Abstract: A system including at least one processing unit and a memory encoded with instructions that, when executed by the at least one processing unit, cause the at least one processing unit to cause a plurality of first terms in a native language to be displayed on a screen; receive a selection of a first term and translate it into a first target term in a target language, and use grammatical rules to determine a plurality of second terms that are grammatically compatible with the first term; display the plurality of second terms; receive a selection of a second native term; use the grammatical rules to translate the second term into a second target term based on the first target term and the second term; and build a phrase in the target language based on the first and second target terms, the phrase being grammatically correct in the target language.

Abstract: An establishing method of a knowledge library engine based on emergency management includes steps of extracting relevant data from unstructured text information, filtering out unwanted data and improving a structure of the relevant data, so as to ensure homogeneity of the relevant data; structurally analyzing a text content of obtained information, extracting a feature thereof and storing the feature in the feature library as a key word of a knowledge; and processing the obtained information with the automatic word segmentation module, then classifying a result according to an information classification standard, and establishing an association of the knowledge and information classification. A querying method of the knowledge library engine based on emergency management includes steps of: understanding semanteme, understanding knowledge, and querying knowledge.

Abstract: Mechanisms are provided for translating messages output by processes associated with the data processing system. Language based user activities occurring via a data processing system are tracked and automatically analyzed to determine a relative priority ordering of one or more languages used by a user performing the user activities. A sequence of preferred language translation substitutions is generated for outputting messages based on results of the analysis and is automatically applied to a received message from a process associated with the data processing system to generate a translated message using one of the preferred language translation substitutions in the sequence of preferred language translation substitutions. The translated message is output via an output device of the data processing system.

Abstract: Systems and methods of various embodiments may enable or refine translation of text between a first language and a second language. In particular, systems and methods may enable or refine a text translation by soliciting and/or receiving feedback for: translation of a first word or phrase from a first language to a second language; transformation of the first word or phrase (in the first language) to a second word or phrase in the first language; or transformation of the first word or phrase (in the first language) to a second word or phrase in the second language. The systems and methods of various embodiments may incentivize user feedback for failed translations in order to encourage user feedback, improve the quality of user feedback received, and to permit development of translation corpora that can evolve with time.

Abstract: Contextual, focus-based language translation within a broadcast radio automation software application allows the user interface to enable interaction with users having various written or graphical language requirements. The software does not require special training or learning to enable the translation of text within the application. When the user enters a translation mode and selects a text string in the original language, the program displays a corresponding text string in a second language based on the specific display area, e.g., child window. The translated text string can be displayed proximate to original first text string within the specific display area, and can further be displayed as an entry in a table of a translation window which also displays additional text strings in the second language corresponding to additional text string fields for the specific display area.

Abstract: According to an aspect, a first word in a first language and a second word in a second language in a bilingual corpus are stemmed. A probability for aligning the first stem and the second stem and a distance metric between the normalized first stem and the normalized second stem are calculated. The first word and the second word are identified as a cognate pair when the probability and the distance metric meet a threshold criterion and stored as a cognate pair in a set of cognates. A candidate sentence in the second language is retrieved from a corpus. The candidate sentence is filtered by the active vocabulary of a user in the second language and the set of cognates. A sentence quality score is calculated for the candidate sentence; and the candidate sentence is ranked for presentation to the user based on the sentence quality scorer.

Abstract: Webpage content is sent to a user's web browser in a neutral format. On the user's web browser the webpage content is translated by a transformation application into the appropriate language and locale. A Language template file that contains the internationalization and localization data and processing components is loaded onto a browser and is used by the transformation application to render the final data to the user. The transformation application can render a new language and locale by loading an appropriate language template file.

Abstract: The embodiments determine the recommendations for a search term and its criteria, whereby a threshold is used for accepting a result, whether it is a document, message, file, or any other form of communication. The input may be part of a larger repository, and there is no restriction on how many documents constitute the returned recommendation set.

