International Business Machines Corporation

Biscom Inc.

Daniel M Jamrog

Kristin A Witherspoon

Jason D LaVoie

Nicholas W Orrick

Donna Karen Byron

Renee F Decker

Suzanne L Estrada

John A Morganti

Aditya S Gaitonde

Samir J Patel

Joseph F Zaffarano

Corville O Allen

Elie Feirouz

Chester E Ryder, Iii

Stanley John Vernier

Alexander Pikovsky

Elinna Shek

Chengmin Ding

Octavian Florin Filoti

David S Kern

Daniel C Gurney

A system for question answering (QA) documents data ingestion decides to ingest the documents data through a first plurality of sub-pipelines including a first sub-pipeline having a first set of engines and a second sub-pipeline having a second set of engines being independent from the first set of engines. The system determines a subset of the documents data and decides to ingest the subset through a second plurality of sub-pipelines including a third sub-pipeline having a third set of engines and a fourth sub-pipeline having a fourth set of engines being independent from the third set of engines. A set of engines of the second plurality of sub-pipelines and a set of engines of the first plurality of sub-pipelines are in a common class. The system selects output data from the second plurality of sub-pipelines over corresponding output data from the first plurality of sub-pipelines and generates a knowledge base.

Plug-and-ingest framework for question answering systems

Embodiments are directed to a question and answer (QA) pipeline system that adjusts answers to input questions based on a user criteria, thus implementing a content-based determination of access permissions. The QA system allows for information to be retrieved based on permission granted to a user. Documents are ingested and assigned an access level based on a defined information access policy. The QA system is implemented with the defined information access policy, the ingested documents, and the inferred access levels. For the QA system implementation, a user enters a question; primary search and answer extraction stages are performed; candidate answer extraction is performed using only content the user is allowed to access; the candidate answers are scored, ranked, and merged; ranked answers based on user permissions are filtered; and answers are provided to the user.

Method to allow for question and answer system to dynamically return different responses based on roles

An approach is provided to adjust time dependent terminology in a question and answering (QA) system. The approach ingests a set of documents to produce a corpus utilized by the QA system. A base time is established and the approach acquires a temporally accurate lexicon of terms that correspond to the base time. A corpus of the QA system is updated according to the lexicon. The QA system answers a question according to the updated corpus.

Adjusting time dependent terminology in a question and answer system

Aspects include creating a profile for a user from user account data and a user activity log, searching an indexed repository for keywords matching elements from the user account data and the user activity log, and retrieving documents from the indexed repository that are identified based on the searching. Aspects further include creating a custom portal for the user that includes the documents.

Dynamic document collection and custom portal creation

Aspects of processing misaligned annotations include receiving a tokenized document and offset annotation file at a processor. The tokenized document includes a source document and corresponding tokens resulting from a low-level segmentation process. Annotations from the annotation file are applied, in conjunction with tokenization rules, to the source document, and a misalignment responsive to the applying is determined. If the misalignment is caused by an offset mismatch, an offset number of characters between the position counts in the annotation file and the source document is calculated, and the position count in the annotation file is adjusted to coincide with the position count in the source document. If the misalignment is not caused by an offset mismatch, a current position count in the source document is reset to a position count of a previous location in which a most recent alignment between the annotation file and the source document was ascertained.

Misaligned annotation processing

A mechanism is provided, in a data processing system comprising a processor and a memory configured to implement a question and answer (QA) system, for weighting search criteria based on similarities to an ingested corpus in the QA system. A set of question characteristics found in a received input question are compared to a set of data characteristics respectively describing data in each corpus of a corpora. For each question characteristic in the set of found question characteristics, a first weight is assigned to the corpus within which data associated with the data characteristic resides in response to the question characteristic being more related to a data characteristic; otherwise a second weight is assigned, where the first weight is greater than the second weight. A selective search is then performed for an answer to the received input question in one or more corpora with a higher weighting.

Weighting search criteria based on similarities to an ingested corpus in a question and answer (QA) system

A question-answering (QA) system first receives input questions. Each question is then assigned to a first question category of a plurality of question categories. The QA system then identifies a set of candidate answers to each question using a core information source. A set of confidence scores, including a confidence score for each candidate answer, is then calculated. The QA system then determines that the first set of confidence scores fails to satisfy confidence criteria. In response to this determination, an updated information source is ingested.

Handling information source ingestion in a question answering system

An approach is provided for updating corpora in a Question and Answer (QA) system. Ingestion of a first set of sources into a first corpus and a second set of sources into a second corpus with the second set of sources including updates to the first set of sources. A question is received. The system identifies candidate answers to the question using the sources included in the corpuses. Each candidate answer has a weighting. The system determines whether the first and second corpuses have an overlapping source from which two candidate answers were identified. If an overlapping source is found, the candidate answer from the overlapping source in the first corpus is assigned a lower weighting than the candidate answer from the second corpus. Likely answers are selected from the candidate answers based on the weighting and returned to the requestor.

Realtime ingestion via multi-corpus knowledge base with weighting

An approach is provided for utilizing unstructured corpora in a Question and Answer (QA) system. A question is received at the QA system. A private corpora is generated with the private corpora being associated with the user. The private corpora is generated from private data stores associated with the user as well as private data stores associated with other users. Access to the other user's private data sources is provided to the user by the other users. The system retrieves data responsive to the question from the private corpora. The responsive data is ranked based on its relevance to the question. Likely answers are identified based on the ranked responsive data and the likely answers are provided back to the user.

Growing community of inventors

Acton, MA, United States of America

Daniel M Jamrog