Abstract: A machine learning based system and method for automated recognition of consumer products in images and video using a camera system, a neural network using a ranked tagging system, a two-stage recognition application, and a training module. Training image sets of items captured by the camera system are assigned identification tags through template matches to training sets within the neural network. Tags are assigned from various levels of specificity to identify exact product matches. A user recognition application captures images and generates bounding boxes for detected objects and assigns a general classification within the image using a single, fast, convolutional neural network (CNN) layer. General classification narrows subsets for each generated bounding box and multi-scale template matching is applied to achieve detailed identification of single or multiple items detected within single images or video. The training module adjusts smart camera systems and the neural network based on accuracy feedback.

Abstract: Systems and methods for processing events are disclosed. Event data comprising passive event data, active event data, or both is received. It is determined whether the received event data is available for a pattern of passive event data and active event data. In response to determining that the received event data is available for the pattern of passive event data and active event data, one or more constraints between the passive event data and the active event data are converted into one or more query terms. The query terms are used to construct at least one query. Remaining passive event data that is related to some, but not all, of the active event data is obtained using the constructed at least one query.

Abstract: Method, system, and programs for providing enhanced search term suggestions. A set of incomplete search terms indicating a sequence of search terms entered may be received. It may be detected that the sequence contains a descending phase followed by an ascending phase. In response to the detection, a pair of misinput term and corresponding corrected term may be identified in the set of incomplete search terms. A probability with respect to the misinput term is a misinput of the corresponding corrected term may be determined based on a historical context. Using such a probability, an incomplete search term containing the misinput term may be corrected. One or more proposed search terms may be determined based on the corrected incomplete search term for suggestion to the user.

Abstract: Techniques are provided for automatically detecting homologies between documents based on structural characteristics. Various statistics relating to the COS structure of a PDF document are compiled. The statistics are input into a rolling hash function to generate a digital fingerprint of the document. Fingerprints from two similar documents will have small edit distances between them, and can therefore be classified similarly or provided as results to a fingerprint-based search. For example, an unclassified document may be classified in the same class as a representative document where the fingerprints of the two documents have a small edit distance between them. Since the structure of the document is used instead of the text content or renderings, it is possible to operate on encrypted documents. Further, representative elements of a particular class of documents can be selected for comparison against a target document for improved resolution of the results.

Abstract: In an embodiment, a method comprises performing a first search based on a first text string; based on the first search, identifying one or more search results; after identifying the one or more search results, causing the one or more search results to be displayed; while the one or more search results are displayed: receiving input that changes the first text string to a modified text string; in response to receiving the input: performing a second search based, at least in part, on a portion of the modified text string, and determining that no search results are identified based on the second search; after determining that no search results are identified based on the second search, continuing to display the one or more search results; wherein the method is performed by one or more computing devices.

Abstract: Method and apparatus for managing versions of a portion of a document on a computing system are described. Multiple versions of an object in the document are maintained on the computing system. A graphical representation of at least two versions of the object is displayed on a display to a user of the computing system. A previous version of the object is selected from the at least two versions of the object through an input device, and the object in the document is restored to the selected previous version. The object is a portion of the document less than the entirety of the document.

Abstract: A remote processor interacts with a transactional memory that has a memory, local BWC (Byte-Wise Compare) resources, and local NFA (Non-deterministic Finite Automaton) engine resources. The processor causes a byte stream to be transferred into the transactional memory and into the memory. The processor then uses the BWC circuit to find a character signature in the byte stream. The processor obtains information about the character signature from the BWC circuit, and based on the information uses the NFA engine to process the byte stream starting at a byte position determined based at least in part on the results of the BWC circuit. From the time the byte stream is initially written into the transactional memory until the time the NFA engine completes, the byte stream is not read out of the transactional memory.

Abstract: Technologies and implementations for allocating datacenter resources are generally disclosed.