Abstract: A method, system and computer program product for providing translated web content is disclosed. The method includes receiving a request from a user on a web site, the web site having a first web content in a first language, wherein the request calls for a second web content in a second language. The method further includes dividing the first web content into a plurality of translatable components and generating a unique identifier for each translatable component. The method further includes identifying a plurality of translated components of the second web content using the unique identifier of each of the plurality of translatable components of the first web content and putting the plurality of translated components of the second web content to preserve a format that corresponds to the first web content. The method further includes providing the second web content in response to the request that was received.

Abstract: An automated communication system with an associated method for presenting customized voices is disclosed. The system which performs a predetermined task accepts information regarding an intended user indicating the intended user's identity, preferences, etc. Next, the system customizes one or more voices for the intended user based on the accepted information. The system then presents to the intended user one or more audible communications converted from text associated with a predetermined task performed by the system using the one or more customized voices.

Abstract: The present disclosure provides methods and systems for performing syntactic analysis of a text. In some implementations the method includes performing rough syntactic analysis of the text, generating a graph of generalized constituents of the text and filtering arcs of the graph of generalized constituents with a combination classifier which includes a tree classifier and one or more linear classifiers. The combination classifier is trained using parallel analysis of an untagged two-language text corpus.

Abstract: A system and method operable for providing contextually triggered vocabulary information on at least one display screen of an E-book reader are disclosed. A character group comprising a target language character group is displayed on the at least one display screen. A local or a remote electronic pre-translated primary language resource is provided in response to a user selecting a word or sentence in the target language character group. The electronic pre-translated primary language electronic resource is searched to obtain a pre-translated primary language character group corresponding to the target language character group. A synchronized pre-translated primary language character group is obtained by synchronizing the pre-translated primary language words or sentences to words or sentences in the target language. The synchronized pre-translated primary language character group is displayed on the at least one display screen.

Abstract: An arrangement and corresponding method are described for natural language processing. A natural language understanding (NLU) arrangement processes a natural language input to determine a corresponding sentence-level interpretation. A user state component maintains user context data that characterizes an operating context of the NLU arrangement. Operation of the NLU arrangement is biased by the user context data.

Abstract: Systems and methods are presented for producing datasets as symbolic or associative representations of terms or objects or entities. A term can be a word or a phrase, which can also be the name of an object or a topic or concept. A dataset is produced for a specific term or object. The dataset contains a plurality of other terms or values associated with the specific term, and can serve as a representation of the specific term by other terms or values obtained using machine-based algorithms from text contents. Terms or values in the datasets also represent information about an object, or information about properties associated with the object. Methods for obtaining the datasets include discovering the relationships between terms in a plurality of text contents, based on occurrence, location, and attributes associated with the terms in the text contents.

Abstract: Embodiments of the invention relate to computerized systems and methods for automatically translating text between natural languages. Text for translation may be broken into segments, which may themselves be templates including placeholders. If an exact match for a segment is found in a translation database, the segment may be matched with its translation from the database. Otherwise, alternate source language text, corresponding to the segment, for which an exact translation is available, may be identified and presented to the user for selection. Text selected by the user may replace the original segment and be matched with its translation. Translation of a segment may be influenced by other translated segments. Additionally, translations may be modified as the user interactively enters and edits text.

Abstract: A method for translating content includes providing a plurality of content on a user device. The plurality of content is associated with a first region display mode. The plurality of content is displayed on the user device in the first region display mode. A predefined orientation is then detected in the user device. The plurality of content is then translated according to a second region display mode to produce a plurality of translated content. The plurality of translated content is then displayed on the user device. In different embodiments, the plurality of content may have been previously provided on the user device, may be retrieved over a network by the user device, or my be received through a user interface on the user device. A specific embodiment allows a user to rotate the user device 180 degrees to quickly and easily translate text displayed on the user device.

Abstract: A system and method predict the translation quality of a translated input document. The method includes receiving an input document pair composed of a plurality of sentence pairs, each sentence pair including a source sentence in a source language and a machine translation of the source language sentence to a target language sentence. For each of the sentence pairs, a representation of the sentence pair is generated, based on a set of features extracted for the sentence pair. Using a generative model, a representation of the input document pair is generated, based on the sentence pair representations. A translation quality of the translated input document is computed, based on the representation of the input document pair.

