Abstract: An approach is provided in which an information handling system identifies candidate answers from a knowledge base that correspond to a question comprising a lexical answer type (LAT). The information handling system determines sub-LATs corresponding to the LAT in a database that are child LATs of the LAT. The information handling system then identifies a set of the candidate answers that correspond to at least one of the sub-LATs and ranks the set of candidate answers based on one or more ranking criterion. In turn, the information handling system provides the ranked set of candidate answers to a user.

Abstract: A method includes receiving, by a computer system configured to generate search result, a query, identifying a plurality of entities in the query, determining, using a knowledge graph, a measure of a relationship between the entities in the query, identifying a replacement entity for a first entity of the entities in the query, wherein the replacement entity is determined to improve the measure of the relationship between the entities when the first entity is replaced by the replacement entity, and generating a search result using the query modified by the replacement entity, which replaces the first entity in the query.

Abstract: Embodiments for processing questions based on equivalence classes in a cognitive question answering system. A plurality of syntactic representations of a plurality of questions asked of the cognitive question answering system are provided. A plurality of syntactic representations of a plurality of passages ingested by the cognitive question answering system are provided. Question focus to candidate passage pairs are mapped to form an equivalence class mapping, and the equivalence class mapping is used to determine an answer to one of the plurality of questions asked of the cognitive question answering system.

Abstract: A system for handling a graphical representation of data associated with a question answering (QA) input document includes a memory having instructions therein and includes at least one processor in communication with the memory. The at least one processor is configured to execute the instructions to derive, at least from a portion of the QA input document, first metadata regarding a context of the graphical representation of data. The at least one processor is also configured to execute the instructions to derive, at least from a portion of the graphical representation of data, tabular data. The at least one processor is also configured to execute the instructions to determine, at least in part by comparing at least a portion of the first metadata to existing table annotations from a QA knowledge base, how to incorporate the tabular data into the QA knowledge base.

Abstract: A controller receives a multi-facet natural language question from a user at a question answering system and generates one or more inquiry questions to submit to the question answering system for an order type requirement of the multi-facet natural language question. The controller evaluates responses, each with a respective confidence score by the question answering system, to each of the one or more inquiry questions to identify a selection of answers from among the responses, the selection of answers capped at a number of specific answers assessed for the multi-facet natural language question. The controller merges the selection of answers into a single aggregated answer to return to the user in response to the multi-facet natural language question.

Abstract: A computer converts a question received in a natural language format into a string of text elements. The computer searches a corpus comprising unstructured passages with the string of the text elements as search terms to identify a selection of unstructured passages from the corpus relevant to the text elements. The computer annotates the selection of relevant unstructured passages with one or more annotations according to at least one natural language annotation type to generate an annotated selection knowledge base. The computer modifies the string of text elements by annotating at least one of the text elements according to the at least one natural language annotation type. The computer searches the annotated selection knowledge base using the modified string of text elements to generate a selection of ranked passages. The computer identifies an answer to the question based on the selection of ranked passages.

Abstract: A method, apparatus and computer program product for creating parse trees from mixed records is described. A mixed record is received including a first entry containing a first natural language phrase and a second entry containing a first structured data entry. Using the first natural language phrase, a first parse tree structure is created joining words in the first natural language phrase with natural links indicating their semantic relationship within the first natural language phrase. A first synthetic node is created which represents the first structured data entry. The first synthetic is joined to the first parse tree structure using a synthetic link to a node in the parse tree to produce a pseudo parse tree. The pseudo parse tree is sent to a question answer system for answering user queries by reference to the pseudo parse tree.

Abstract: A method includes: receiving a question including multiple foci; determining one or more passages (OMPs) that include one or more sets of candidate answers (OMSCAs) to the question; determining matchings between words in the question and words in the OMPs; determining skipngrams of the question (“first skipngrams”); determining one or more sets of skipngrams of the OMPs (“second skipngrams”); determining one or more matching numbers (OMMNs) for the second skipngrams by determining, for each set of the second skipngrams and using the matchings, a number of the first skipngrams that match skipngrams of the set of skipngrams; determining a feature input to a machine learning model (MLM) using the OMMNs; determining one or more answers (OMAs) to the question by determining, by the MLM, whether the OMSCAs are correct answers to the question based at least in part on the feature input; and sending the OMAs to a user device.

Abstract: An approach is provided that receives a question at a question-answering (QA) system. The received question includes one or more terms, and the question pertains to a subject matter domain that is supported by the QA system. Analyzing a number of expressions included in a set of question-answer pairs (QA pairs), with the QA pairs being ground-truths established to in support of the subject matter domain. The analysis identifies whether a selected term from the question is a synonym for any of the expressions. The expressions that are identified as synonyms are then used in a QA pipeline that generates one or more candidate answers to the received question.

Abstract: A method, system and computer program product are provided for implementing misleading title identification. A title and a list of passages are received. The title is supplied to a question answering factoid pipeline as a question. The title is analyzed against the list of passages or content as an answer source material. A respective answer score of all passages is compared with a tunable threshold to identify titles that are misleading or inaccurate based on the content of all passages. A selected action for the identified misleading titles is provided for a user.

