International Business Machines Corporation

Lev Kozakov

Youssef Drissi

Yurdaer Nezihi Doganata

Genady Grabarnik

Moon Ju Kim

Larisa Shwartz

Moon Jung Kim

Tong-Haing Fin

Juan Leon Rodriguez

Gregory T Brown

Juan Leon-Rodriquez

Sheng Ma

Dikran S Meliksetian

Thomas Anthony Cofino

Cesar Augusto Gonzales

Jun-Jang Jeng

Zvi Yosef Yehudai

Mila Keren

Richard Kirchhofer

Yuval Ishai

Roberto J Bayardo, Jr

Juan L Rodriguez

Christopher Thomas Kello

Youssef Dressi

David Angelo Ferrucci

Pnina A Vortman

Samuel Kallner

Alexey Roytman

Uri Shani

John Williams Miller

Ricardo Vilalta

Yefim Haim Michlin

Juan-Leon Rodriguez

Noon Ju Kim

Blanche M Waddell

Galina Kofman

Volodimir F Lemberg

A document corpus is annotated by given annotators and aggregates of the given annotators to generate a corpus of pre-annotated documents with given annotations. The same document corpus is annotated by a subject annotator to have subject annotations. The subject annotations are decomposed into superpositions of the given annotations. The decomposed subject annotations are mapped to a most representative superposition of the given annotations. The subject annotator may be mapped to the aggregate of given annotators corresponding to the most representative superposition.

System and method of automatically mapping a given annotator to an aggregate of given annotators

Evaluating the performance of a text analysis engine is provided. A plurality of pre-annotated reference documents and a set of annotation types associated with the pre-annotated reference documents are received. Annotation contexts of reference annotations in the plurality of pre-annotated reference documents are analyzed using the set of annotation types. Similar annotation contexts are identified between the reference annotations and the set of annotation types. Responsive to identifying the similar annotation contexts, the similar annotation contexts are clustered thereby forming a plurality of reference annotation clusters. A set of reference content heterogeneity scores are computed based on the number of reference annotation clusters for each annotation type in the set of annotation types. An integral reference content rate for the set of annotation types is then computed and output to a user.

Credibility of text analysis engine performance evaluation by rating reference content

Techniques are presented that include sending a query to one or more information sources associated with a category comprising one or more terms, wherein said one or more information sources are identified by a ranked list of information sources, and wherein said ranked list is generated by ranking a plurality of information sources based on scores for results returned by said plurality of information sources in response to a ranking query generated from said category, and obtaining results from said one or more information sources.

Techniques for personalized and adaptive search services

Techniques are presented that include defining one or more personalized categories, wherein each personalized category comprises one or more terms, identifying one or more of said personalized categories that are associated with a user query, identifying one or more terms that are associated with said user query and with said personalized category, and adding said identified one or more terms to said user query.

A document corpus is annotated by given annotators and aggregates of the given annotators to generate a corpus of pre-annotated documents with given annotations. The same document corpus is annotated by a subject annotator to have subject annotations. The subject annotations are decomposed into superpositions of the given annotations. The decomposed subject annotations are mapped to a most representative superposition of the given annotations. The subject annotator may be mapped to the aggregate of given annotators corresponding to the most representative compatible superposition, resulting in a meaningful new aggregate annotator.

Under the present invention, when an application is deployed, certain information corresponding thereto will be stored in an Enterprise Application Directory (EAD). Thereafter, when an instance of the application is requested (e.g., by a requestor), the information is retrieved from the EAD. In addition, the instance is registered with an Application Access Registry (AAR). Registering the instance involves creating a record in the AAR, generating a unique application access key for the instance, and storing the access key with the information in the record. The access key (and optionally the information) is then returned to the requestor, which can then issue service requests to one or more grid services in the shared computer infrastructure. The service request will include the access key so that the grid service can identify the application by searching for the record in the registry using the access key.

Method, system and program product for managing applications in a shared computer infrastructure

In general, the present invention protects actual search queries submitted to web search engines using a set (i.e., at least one) of supplemental queries (hereinafter referred to as securing search queries). As a result, collections of search queries will not form statistically stable categories, and will not disclose the search subject. Any hits resulting from securing search queries will be filtered from results that are returned to the requestor. In addition, the securing search queries can be associated with protective Internet Protocol addresses to reduce the possibility of the requestor of the actual search query to be identified.

Securing search queries

Techniques are presented for automatically selecting information sources that are most relevant to user queries. Results of searches returned by information sources for queries are analyzed and the information sources are ranked based on this analysis. The information sources that have high rankings for a query are subsequently used to search for relevant results. This process can be adaptive, as the returned results of old queries can be analyzed at a later date to update the ranking of the information sources, automatic searches can be performed to update the ranking of the information sources, new queries can be used for analysis and stored, new information sources added, and old information sources deleted. A linguistic library is used to store personal categories for one or more users and general categories. Each category is associated with keywords and ranked lists of information sources. The library also contains general categories, taxonomies, and dictionaries.

Under the present invention, it is determined whether sufficient resources exist for deploying a platform dependent application on its specific 'native' platform in a grid environment. If not, a platform dependent portion of the application is identified, and the application is split into the platform dependent portion and a platform independent portion. Thereafter, the platform dependent portion is deployed on its corresponding native platform, while the platform independent portion is deployed on another platform in the grid environment based on available resources. Interconnections between the two portions can then be automatically generated. This can include, for example, creating remote method invocation facade interfaces, creating remote method invocation-enabled facade classes, and building a remote method invocation server for the first platform.

Method, system, and program product for deploying a platform dependent application in a grid environment

A method, software and apparatus are provided which enable a user to obtain products and services in a manner which counteracts conflicting actions modifying raw ranking data based on merchants interests. For this purpose, an information a user's web browser is provided with modules that modify search results obtained from one or more sources including directly from merchants. When the user enters a set of key words in an entry field in an on-screen form for a web server to obtain a list of items (products and/or services) of interest to the user, the search terms are modified in the users or his agents browser to reflect the interest of the user, his management or his parents. The user or his agent receives ranking information taking the highest ranked products and obtains the product information for the most highly ranked products. The browser then reranks the products using a ranking algorithm preferred by the user that offsets information sources biases in the ranking of the product information. The algorithm adds weighting factors reflecting the modified search terms. The weighting factors are combined with the mentioned ranking mechanisms to the documents to increase the probability that certain items come to the top when the search results are presented to the shopper.

Promoting strategic documents by bias ranking of search results on a web browser

A method for extending the service life of a portable device includes monitoring a battery of a portable device; identifying a problem bank; reconfiguring a connection schema for the battery to replace the problem battery bank with at least one spare bank; conditioning or exercising the problem bank; connecting the portable device to a power supply to recharge the problem bank; and reconnecting the recharged or repaired bank according to the connection schema without the at least one spare bank. A method for extending the service life of a portable device includes monitoring power consumption of at least one of the hardware or software of a portable device; and reconfiguring the connection schema of the battery banks to redistribute power consumption of at least one of the hardware or software.

System to extend the service life of portable devices

A system and method of searching a database in which documents of different languages are included. The system includes a synonym or keyword dictionary which is bi-directional and allows for translation of keywords between a first language and other languages. The translated words keywords for the document are stored in an inverted index which is then used for searching, either in a selected language, a second language or in all languages, as determined by the user. This use of multiple searching and a translated synonym dictionary avoids the need for translation of the entire document and avoids inaccuracies which may result from translations.

Method and system for searching a multi-lingual database

Searching a multi-lingual database

System and methods to extend the service life of portable devices

Growing community of inventors

Stamford, CT, United States of America

Lev Kozakov