Abstract: Disclosed are embodiments for use in an articulatory-based text-to-speech conversion system configured to establish an articulatory speech synthesis model of a person's voice based on facial characteristics defining exteriorly visible articulatory speech synthesis model parameters of the person's voice and on a predefined articulatory speech synthesis model selected from among stores of predefined models.

Abstract: A system and method for translating data from a source language to a target language is provided wherein machine generated target translation of a source sentence is compared to a database of human generated target sentences. If a matching human generated target sentence is found, the human generated target sentence may be used instead of the machine generated sentence, since the human generated target sentence is more likely to be a well-formed sentence than the machine generated sentence. The system and method does not rely on a translation memory containing pairs of sentences in both source and target languages, and minimizes the reliance on a human translator to correct a translation generated by machine translation.

Abstract: Natural language (NL) processing of NL commands is performed to determine one or more of (i) content to be included in a hypertext document, (ii) a layout for the hypertext document, or (iii) parameters corresponding to behavior of the hypertext document. Performing the NL processing includes: determining assumptions concerning one or both of (a) ambiguities associated with the NL commands or (b) information necessary for creating the hypertext document not specified by the NL commands. The assumptions correspond to the one or more of (i) content to be included in the hypertext document, (ii) the layout for the hypertext document, or (iii) the parameters corresponding to behavior of the hypertext document. Hypertext document code or data for displaying the hypertext document is generated based on the NL processing of the NL commands including the determined assumptions.

Abstract: According to one embodiment, an interpretation apparatus includes a translator, a calculator and a generator. The translator performs machine translation on a speech recognition result corresponding to an input speech audio from a first language into a second language to generate a machine translation result. The calculator calculates a word number based on a first time when the machine translation result is generated and a second time when output relating to a prior machine translation result generated prior to the machine translation result ends, the word number being 0 or larger. The generator omits at least the word number of words from the machine translation result to generate an abridged sentence output while being associated with the speech audio.

Abstract: A translation system includes translation subsystems by a plurality of translation policies with different translation output tendencies, performs forward translation and reverse translation by a consistent translation policy, and ranks the plurality of translation policies based on a similarity between the reverse translation sentences and the original sentence. As a result, a translation policy that best matches each input object to be translated is selected from the plurality of translation policies for each input sentence and, as a consequence, it is possible to cover a wide range of natures of sentences to be translated.

Abstract: A method for generating a phrase table for a target domain includes receiving a source corpus for a target domain and, for each of a set of comparative domain phrase tables, computing a measure of similarity between the source corpus and the comparative domain phrase table. Based on the computed similarity measures, a subset of the comparative domain phrase tables may be identified from the set of comparative domain phrase tables, and/or weights for combining them, and a phrase table is generated for the target domain based on the at least a subset of phrase tables.

Abstract: Aspects of the present invention disclose a method, computer program product, and system for parsing a domain-specific language (DSL) statement. The method includes one or more processors accessing a DSL statement that includes contracted phrases. The method further includes one or more processors identifying one or more contracted phrases in the DSL statement utilizing an annotated domain vocabulary for a DSL associated with the DSL statement and grammar rules for the DSL. The method further includes one or more processors determining expanded phrases corresponding to the identified one or more contracted phrases based on the annotated domain vocabulary and the grammar rules. The method further includes one or more processors creating an expanded abstract syntax tree (AST) that is representative of the DSL statement with the determined expanded phrases replacing the identified one or more contracted phrases.

Abstract: A system and method for assessing translation quality are disclosed. In one example, a computer system receives data associated with a translation from a first text in a first natural language to a second text in a second natural language, wherein the data comprises a first value that is based on one or more semantic errors of the second text, a second value that is based on extraneous content in the second text, and a third value that is based on missing content in the second text; and determining by the computer system, based on the first value, the second value, and the third value, a score that quantifies translation quality of the second text with respect to the first text.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for augmenting clusters of news articles with translated news articles. In one aspect, a method includes obtaining data characterizing a first cluster of electronically published news articles in a presentation language, wherein the data characterizing the first cluster of news articles comprises one or more terms in the presentation language and region data identifying a first geographical region associated with the first cluster; generating translated terms by translating one or more of the terms from the presentation language to a first language, wherein the first language corresponds to the first geographical region; and obtaining one or more news articles in the first language in response to a search query derived from the one or more translated terms.