Abstract: The disclosed embodiments include a Question and Answer (QA) system, computer program product, and computer-implemented method configured to adjust passage relevancy to a question based on time. The disclosed embodiments seek to improve the process of providing answers to questions that include a temporal reference. In particular, the disclosed embodiments are better able to provide answers to questions that include a temporal reference that does not have an exact matching temporal reference in an answer source of a QA system.

Abstract: An approach is provided in which an information handling system converts a first set of text to synthesized speech using a text-to-speech converter. The information handling system then converts the synthesized speech to a second set of text using a speech-to-text converter. In response to converting the synthesized speech to the second set of text, the information handling system analyzes the second set of text against a filtering criterion and prevents usage of the synthesized speech based on the analysis.

Abstract: A method trains an artificial intelligence system to respond to a complex query. A query/answer (Q/A) system sends a first query to an information server, where the first query is a query about a first topic with regard to a first entity, and where the information server provides access to unstructured information sources. In response to receiving a negative response to the first query for the first entity, the Q/A system sends variations of the first query and a second query about a second topic to the information server. The Q/A system identifies a pattern in the table of information in order to create the answer to the first query for the first entity. The Q/A system then trains an artificial intelligence system to respond to a complex query about the first entity with information in the table of information.

Abstract: A computer-implemented method that includes receiving a search query and, after receiving the search query, identifying a first subset of documents that include the search query by searching, based on the search query, an inverted index of a corpus of documents that includes the first subset of documents. The computer-implemented method additionally includes processing the first subset of documents to identify entities related to the search query and relations between the entities and the search query. The computer-implemented method additionally includes generating a knowledge graph based on the entities and the relations.

Abstract: A computer-implemented method includes receiving an input query including a comparative question referencing a plurality of entities. The comparative question includes a comparative term regarding a metric associated with the plurality of entities. The computer-implemented method includes generating a plurality of polar questions. Each of the plurality of polar questions seeks a polar answer to a comparison, along the metric and according to the comparative term, between a target entity of the plurality of entities and at least one second entity of the plurality of entities. The computer-implemented method includes determining confidence scores for the plurality of polar questions. The computer-implemented method includes identifying a particular polar question of the plurality of polar questions based on the confidence scores. The computer-implemented method includes outputting a response to the input query that identifies the target entity of the particular polar question as an answer to the input query.

Abstract: A system comprising a memory storing an inverted index of a corpus of documents, and a processor coupled to the memory and configured to receive a search query, after receiving the search query identify a first subset of documents that include the search query by searching, based on the search query, the inverted index of the corpus of documents, wherein the corpus of documents includes the first subset of documents, process the first subset of documents to identify entities related to the search query and relations between the entities and the search query, and generate a knowledge graph based on the entities and the relations.

Abstract: Current plagiarism checkers can determine if a document has extensively copied from another but these checkers fail when a document is filled with synonyms. The present invention offers a means of detecting plagiarism involving the use of synonyms or heaving paraphrasing. A source passage from a pertinent corpus is modified by generating augmented tokens which include synonyms for the various terms in the passage. Text analysis may be required to confirm the semantic meaning of a term. The modified passage may be a regular expression. The modified passage can then be compared to a suspect passage using a trigram scorer. If possible plagiarism is detected, an alert message is sent to a reviewer with the passage of interest, the source passage, and an identification of the related corpus.

Abstract: An approach is provided that receives words that are input by a user of an application with the words being displayed on a display device. Each of the words are compared to words from a dictionary. Based on the comparisons, words that are not found in the dictionary and only appear a single time are highlighted as being misspelled words. However, words that are not in the dictionary but appear multiple times in the document are highlighted differently to indicate that these words are possible misspelled words with the difference in highlighting allowing the user to easily discern between misspelled and possibly misspelled words.

Abstract: An approach is provided that receives words that are input by a user of an application with the words being displayed on a display device. Each of the words are compared to words from a dictionary. Based on the comparisons, words that are not found in the dictionary and only appear a single time are highlighted as being misspelled words. However, words that are not in the dictionary but appear multiple times in the document are highlighted differently to indicate that these words are possible misspelled words with the difference in highlighting allowing the user to easily discern between misspelled and possibly misspelled words.

Abstract: A method, system and computer program product are provided for implementing unstructured content utilization from a structured source in a system capable of answering questions. A question is received by the system. The question is provided as a list of tokens T (T1, T2, . . . , Tn). A structured source is received containing records R (R1, R2, . . . , Rk). Headers H (H1, H2, . . . , Hi) are mapped to records. Each record R1, R2, . . . , Rk is searched for any token Ti in the list of tokens T (T1, T2, . . . , Tn) in the question. Responsive to identifying the token Ti in the question in a record Rj, the token Ti is added to a tracking structure. Any header mapped to the record Rj is compared to the list of tokens T (T1, T2, . . . , Tn) in the question. Responsive to identifying a header Hj mapped to the record Rj in the question, the header Hj is added to tracking structure.