Abstract: Systems, methods, and machine-readable media for prioritizing points of interest for display in a map have been described. In certain aspects a system may include an interface module, a search module, and a priority module. The interface module may be configured to obtain pattern recognition data and location data associated with an image of a geographical area. The search module may be configured to access a set of point of interest listings and determine whether the pattern recognition data and the location data associated with the image correspond with a point of interest listing in the set of point of interest listings. The priority module may be configured to prioritize the point of interest listing for display in a map if a point of interest listing in the set of point of interest listings is determined to correspond.

Abstract: A method for analyzing a character string, the method including: analyzing a character string to determine one of more characters of the character string; determining from a dictionary source, an alternative character string to the analyzed character string; comparing the analyzed character string with the alternative character string to determine a weighting factor for each of the characters of the analyzed character string relative to the positional arrangement of the characters in the alternative character string; and for each determined weighting factor, generating for each of the characters in the analyzed character string a corresponding character of a particular size as determined by the weighting factor.

Abstract: One embodiment relates to a computer-implemented method for multiple-keyword matching performed using a computer including at least a processor, data storage, and computer-readable instructions. A keyword set and a text input to be searched are obtained. The keyword set is processed to create a reverse trie. A search procedure which starts from the end of the text is then applied using the reverse trie to find keyword occurrences in the text input. Other embodiments, aspects, and features are also disclosed.

Abstract: Systems and methods to extract a string from a communication. A method includes: receiving a communication comprising a plurality of strings; assigning a score to each of the strings, wherein the score assigned to each of the strings corresponds to a frequency of usage of the respective string for a first function relative to an overall frequency of usage of the respective string; determining a respective total sum for each of a plurality of sequences in the communication, the respective total sum determined as a sum of the scores for each string in the respective sequence; and extracting a first sequence of the sequences from the communication based on the total sum for the first sequence. In one embodiment, the total sum includes an additional score for each of a starting word and an ending word of the first word sequence, wherein each respective additional score is associated with a probability that the starting (or ending) word is used as the first (or last word) of an address.

Abstract: A system, method, and computer program product are provided for performing a string search. In use, a first string and a second string are identified. Additionally, a string search is performed, utilizing the first string and the second string.

Abstract: A method for submatch extraction may include receiving an input string, receiving a regular expression, and converting the regular expression with capturing groups into a plurality of finite automata to extract submatches. The method further includes using a first automaton to determine whether the input string is in a language described by the regular expression, and to process the input string, and using states of the first automaton in a second automaton to extract the submatches. In addition, input partitioning and automaton minimization techniques may be employed to reduce the storage area consumed by the plurality of finite automata.

Abstract: In some embodiments, a method includes receiving a command to update a form component within a set of one or more form templates. The method also includes receiving an alternative form component for the form component that is to be updated. The method includes performing a pattern matching across a number of form templates to identify the set of one or more form templates in the number of form templates. The method includes refactoring the set of one or more form templates. The refactoring includes changing a reference in the set of one or more form templates to reference the alternative form component.

Abstract: An embodiment for evaluating rules pertaining to an event includes providing a rule trie comprising multiple nodes, wherein match conditions must be met to trigger a given rule, and each match condition corresponds to a different node. An attribute trie is provided for a string attribute, which includes a node sequence having a value. The string attribute is extracted from the event and also has a value. The value of the extracted string attribute and the value of the node sequence are compared, by traversing a path extending along the node sequence. Responsive to determining that the values of the extracted string attribute and the node sequence are the same, notice is provided that the match condition of a particular node of the rule trie has been met by an attribute of the received event.

Abstract: Website and newsletter operators may only have users' email addresses but not their person names. It can be beneficial for such operators to derive additional information about the email address owners so they can improve their services. However, there is no easy way to retrieve owners' information based on only email addresses. The present disclosure provides information such as the owners' names, professional profiles and aggregated statistics with only a list of email addresses as the input.

Abstract: A regular expression matching hardware implementation comprises two tightly coupled hardware engines. A regex engine performs state transitions and accepts (as matching) or rejects (as not matching) an input string. The regex engine takes also care of the logic of the operators and deals with the complexity of the state machine. A lookup engine reads characters from an input (e.g., tape, memory, network packets, or the like), and provides them to the regex engine. A preprocessing procedure transforms a regular expression into a regex state transition table and a lookup table, for use by the regex engine and lookup engine, respectively. The two hardware engines are synchronized by a global state machine. The regex engine advances the state machine, and the lookup engine reads it.

