121 results about "Approximate matching" patented technology

Improved techniques for enhancing, associating, and linking various sources of metadata for music files, to allow integration of commercially generated metadata with user-entered metadata, and to ensure that metadata provided to the user is of the highest quality and accuracy available, even when the metadata comes from disparate sources having different levels of credibility. The invention further provides improved techniques for identifying approximate matches when querying metadata databases, and also provides improved techniques for accepting user submissions of metadata, for categorizing user submissions according to relative credibility, and for integrating user submissions with existing metadata.

A method and system are provided for approximate string matching of a target string to a trie data structure. The trie data structure has a root node and generations of child nodes each node representing at least one character in an alphabet to provide a lexicon of words and word fragments. The method involves traversing the trie data structure starting from the root node by comparing each node of a branch of the trie data structure to characters in the target string and adding characters traversed in a branch of the trie data structure to a gathered string to provide suggestions of approximate matches. If the method reaches a node flagged as a node for a word or a word fragment and, if the target string is longer than the gathered string, the method loops back to the root node, and continues the traverse from the root node. This enables the trie data structure to use word fragments for compound words and to split non-delimited words where appropriate. The method also includes, at each node, determining if there is a correction rule for one or more characters in the remainder of the target string from the current node, and if so, applying the correction rule to the target string to obtain a modified target string.

Disclosed herein is a method and system for hardware-accelerating the generation of metadata for a data stream using a coprocessor. Using these techniques, data can be richly indexed, classified, and clustered at high speeds. Reconfigurable logic such a field programmable gate arrays (FPGAs) can be used by the coprocessor for this hardware acceleration. Techniques such as exact matching, approximate matching, and regular expression pattern matching can be employed by the coprocessor to generate desired metadata for the data stream.

The present invention relates generally to a system and method for categorization of strings of words. More specifically, the present invention relates to a system and method for normalizing a string of words for use in a system for categorization of words in a predetermined categorization scheme. A method for adaptive categorization of words in a predetermined categorization scheme may include receiving a string of text, tagging the string of text, and normalizing the string of text. Normalization may be performed with a three-stage algorithm including a literal match processing stage, an approximation match processing stage, and a nearest neighbor match processing stage. The normalized string of text can be compared to a number of sequences of text in the predetermined categorization scheme.

New language constantly emerges from complex, collaborative human-human interactions like meetings—such as when a presenter handwrites a new term on a whiteboard while saying it redundantly. The system and method described includes devices for receiving various types of human communication activities (e.g., speech, writing and gestures) presented in a multimodally redundant manner, includes processors and recognizers for segmenting or parsing, and then recognizing selected sub-word units such as phonemes and syllables, and then includes alignment, refinement, and integration modules to find or at least an approximate match to the one or more terms that were presented in the multimodally redundant manner. Once the system has performed a successful integration, one or more terms may be newly enrolled into a database of the system, which permits the system to continuously learn and provide an association for proper names, abbreviations, acronyms, symbols, and other forms of communicated language.

The method discloses a fuzzy matching-based Chinese geo-code determination method, which comprises the following steps: A1, reading descriptive Chinese address information in and adopting a forward maximum searching method to split an original address to obtain an original address element array in a way that the levels of administrative regions are taken as breakpoints; A2, standardizing original address elements through an address dictionary; and A3, reading a standard address tree, adopting a branch-bound algorithm to match the original address element array, simultaneously, utilizing fuzzy rules to control the matching operation, and after acquiring keywords after the original address is split, taking a matching result with the highest evaluation score as the most approximate matching result to obtain a more accurate matched address. The invention provides the fuzzy matching-based Chinese geo-code determination method, which has the advantages of rational address model, relatively higher matching rate and high speed.

The functions and operations of a matching service are disclosed. This includes approximating the satisfaction that a user of the matching service has in the relationships that the user forms with others and identifying candidates for a relationship with the user based on the approximated satisfaction. This also includes approximating the satisfaction that the user will have in a relationship with a particular candidate. The matching service also identifies two parties for a relationship. The matching service makes available a plurality of communication levels at which the parties can communicate. Each communication level allows the parties to exchange information in a different format. The parties are permitted to exchange information at one of the communication levels.

Establishments are identified in geo-tagged images. According to one aspect, text regions are located in a geo-tagged image and text strings in the text regions are recognized using Optical Character Recognition (OCR) techniques. Text phrases are extracted from information associated with establishments known to be near the geographic location specified in the geo-tag of the image. The text strings recognized in the image are compared with the phrases for the establishments for approximate matches, and an establishment is selected as the establishment in the image based on the approximate matches. According to another aspect, text strings recognized in a collection of geo-tagged images are compared with phrases for establishments in the geographic area identified by the geo-tags to generate scores for image-establishment pairs. Establishments in each of the large collection of images as well as representative images showing each establishment are identified using the scores.

