Scarica Computational Linguistics: Processing Human Language with Computers e più Sintesi del corso in PDF di Filologia Inglese solo su Docsity! CAP. 11 – Computer Processing of Human Language Until a few decades ago, language was strictly “human only”, but today, it is common for computers to process language. Computational linguistics is a subfield of linguistics and computer science that is concerned with the interactions of human language and computers. For many purposes it would be helpful if we could communicate with computer sas we communicate with other humans, throght our native language. Computional linguistics is concerned with the interaction between language and computers in all dimensions: Computational phonetics and phonology is concerned with processing speech; its main goal are converting speech to text on the comprehension side, and text to speech on the production side. The two sides of this computational are speech recognition and speech synthesis: the first is the process of analyzing the speech signal into its component phones and phonemes, and producing a phonetic transcription og the speech; the second processi s the one of creating electronic signals that simulate the phones and prosodic features of speech and assemble them into words and phrases for output to an electronic speaker, or for further processing as in a language generation application. Computational syntax is concerned with the syntactic categories of words and with the larger syntactic units of phrases and sentences. Computational semantics is concerned with representing meaning inside the computer or semantic representation. To communicate with a person, the computer creates a semantic representation of what the person says to it, and another semantic representation of what it wants to say back. Computational pragmatics may influence the understanding or the response of the computer by taking into account knowledge that the computer system has about the real world. Computational lexicology is the use of computers both to construct “ordinary” dictionaries and to construct electronic dictionaries with far more information, suitable for the goals of language understanding and generation. Computers may be programmed to model a linguist’s grammar of a human language and thus rapidly and thoroughtly test that grammar. To analyze a corpus, or body of data, a computer can do a frequency analysis of words; compute a concordance which locates words in the corpus and gives their immediate context. A way to refine a concordance idìs through collocation analysis; the point is to find evidence that the presence of one word in the next affects the occurrence of other words. The Cultural Revolution A frequency analysis on a 361 billion word corpus of English revealed half a million undocumented English words. Twitterology is the computer anaslysis of a vast corpus of microblogs known as tweets, is becoming a subfield of culturomics, but also is confined to a very narrow form of communiaction that’s been called trivial and superficial. Information Retrieval Computers are also useful for information retrieval based on keywords, automatic summarization, trough which computers can eliminate redundancy and identify the most salient features of a body of information, and spell checking, applications of computational linguistics. Machine Translation The need to translate between languages has never been greater than it is in today’s global society. The first use of computers for natural language processing began in the 1940s with the attempt to develop automatic machine translation. The aim of this is to input a spoken utterance or a written passage in the source language and to receive a grammatical passage of equivalent meaning in the target language. Translation is more than word-for-word replacement. Often there is no equivalent word in the target language, and the order of words may differ. There are, however, challenges in morphology when translating between languages. For example, a word like “ungentlemanliness” is certainly translatable into any language, but few languages are likely to have an exact word that meaning. Computational Forensic Linguistics Forensic linguistics is a subfield of linguistics that applies to language as used in the legal and judicial fields. Computational forensic linguistics is a sub-area that concerns itself with computer applications in forensic linguistics. Other applications of computational linguistics are found in the forensic fields, where computational forensic linguists takes up such legal problems as trademark protection and infringement, in which computers are used to examine huge corpuses to infer how people interpret trademarks such as the Mc- in McDonald’s; and speaker identification, where a computational analysis of speech used in a crime such as a bomb threat can assist in identifying, or exonerating, a suspect.