Abstract: A method includes identifying at least one document associated with content from at least one digital content source. The at least one document includes information identifying at least one aspect of the content. The method also includes determining a document index for the at least one document based on keywords included in the at least one document. An activity field is inserted into the document index. The method includes accessing activity information. The activity information identifies at least one activity keyword associated with at least one activity. The method further includes identifying at least one present activity keyword in the document based on the activity information. An indicator of at least one present activity is determined based on the at least one present activity keyword. The method includes indexing the indicator of the at least one present activity in the activity field of the document index.

Abstract: A record-pattern expression of a record pattern of data records is translated to a character-pattern expression of a character pattern. A data-record series of data records is converted to a character string in which each character in the character string corresponds to a respective data record in the data-record series. Pattern instances of the character pattern in the character string are searched to identify ordinal character positions in the character string of characters in the character pattern instances. Data records at ordinal record positions in the record series of data records corresponding to the ordinal character positions are reported.

Abstract: In an embodiment, a method comprises performing a first search based on a first text string; based on the first search, identifying one or more search results; after identifying the one or more search results, causing the one or more search results to be displayed; while the one or more search results are displayed: receiving input that changes the first text string to a modified text string; in response to receiving the input: performing a second search based, at least in part, on a portion of the modified text string, and determining that no search results are identified based on the second search; after determining that no search results are identified based on the second search, continuing to display the one or more search results; wherein the method is performed by one or more computing devices.

Abstract: A method of operation of a navigation system includes: detecting a search string; determining a forward-element using a forward-boundary within the search string in a forward direction; determining a backward-element in a backward direction up to the forward-boundary using a backward-boundary within the search string; and determining a matching result for the search string using the forward-element and the backward-element for displaying on a device.

Abstract: A method accomplishes matching a particular media metadata record to one of a plurality of records of a first media metadata database. For the particular media metadata record, for each of at least some of the tokens of the particular media metadata record, a table is processed that correlates that token to records of the first media metadata database. Based on the processing, it is determined in which of the plurality of records (match candidate records) of the first media metadata database that token of the particular record appears. For each of at least some of the match candidate records of the first media metadata database, a value is determined that indicates a degree of match between that match candidate record and the particular media metadata record. The determined values are processed to determine one of the match candidate records is a nominal match to the particular media metadata record.

Abstract: An information processing method for defining a data structure for pattern matching, the method executed by an information processing apparatus, comprises generating, by the apparatus, an ordered tree structure by defining transition edges between nodes using, as transition conditions, respective constraints from one or more constraint patterns each including plural constraints; searching, by the apparatus, for a second substructure similar to a first substructure from a root node by determining a set relation between transition conditions of respective transition edges; and defining, by the apparatus, an additional transition link from a tail node of the second substructure to a child node at a tail end of the first substructure, the additional transition link adding a constraint to be met by an indeterminant identified from the set relation.

Abstract: A system and method for searching a product model number database having a plurality of product model numbers stored therein receives a string to be searched in the product model number database from a client device. Multiple queries are then created based on the string and the model number database is queried using the created queries. Query results responsive to the created queries are then received from the model number database and transmitted to the client device.

Abstract: An entity summarization system is described herein that mines the Internet and other data source to provide answers to questions such as the relative sentiment of users towards various brands. The system uses a controlled vocabulary list describing a specific aspect of entities of interest. Given an entity name, the system scans the whole content corpus to collect statistics on the words that occur most frequently in the context of the entity name, taking into account proximity information, to produce a weighted list of vocabulary terms describing the entity. Two entities can be compared by normalizing and comparing their weighted term lists. In some embodiments, the system performs these procedures efficiently by leveraging an N-gram web model. Thus, the system provides an automated way to compare two entities to derive information about how users feel about the entities at any given time.

Abstract: Techniques for performing searches using models generated from data mining. These techniques may include receiving a query including a search string. If a portion of the search string matches with a category of relationship data, the portion of the search string may be replaced with a label of the category to generate a modified query. Based on the modified query and models generated from mining historical search data, intent of the query may be determined, and searches may be then performed.