Abstract: Disclosed herein is an ability to identify a supported language. A current language tag including one or more subtags, each subtag providing language identification information. In response to determining that a language referenced by the current language tag is not a supported language, accessing a stored set of mappings including one or more mappings, each mapping associating a first language tag with a second language tag, determining whether or not a mapping in the set of mappings includes the current language tag as one of the first and second language tags and a new language tag as the other of the first and second language tags, and using the new language tag as the current language tag to identify the supported language.

Abstract: In a publication system, such as an ecommerce system, machine translation translates a query in a first language to a second language to query an ecommerce database maintained in the second language and obtain a result set responsive to the query. Human activity feedback relating to the result set is detected. If the feedback is positive the system increases the probability that the translation is correct. If the feedback is negative the system decreases the probability that the translation is correct. For positive feedback, the system detects whether a clue is recognized in the query. If a clue is recognized the system increases the value of the clue for making the translation. The system may detect the identity of the product in the query, accesses the product vendor website that is maintained in the first language, and detect information that is in the first language for use translation process.

Abstract: Content items, such as e-books, audio files, video files, and the like, may be tagged as associated with a location based on observing the locations at which users access the content items. A rich set of tag data may be gathered by additionally observing such things as the date and time when users access the content items as well as allowing the users to tag the content items with comments or ratings. A fine granularity of tagging may be achieved by associating the tags with specific portions of the content items. Content recommendations based on the tags may be provided to other users when those users are in approximately the same location.

Abstract: Techniques for performing multi-style speech synthesis. The techniques include using at least one computer hardware processor to perform: obtaining input comprising text and an identification of a first speaking style to use in rendering the text as speech; identifying a plurality of speech segments for use in rendering the text as speech, the identified plurality of speech segments comprising a first speech segment having the first speaking style and a second speech segment having a second speaking style different from the first speaking style; and rendering the text as speech having the first speaking style, at least in part, by using the identified plurality of speech segments.

Abstract: A system that incorporates the subject disclosure may include, for example, receive user speech captured at a second end user device during a communication session between the second end user device and a first end user device, apply speech recognition to the user speech, identify an unclear word in the user speech based on the speech recognition, adjust the user speech to generate adjusted user speech by replacing all or a portion of the unclear word with replacement audio content, and provide the adjusted user speech to the first end user device during the communication session. Other embodiments are disclosed.

Abstract: A system and method for localizing a spoken dialog system is disclosed. Source data from a source language spoken dialog system is accessed, including semantic annotations and transcriptions of a plurality of utterances. The transcriptions are machine-translated into a target language. Semantic classifiers are trained on the machine translated transcriptions and the source language semantic annotations.

Abstract: Techniques are disclosed for generating phrases for selection by users communicating in an online virtual world environment. A phrase generation engine automatically constructs new phrases using pre-approved words from a dictionary of frequently used words and language-specific rules for phrase formation to ensure safety and supplement an existing database of commonly used, pre-approved phrases. Increasing the number of phrases that are available for selection by a user increases the user expressivity. Each word in the dictionary is annotated with semantic and grammatical information that constrains how the word is combined with other words to generate a new phrase. Each new phrase may also be tagged to enable translation into a different language so a phrase in a first language selected by a first user may be displayed in a second language to a second user.

Abstract: Systems and methods are described herein for selecting a closed captioning language track. Control circuitry may capture an image or video of a user while the user is performing a gesture of a sign language. The control circuitry may access a database comprising a plurality of entries corresponding to a plurality of gestures in respective sign languages. The control circuitry may compare a first pixel map corresponding to the image or video of the user with pixel maps corresponding to each of the plurality of entries in order to identify a preferred sign language. The control circuitry may receive metadata associated with a media asset comprising a plurality of closed captioning language tracks, each closed captioning language track comprising closed captioning for the media asset in a different language and select one of the closed captioning language tracks that corresponds to the preferred sign language.

