcame by unusual means, effectively floating the line across it.^[2] The method he used was almost exactly the same as that used by John Metcalf during his construction of many miles of road through marshes in the Pennines. He would lay a foundation of heather and branches bound together by the weight of the passing coaches with a layer of stones on top.

As the L&MR approached completion in 1829, its directors arranged for a competition to decide who would build its locomotives, and the Rainhill Trials were run in October 1829. Entries could weigh no more than six tons and had to travel along the track for a total distance of 60 miles (97 km). Stephenson's entry was Rocket, and its performance in winning the contest made it famous. George's son Robert had been working in South America from 1824 to 1827 and had returned to run the Forth Street Works while George was living in Liverpool and overseeing the construction of the new line. Robert was very much responsible for the detailed design of Rocket, although he was in constant postal communication with George, who made many suggestions on the design. One significant innovation was the use of a fire-tube boiler, invented by French engineer Marc Seguin that gave improved heat exchange. This was suggested by Henry Booth, the treasurer of the L&MR.^[2]

The opening ceremony of the L&MR, on 15 September 1830, was a considerable event, drawing luminaries from the government and industry, including the Prime Minister, the Duke of Wellington. The day started with a procession of eight trains setting out from Liverpool. The parade was led by "Northumbrian" driven by George Stephenson, and included "Phoenix" driven by his son Robert, "North Star" driven by his brother Robert and "Rocket" driven by assistant engineer Joseph Locke. The day was marred by the death of William Huskisson, the Member of Parliament for Liverpool, who was struck and killed by Rocket. Stephenson in person evacuated the injured Huskisson to Eccles with a train, but the injury was beyond help. Despite this tragedy the railway was a resounding success. Stephenson became famous, and was offered the position of chief engineer for a wide variety of other railways.^[2]

Stephenson's skew arch bridge[link]

Stephensons Bridge

1830 also saw the grand opening of the skew bridge in Rainhill as part of the grand opening of the Liverpool and Manchester Railway. The bridge was the first to cross any railway at an angle.^[9] This required the structure to be constructed as two flat planes (overlapping in this case by 6') between which the stonework forms a parallelogram shape when viewed from above. This has the effect of flattening the arch and the solution is to lay the bricks forming the arch at an angle to the abutments (the piers on which the arches rest). This technique, which results in a spiral effect in the arch masonry, provides extra strength in the arch to compensate for the angled abutments.^[10]

The bridge is still in use at Rainhill railway station, and carries traffic on the busy A57 (Warrington Road). The bridge is now a listed structure.

Later career[link]

The next ten years were the busiest of Stephenson's life, as he was besieged with requests from railway promoters. Many of the first American railroad builders came to Newcastle to learn from Stephenson, and indeed, the first dozen or so locomotives utilized in the U.S. were purchased from the Stephenson shops. Other talented men were starting to make their marks, such as his son Robert, his pupil Joseph Locke and finally Isambard Kingdom Brunel. His conservative views on the capabilities of locomotives meant that he favoured circuitous routes and civil engineering that were more costly than his successors thought necessary. For example, rather than the West Coast Main Line taking the direct route favoured by Joseph Locke over Shap between Lancaster and Carlisle, Stephenson was in favour of a longer sea-level route via Ulverston and Whitehaven. Locke's route was the one built. Stephenson also tended to be more casual in estimating costs and paperwork in general. He worked with Joseph Locke on the Grand Junction Railway with one half of the line allocated to each man. Stephenson's estimates and organising ability proved to be inferior to those of Locke and the board's dis-satisfaction led to Stephenson's resignation. This caused a rift between Stephenson and Locke, which was never healed.^[2]

Despite Stephenson's losing some routes to competitors due to his caution, he was offered more work than he could cope with, and was unable to accept offers for additional work. He worked on the North Midland line from Derby to Leeds, the York and North Midland line from Normanton to York, the Manchester and Leeds, the Birmingham and Derby, the Sheffield and Rotherham among many others.^[2]

Stephenson tended to become a reassuring name, rather than a cutting-edge technical adviser. He was the first president of the Institution of Mechanical Engineers on its formation in 1847. He had by this time settled into semi-retirement, supervising his mining interests in Derbyshire - tunnelling work for the North Midland Railway had revealed unworked coal seams, and Stephenson put much of his money into their exploitation.

Personal life[link]

George first courted Elizabeth (Betty) Hindmarsh, a farmer's daughter from Black Callerton, whom he would meet secretly in her orchard. Her father Thomas refused marriage because of Stephenson's lowly status as a miner.^[11] George next paid attention to Anne Henderson where he lodged with her family, but when she also rejected him he transferred his attentions to her sister Frances (Fanny), who was nine years his senior.

George and Fanny married at Newburn Church on 28 November 1802. They had two children Robert (1803) and Fanny (1805) but the girl died within months, and George's wife died, probably of tuberculosis or, the year after. While George was away working in Scotland, Robert was brought up by a succession of neighbors and then by George's unmarried sister Eleanor (Nelly), who continued living with them in Killingworth on George's return.