Abstract: An information processing apparatus according to the present invention comprises a generation section configured to generate an ordered tree structure by defining transition edges between nodes using, as transition conditions, respective constraints from one or more constraint pattern each including plural constraints, a search section configured to search for a second substructure K similar to a first substructure P from a root node by determining a set relation between transition conditions of respective transition edges, and a transition defining section configured to define an additional transition link from a tail node sKn of the second substructure K to a child node sPn+1 of the first substructure P, the additional transition link adding a constraint to be met by an indeterminant identified from the set relation, wherein a data structure for pattern matching is defined.

Abstract: One or more techniques and/or systems are disclosed for matching a target string to a known string. A target string is broken into one or more target terms, and the one or more target terms are matched to known terms in an index tree. The index tree comprises one or more known terms from a plurality of known strings, where the respective known terms in the index tree are associated with one or more known string IDs. A known term that is associated with a known string ID (in the index tree, and to which a target term is matched), is comprised in a known string, which corresponds to the known string ID. The target string can be matched to the known string using the known string's corresponding known string ID that is associated with a desired number of occurrences in the matching of the one or more target terms.

Abstract: Systems and methods are disclosed for comparing selections of text to show differences between the two selections. The text may be selected from the same source or from two different sources. In one implementation, a system receives a first selection of text for comparison and places the selection in a first buffer. The system receives a second selection of text for comparison and places the second selection in a second buffer. The system compares the first buffer and the second buffer to determine differences and displays the differences. In some embodiments, the system may allow a user to choose two buffers from among a plurality of buffers for comparison.

Abstract: A method and system are disclosed for conducting text-based searches of images using a visual signature associated with each image. A measure of string similarity between a query and an annotation associated with each entry in a first database is computed, and based upon the computed string similarity measures, a set of entries from the first database is selected. Each entry of the first database also includes an associated visual signature. At least one entry is then retrieved from a second database based upon a measure of visual similarity between a visual signature of each of the entries in the second database and the visual signatures of the entries in the selected set. Information corresponding to the retrieved entries from the second database is then generated.

Abstract: A spatial and temporal memory system (STMS) processes input data to detect whether spatial patterns and/or temporal sequences of spatial patterns exist within the data, and to make predictions about future data. The data processed by the STMS may be retrieved from, for example, one or more database fields and is encoded into a distributed representation format using a coding scheme. The performance of the STMS in predicting future data is evaluated for the coding scheme used to process the data as performance data. The selection and prioritization of STMS experiments to perform may be based on the performance data for an experiment. The best fields, encodings, and time aggregations for generating predictions can be determined by an automated search and evaluation of multiple STMS systems.

Abstract: Methods and systems for obtaining a scalable variable-length nonce are provided. An exemplary method for obtaining a scalable variable-length nonce includes obtaining an initial nonce candidate. The exemplary method also includes transmitting a query to determine whether the initial nonce candidate has already been assigned. When the initial nonce candidate has already been assigned, a combined nonce candidate is obtained. The combined nonce candidate includes characters appended to the initial nonce candidate. An exemplary system for obtaining a scalable variable-length nonce includes a transmitting engine that transmits a request associated with a current token. The exemplary system also includes a receiving engine that receives a message associated with the current token. The exemplary system further includes a token acquiring engine that obtains a new token, and an appending engine that appends the new token to the current token.

Abstract: Systems and methods are disclosed for identifying objects from a collection of objects of diverse types that match a pattern. In one implementation, a system includes a processor and a memory. The memory stores instructions that cause the processor to receive a pattern input text identifying the pattern to be matched. The pattern input text comprises one or more conditions and a match statement. The processor also receives a collection of objects of diverse types, where each object comprises at least an attribute and a value associated with the attribute. The processor analyzes each object to determine whether the object satisfies the match statement based on an application of the one or more conditions to the attributes and attribute values for the object. If the processor determines that the object satisfies the match statement, the processor creates an output comprising an indication of the object and the conditions that caused the match statement to be satisfied.