Abstract: Some embodiments of an efficient string search have been presented. In one embodiment, a string of bytes representing content written in a non-delimited language is received, wherein the content has been classified into a predetermined category. In a single pass through the string of bytes, a set of N-grams is searched for simultaneously. Statistical information on occurrences of the N-grams, if any, in the string of bytes is collected. In some embodiments, a model is generated based on the statistical information, where the model is usable by a content filter to classify content.

Abstract: Disclosed are methods, systems, and computer-readable mediums for automatic training of a syntactic and semantic parser using a genetic algorithm. An initial population is created, where the initial population comprises a vector of parameters for elements of syntactic and semantic descriptions of a source sentence. A natural language compiler (NLC) system is used to translate the sentence from the source language into a target language based on the syntactic and semantic descriptions of the source sentence. A vector of quality ratings is generated where each quality rating in the vector of quality ratings is of a corresponding parameter in the vector of parameters. Quality ratings are evaluated according to specific criterion, which comprise parameters such as a BLEU score and a number of emergency sentences. A number of parameters in the vector of parameters are replaced with adjusted parameters.

Abstract: Methods and apparatus for language translation in a computing environment associated with a virtual application are presented. For example, a method for providing language translation includes determining languages of a user and a correspondent; determining one or more sequences of translators; determining a selected sequence of selected translators from the one or more sequences of the translators; requesting a change in virtual locations, within the computing environment associated with the virtual application, of one or more selected translator virtual representations of the selected translators to a virtual meeting location within the computing environment associated with the virtual application; and changing virtual locations of the one or more selected translator virtual representations to the virtual meeting location.

Abstract: Techniques are described for predicting the language of a text excerpt. The language prediction is accomplished by comparing n-grams of the text excerpt with n-grams of different language references. A probability is calculated for each n-gram of the text excerpt with respect to each of the language references. The calculated probabilities corresponding to a single language are then averaged to yield an overall probability corresponding to that language, and the resulting overall probabilities are compared to find the most likely language of the sample text.

Abstract: Computing systems are provided for facilitating communication to the hearing-impaired. The computing systems obtain and modify a video that includes a signer performing sign language about a topic. The video is then modified and presented to a viewer in such a way that the viewer can simultaneously view the sign language and the one or more objects corresponding to the topic. In some instances, the modified video is blended with another video corresponding to the same topic so as to enable a user to view the sign language at the same time as the corresponding video presentation. This can be done in real-time or with a time delay. In other instances, the modified video is rendered on a display through which the user is also viewing the one or more related items associated with the sign language.

Abstract: User interfaces receive a first plurality of user queries and result sets that are in a category, exhibit a constraint, and exhibit user behavior. Also received are a second plurality of user queries and result sets that that are in the category, exhibit the constraint, and exhibit user behavior. The second user queries and results are received either from a plurality of user interfaces coupled to a second system, or from the second system itself. Responsive to detecting that the first plurality of user queries and result sets and the second plurality of user queries and result sets satisfy respective thresholds, a signal indicates that at least one of the first plurality of user queries and at least one of the second plurality of user queries are translations of each other.

Abstract: A computer-implemented technique can include receiving a machine translation input specifying (i) a source text, (ii) a source language of the source text, and (iii) a target language for the source text, and obtaining a machine translation of the source text from the source language to the target language to obtain a translated source text. The technique can include determining whether to swap the source and target languages based on (i) the source text and (ii) at least one language model, and in response to determining to swap the source and target languages: swapping the source and target languages to obtain modified source and target languages, utilizing the translated source text as a modified source text, obtaining a machine translation of the modified source text from the modified source language to the modified target language to obtain a translated modified source text, and outputting the translated modified source text.

Abstract: A system is described herein which uses a neural network having an input layer that accepts an input vector and a feature vector. The input vector represents at least part of input information, such as, but not limited to, a word or phrase in a sequence of input words. The feature vector provides supplemental information pertaining to the input information. The neural network produces an output vector based on the input vector and the feature vector. In one implementation, the neural network is a recurrent neural network. Also described herein are various applications of the system, including a machine translation application.