On 29 March 1820, George (now considerably wealthier) was finally allowed to marry Betty Hindmarsh at Newburn. The marriage seems to have been a happy one, but there were no children from this union, and Betty died in 1845.

On 11 January 1848, at St John's Church in Shrewsbury, George married for the third time, to Ellen Gregory, another farmer's daughter originally from Bakewell in Derbyshire, who had been his housekeeper. Six months after his wedding, George contracted pleurisy and died, aged 67, on 12 August 1848 at Tapton House in Chesterfield, Derbyshire. He was buried at Holy Trinity Church, Chesterfield, alongside his second wife.^[2]

Descendants[link]

George Stephenson had two children:

• Robert was born on 16 October 1803 and married Frances Sanderson, daughter of a City of London professional John Sanderson, on 17 June 1829. Robert died in 1859 having no children.
• His daughter was born in 1805 but died within weeks of her birth.

Legacy[link]

Britain led the world in the development of railways and this acted as a stimulus for the industrial revolution, by facilitating the transport of raw materials and manufactured goods. George Stephenson cannot claim to have invented the locomotive. Richard Trevithick deserves that credit. George Stephenson, with his work on the Stockton and Darlington Railway and the Liverpool and Manchester Railway, paved the way for the railway engineers who were to follow, such as his son Robert, his assistant Joseph Locke who went on to carry out much work on his own account and Isambard Kingdom Brunel. These men were following in his footsteps. Stephenson was also farsighted in realising that the individual lines being built would eventually join together, and would need to have the same gauge. The standard gauge used throughout much of the world is due to him.

Memorials and commemorations[link]

Birthplace[link]

George Stephenson statue, Chesterfield March 2011

George Stephenson's Birthplace is an 18th century historic house museum in the village of Wylam, and is operated by the National Trust.

Memorabilia[link]

Chesterfield Museum in Chesterfield, Derbyshire, has a gallery of Stephenson memorabilia, including the straight thick glass tubes that he invented for the purpose of growing straight cucumbers. The museum is located within the Stephenson Memorial Hall^[12] (which also incorporates the Pomegranate Theatre, formerly the Chesterfield Civic Theatre) and adjacent to Stephensons Place, located roughly mid distance between Stephenson's final home at Tapton House and his resting place in Trinity Church. In Liverpool, where he lived at 34 Upper Parliament Street, a City of Liverpool Heritage Plaque is situated next to the front door of the property which has now been divided into private flats.

Academic buildings[link]

George Stephenson College, founded in 2001 on the University of Durham's Queen's Campus in Stockton-on-Tees, is named after him, with the student union bar being named The Rocket. Also named after him and his son is George Stephenson High School in Killingworth, Stephenson Memorial Primary School in Howdon, the Stephenson Railway Museum in North Shields and the Stephenson Locomotive Society. The Stephenson Centre, an SEBD Unit of Beaumont Hill School in Darlington, is also named after George Stephenson.

Statues and monuments[link]

As a tribute to his life and works, a bronze statue of Stephenson was unveiled at Chesterfield railway station (which is overlooked by Tapton House, where Stephenson spent the last ten years of his life) on 28 October 2005, marking the completion of improvements to the station. At the event a full-size working replica of the Rocket was on show, which then spent two days on public display at the Chesterfield Market Festival. A statue of George dressed in classical robes stands in Neville Street, Newcastle, facing the building that houses the Literary and Philosophical Society and the Mining Institute, and near to Newcastle railway station.

Banknotes[link]

From 1990 until 2003, Stephenson's portrait appeared on the reverse of Series E £5 notes issued by the Bank of England. Stephenson's face is shown alongside an engraving of his Rocket steam engine and the Skerne Bridge on the Stockton to Darlington Railway. His image was replaced in 2003 by that of Elizabeth Fry.^[13]

See also[link]

• History of Science and Technology
• Industrial Revolution
• Train
• Robert Stephenson
• Robert Stephenson and Company

References[link]

Biographical works[link]

• Smiles, Samuel (1857). The Life of George Stephenson. London. 
• Davies, Hunter (2004). George Stephenson: The Remarkable Life of the Founder of Railways. Stroud: Sutton Publishing. ISBN 978-0-7509-3795-5. 
• Rolt, L.T.C. (1960). George and Robert Stephenson: The Railway Revolution. London: Penguin. ISBN 978-0-14-007646-2. 
• Ross, David (2010). George and Robert Stephenson: A Passion for Success. Stroud: History Press. ISBN 978-0-7524-5277-7. 

External links[link]

Wikimedia Commons has media related to: George Stephenson

Wikiquote has a collection of quotations related to: George Stephenson

• Biography
• Biography
• Biography
• Stephenson's birthplace