Abstract: A system for frequent pattern mining uses two layers of processing: a plurality of computing nodes, and a plurality of processors within each computing node. Within each computing node, the data set against which the frequent pattern mining is to be performed is stored in shared memory, accessible concurrently by each of the processors. The search space is partitioned among the computing nodes, and sub-partitioned among the processors of each computing node. If a processor completes its sub-partition, it requests another sub-partition. The partitioning and sub-partitioning may be performed dynamically, and adjusted in real time.

Abstract: Systems and methods are described for efficiently processing, searching and/or rewriting variable width encoded data, such as UTF-8 encoded data, will be described. Embodiments of the systems and methods modify and adapt search algorithms, such as the Horspool and Wu-Manber algorithms, to efficiently process and manage searching of variable width encoded text in large blocks of text, such as text that may be carried via a stream of packets thru a network device, such as an intermediary device.

Abstract: String matching is a ubiquitous problem that arises in a wide range of applications in computer science, e.g., packet routing, intrusion detection, web querying, and genome analysis. Due to its importance, dozens of algorithms and several data structures have been developed over the years. A recent breakthrough in this field is the FM-index, a data structure that synergistically combines the Burrows-Wheeler transform and the suffix array. In software, the FM-index allows searching (exact and approximate) in times comparable to the fastest known indices for large texts (suffix trees and suffix arrays), but has the additional advantage to be much more space-efficient than those indices. This disclosure discusses an FPGA-based hardware implementation of the FM-index for exact and approximate pattern matching.

Abstract: Various embodiments of the present invention provide systems and methods for data processing. For example, some embodiments of the present invention provide data processing circuits including a pattern detection circuit having at least two data detector circuits each operable to receive the same series of data samples and to provide a first detected data output and a second detected data output, respectively. In addition, the data pattern detection circuit includes a result combining circuit that is operable to assert a pattern found output based at least in part on the first detected data output and the second detected data output.

Abstract: The present disclosure relates to identifying consuming entity behavior across domains. Identity records are stored in a memory accessible to a computing system. Each of the identity records comprises at least one attribute, and the identity records originate from a plurality of domains. A determination is made as to whether a first one of the identity records identifies a consumption entity that is identified by a second one of the identity records. The first and the second identity records originate from separate ones of the domains, and the second one of the identity records is associated with a persistent key. The persistent key is associated with the consumption entity. The first identity record is associated with the persistent key if the first identity record is determined to identify the consumption entity.

Abstract: A data structure comprises a clump header table and an inline tree data structure. The inline tree, representing filterable data fields of hierarchically organized data records, comprises an alternating sequence of first-level binary string segments, each followed by one or more corresponding second-level binary string segments. Each clump header record includes an indicator of a location in the inline tree of corresponding binary string segments. A dedicated, specifically adapted conversion program generates the clump header file and the inline tree for storage on any computer-readable medium, and the inline tree can be read entirely into RAM to be searched or filtered. A dedicated, specifically adapted search and filter program is employed to list or enumerate retrieved data records. Run-time computer code generation can reduce time required for searching and filtering. One example includes spatial searching and filtering of data records that include spatial coordinates as data fields.

Abstract: Proposed is a method for naturally and efficiently setting metadata in the daily process of searching for files. On a file search screen, there is provided a function of determining the candidate metadata of a metadata-nonregistered file, and initiating entry of metadata with the candidate metadata having been set. Determination of the candidate metadata is performed with any of the three following methods: a method of designating as a candidate a character string of a matched search keyword described in regular expression, a method of designating as a candidate a file path or a character string in a file that matches a keyword dictionary, and a method of designating as a candidate metadata that frequently appears in metadata-registered files.

Abstract: There is provided a technique for automatically acquiring metadata with respect to various organizations which significantly reduces the man-hours required to prepare models for metadata extraction. With a pair comprising a document and metadata appearing therein as input, using a layout feature, and proximate text string and partial text string features with respect to metadata and a text string that is not metadata, the use of the layout feature, the proximate text string and the partial text string with respect to the automatic acquisition of metadata is automatically configured (see FIG. 1).