Abstract: Computer-implemented systems and methods for extracting information during a human-to-human mono-lingual or multi-lingual dialog between two speakers are disclosed. Information from either the recognized speech (or the translation thereof) by the second speaker and/or the recognized speech by the first speaker (or the translation thereof) is extracted. The extracted information is then entered into an electronic form stored in a data store.

Abstract: Disclosed herein are systems, computer-implemented methods, and tangible computer-readable media for enriching spoken language translation with dialog acts. The method includes receiving a source speech signal, tagging dialog acts associated with the received source speech signal using a classification model, dialog acts being domain independent descriptions of an intended action a speaker carries out by uttering the source speech signal, producing an enriched hypothesis of the source speech signal incorporating the dialog act tags, and outputting a natural language response of the enriched hypothesis in a target language. Tags can be grouped into sets such as statement, acknowledgement, abandoned, agreement, question, appreciation, and other. The step of producing an enriched translation of the source speech signal uses a dialog act specific translation model containing a phrase translation table.

Abstract: There is provided a method that includes parsing a string of characters into its graphemes, and generating a pattern of characters that represents an abstraction of the graphemes. There is also provided a system that performs the method, and a storage device that contains instructions for controlling a processor to perform the method.

Abstract: An approach provides context-specific language translation services for electronic documents using templates. As used herein, the term “template” refers to a context-specific set of keywords that includes at least a first set of keywords in a first language and a corresponding second set of keywords in a second language that is different than the first language. As used herein, the term “language” refers to a human language and the term “keyword” refers to one or more words or phrases of any length. The context-specific set of keywords may include additional sets of keywords in other languages that correspond to the first and second sets of keywords. A language translation service uses the templates to translate electronic documents. The language translation service also provides capabilities for managing templates and a translation review capability for reviewing translated documents to identify errors. Templates can then be edited to address the translation errors.

Abstract: Systems and techniques disclosed herein include methods for de-quantization of feature vectors used in automatic speech recognition. A SIMD vector processor is used in one embodiment for efficient vectorized lookup of floating point values in conjunction with fMPE processing for increasing the discriminative power of input signals. These techniques exploit parallelism to effectively reduce the latency of speech recognition in a system operating in a high dimensional feature space. In one embodiment, a bytewise integer lookup operation effectively performs a floating point or a multiple byte lookup.

Abstract: In accordance with a first aspect of the invention, there is provided a method comprising receiving an input as part of a translation request from a requestor, performing a first translation of the input; wherein the first translation is a machine translation, returning the first translation to the requestor; and based on feedback on the first translation from the requestor performing the following (a) fragmenting the input into multiple translation jobs, (b) distributing the multiple translation jobs to a plurality of human translators; (c) generating a second translation of the input based on translations of the multiple jobs by the human translators; and (d) returning the second translation to the requestor.

Abstract: Technology is disclosed for creating and tuning classifiers for language dialects and for generating dialect-specific language modules. A computing device can receive an initial training data set as a current training data set. The selection process for the initial training data set can be achieved by receiving one or more initial content items, establishing dialect parameters of each of the initial content items, and sorting each of the initial content items into one or more dialect groups based on the established dialect parameters. The computing device can generate, based on the initial training data set, a dialect classifier configured to detect language dialects of content items to be classified. The computing device can augment the current training data set with additional training data by applying the dialect classifier to candidate content items. The computing device can then update the dialect classifier based on the augmented current training data set.

Abstract: Systems and methods are disclosed for searching across multi-lingual information. A user makes a query in a first language, and a group of documents that were previously machine-translated into the first language are searched for information responsive to the query. Contextual information derived can be used to improve the accuracy of the machine translation. Responsive documents are returned to the user. Alternatively, a query provided in a user's language may be translated into one or more other languages. Documents written in these languages can then be searched for information responsive to the appropriate translated query. Responsive documents can be translated into the user's language prior to providing them to the user.