Input data queries directed at a plurality of target databases and originating from any of a plurality of sources are first converted to validated canonical forms, which are then used to query the target databases. Specifically, upon receiving an input data query, a relatively accurate reference database is selected based on the type of the input data. This reference is then queried for the input data with the intent of finding an exact matching record or a near-matching record that can be considered an exact match and thereby validating the input data. Otherwise, the requesting source is instructed to provide a new query. Once having a validated record, it is converted to a canonical form, which is then used to query the target databases intended to be searched. In a further embodiment, multiple reference databases are queried to determine a canonical form of the data or to determine multiple canonical forms of the data.

A method of updating a computer file from an old file into a new file comprises blocking the new file and the old file into fixed-size blocks, maintaining a window (collection of contiguous blocks) for each file on which lookup preprocessing has been performed, and performing match processing on each new file block in turn (comparing against both the old and new windows) using a key-sampling technique combined with approximate matching. For each new file block, the match information is then optimized for coding efficiency and encoded into a patch file that describes an algorithm for converting the old file into the new file. The patch file application method and apparatus then performs the algorithm described in the patch file. The method uses a fixed amount of random-access memory regardless of the sizes of the two files and uses no temporary mass storage. In addition, the method has a running time roughly proportional to the size of the new file and allows the use of parallel processing to reduce the time required. The system and method produce patch files which are smaller than prior systems and methods, and allow the operator of the apparatus to perform an efficiency/effectiveness trade-off.

The present invention automatically builds a contracted dictionary from a given list of multi-word proper names and performs fuzzy searches in the contracted dictionary. The contracted dictionary of proper names includes two linked trie-based dictionaries: a first dictionary is used to store single word names, each word name having an ID number; and a second dictionary is used to store multi-word names encoded with ID numbers. Information related to the multi-word names is also stored as a gloss to the terminal node of the multi-word entry of the trie-based dictionary. An approximate lookup for a multi-word name is conducted first for each word of the multi-word name using an approximate matching technique such as a phonetic proximity or a simple edit distance. Accordingly, N suggestions is determined for each word of the multi-word name under consideration. Then, multi-word candidates are assembled in ID notation. Finally, an approximate search for each assembled candidate is performed based on an edit distance or a n-grams approximate string matching. Edit distances and N-grams are used to measure how similar two strings are. The result is a set of multi-word suggestions in an ID notation. This ID notation is encoded back to the original form using the first trie-based dictionary.

A system and method for matching data using probabilistic modeling techniques is provided. The system includes a computer system and a data matching model/engine. The present invention precisely and automatically matches and identifies entities from approximately matching short string text (e.g., company names, product names, addresses, etc.) by pre-processing datasets using a near-exact matching model and a fingerprint matching model, and then applying a fuzzy text matching model. More specifically, the fuzzy text matching model applies an Inverse Document Frequency function to a simple data entry model and combines this with one or more unintentional error metrics/measures and/or intentional spelling variation metrics/measures through a probabilistic model. The system can be autonomous and robust, and allow for variations and errors in text, while appropriately penalizing the similarity score, thus allowing dataset linking through text columns.

Data objects are related by comparing attributes of data objects that belong to different clusters and determining that the data objects are an approximate match based on the comparison. Data elements corresponding to assignments of an identifier are generated, and the data elements are stored in a grouping.

A method and system for approximate string matching are provided for generating approximate matches whilst supporting compounding and correction rules. The method for approximate string matching of an input pattern to a trie data structure, includes traversing a trie data structure to find approximate partial and full character string matches of the input pattern. Traversing a node of the trie data structure to process a character of the string applies any applicable correction rules to the character, wherein each correction rule has an associated cost, adjusted after each character processed. The method includes accumulating costs as a string of characters is gathered, and restricting the traverse through the trie data structure according to the accumulated cost of a gathered string and potential costs of applicable correction rules.

The invention provides efficient searching with fuzzy criteria in very large information systems. The technique of the present invention uses the Pigeonhole Principle approach. This approach can be utilized with different embodiments. but the most effective realization would be to amplify some already given intrinsic approximate matching capabilities, like those in the FuzzyFind method [1][2]. Considering the following problem, data to be searched is presented as a bit-attribute vector. The searching operation includes finding a subset of this bit-attribute vector that is within particular Flamming distance. Normally, this search with approximate matching criteria requires sequential lookup for the whole collection of the attribute vector. This process can be easily parallelized, but in very large information systems this still would be slow and energy consuming. The present invention provides approximate search in very large files using the Pigeonhole Principle, circumvents the sequential search operations and reduces the calculations tremendously.

A method of reducing storage capacity needed to store a target image in a large database of images is presented. A target image is uploaded from a client system to a server system, which are connected through an Internet connection. An image index is queried to find an approximate match to the target image and to identify a most similar reference image to a processed target image stored in an image database. The difference between the target image and a raw image corresponding to the identified most similar reference image is encoded. A pointers corresponding to the processed reference image stored in the image index is updated to reflect the newly stored target image in the image index.