Abstract: Database elements are inserted into a database object by processing each of a plurality of operations in a sequential order within a first processing round to insert the database elements into the database objects, where processing for at least one operation in the order becomes suspended due to a resource request, and where at least one successive operation is initiated in response to suspension of one or more prior operations to enable prefetching of information for processing the operations. Each suspended operation is re-processed with the prefetched information in one or more additional processing rounds until processing of the operations is completed.

Abstract: An information retrieving apparatus is provided which has excellent usability using voice inputs. Search queries input by a user via a touch panel or the like are used as voice recognition vocabulary. Further, the search queries are edited by use of morpheme information and/or other information in a database, thereby providing such voice recognition vocabulary that can be easily uttered by the user.

Abstract: A system and method for searching a document for a query pattern. A plurality of streams may be stored each including a linear sequence of nodes. Each stream may be associated with nodes having a common label in a data tree of the document. A query pattern may be searched for in the streams by executing a plurality of threads. Each of two or more of the threads may be used to search different sub-streams of the plurality of streams. Each of the different sub-streams searched for by each thread in each stream may be uniquely correlated with one or more disjoint sub-trees of a partition of the tree into a plurality of sub-trees. The two or more of the plurality of threads may be executed in parallel. A result of the query pattern search may be generated using at least one of the threads.

Abstract: A comparison device for comparing test pattern files of a wafer tester includes a storage unit and a processing unit. The comparison device stores a first to-be-compared file and a second to-be-compared file into the storage unit. The first to-be-compared file and the second to-be-compared file are text files respectively. The processing unit reads the first to-be-compared file and the second to-be-compared file from the storage unit to process and executes comparison operation, so as to generate a comparison result. The comparison operation is to compare the words in a first section of the first to-be-compared file with the words in a second section of the second to-be-compared file in a one-to-one manner.

Abstract: The present invention extends to methods, systems, and computer program products for decomposing and merging regular expressions. Embodiments of the invention decompose a regular expression into multiple simple keyword graphs, merge those keyword graphs in a compact and efficient manner, and produce a directed acyclic graph (DAG) that can execute a simplified regular expression alphabet. Several of these regular expression DAG's can then be merged together to produce a single DAG that represents an entire collection of regular expressions. DAGs along with other text processing algorithms and a heap collection can be combined in a multi-pass approach to expand the regular expression alphabet.

Abstract: An apparatus for detecting presence of a selected complex string in a data stream has a primary processor, detecting coherent strings of the selected complex string; and a secondary processor, pipelined to the primary processor, and detecting ambiguous strings interleaving said coherent strings. The primary processor is configured to determine a number of coherent strings in the selected complex string, each coherent string preceded by a prefix of indefinite characters, with a last coherent string succeeded by a suffix of indefinite characters; and to locate sequential portions of the data stream that are identical to successive coherent strings of the selected complex string. The secondary processor is configured to ascertain congruence of portions of the data stream preceding and interleaving the sequential portions to prefixes of the coherent strings; and to ascertain congruence of the suffix to a corresponding portion of said data stream following the sequential portions.

Abstract: The present disclosure is directed to systems and methods including receiving a source code at a parser, the source code comprising a statement, the statement comprising a dynamic character string including one or more relation names, each of the one or more relation names defining a relation between business objects, processing the statement using one or more processors to access at least one business object, the at least one business object being stored in machine-readable memory, the at least one business object corresponding to at least one of the one or more relation names, and having business object data associated therewith, and generating an output based on the business object data.

Abstract: A behavior pattern recognition method, system and a computer application program thereof are presented. The method is applicable to an electronic device which has a storage unit for storing multiple sets of behavior record information, and the method includes the following step. Firstly, a first detecting unit acquires first behavior feature information, a collaboration network module acquires at least one second detecting unit having coherence with the first detecting unit. Then, the at least one second detecting unit acquires at least one second behavior feature information, and a processing unit compares the at least one second behavior feature information and the behavior record information to generate at least one comparison result. Finally, a behavior definition represented by the first behavior feature information is determined according to the comparison result.