Techniques for structuring knowledge bases specific to a user or group of users and techniques for using the knowledge bases to answer user inputs are described. A knowledge base may be populated with information provided by users associated with the knowledge base. Users associated with a knowledge base may be proactive in providing content to the knowledge base and/or a system may solicit an answer to a user input from users associated with a particular knowledge base. When the system receives an answer, the system may populate the knowledge base with the answer and may output the answer to the user that originated the user input. The system may output user inputs to be answered using messages or by establishing two-way communication sessions.

A computer based system and method for assigning queries to topics and/or visualizing or analyzing query coverage may include, using a computer processor, searching, using a set of queries, over a set of text documents, to produce for each query a set of search results for the query. Each search result may include a subset of text from a text document of the set of text documents. For each query, a query vector may be calculated based on the set of search results for the query, and for each of a set of topics describing the set of text documents, a topic vector may be calculated. A report or visualization may be generated including the set of queries and the set of topics using the topic vectors and the query vectors.

A mechanism is provided in a data processing system to implement a cognitive natural language processing (NLP) system with descriptor uniqueness identification to support named entity mention clustering. The mechanism annotates a set of documents from a corpus of documents for entity types and mentions, collects descriptor usages from all documents in the corpus of documents, analyzes the descriptor usages to classify the descriptors as base terms or modifier terms, generates compatibility scores for the descriptors, and performs entity merging of entity clusters based on the compatibility scores.

The automated collection of online data is enhanced by generating and saving a context between a document and a related named entity, as well as a credibility level of the online source. The context, credibility level, and quality and quantity of collected data are used to enhance the use of the collected data in automated decision-making. Both the quality and the quantity may be continuously updated and honed through machine learning. Three new algorithms—DUPES, CORRAL, and ONTO—have been introduced to support the above, improving current state-of-the-art engineering practice by sharpening the strategy for named-entity searching, for ensuring that topic modeling produces relevant topic tags, and for handling sentiment which may be NEGATIVE, POSITIVE, and NEUTRAL (which includes MISSING and INCONCLUSIVE).

Systems and methods including one or more processors and one or more non-transitory storage devices storing computing instructions configured to run on the one or more processors and perform: for each respective record in a plurality of records, assigning the respective record: (1) to a first database cluster on a first database server; or (2) to a second database cluster on a second database server; receiving a search request from a requester; executing the search request in the first database cluster to retrieve a first set of results; when the first set of results is greater than a threshold number of results, presenting the first set of results to the requester; and when the first set of results is less than the minimum number of results: executing the search request in the second database cluster to create a second set of results; and presenting the second set of results to the requester, wherein: the threshold number of results is configured such that a probability that the first set of results is greater than the minimum number of results is at least fifty percent. Other embodiments are disclosed herein.

Presented are relation inference methods and systems that use deep learning techniques for data mining documents to discover a relation between terms of interest in a given field covering a specific topic. For example, in the healthcare domain, various embodiments of the present disclosure provide for a relation inference system that mines large-scale medical documents in a free-text database to extract symptom and disease terms and generates relation information that aids in disease diagnosis. In embodiments, this is accomplished by training and using an RNN, such as an LSTM, a Gated Recurrent Unit (GRU), etc., that takes advantage of a term dictionary to examine co-occurrences of terms of interest within documents to discover correlations between the terms. The correlation may then be used to predict statistically most probable terms (e.g., a disease) related to a given search term (e.g., a symptom).

A non-transitory computer-readable recording medium stores a response processing program for causing a computer to execute processing including: receiving a question from a user input to a terminal; extracting, in a case where a plurality of pieces of answer candidate data that corresponds to the received question is retrieved, keywords or key phrases from the plurality of pieces of answer candidate data; classifying the extracted keywords or key phrases on the basis of words included in the keywords or key phrases; and outputting a classification result of the keywords or key phrases to the terminal in a state selectable by a user, together with a response text to the question.

Systems, apparatuses, and methods are provided for identifying a corresponding string stored in memory based on an incomplete input string. A system can analyze and produce phonetic and distance metrics for a plurality of strings stored in memory by comparing the plurality of strings to an incomplete input string. These similarity metrics can be used as the input to a machine learning model, which can quickly and accurately provide a classification. This classification can be used to identify a string stored in memory that corresponds to the incomplete input string.

Systems and methods are directed to providing efficient and fast invocation of concurrent sheet defined functions (SDFs) including dynamic arrays by front-loading the work. At SDF creation time, a SDF cell table, a formula table, and a spill area table are generated. The SDF cell table represents cells from a worksheet that are used for the SDF. The formula table comprises an index of formulas used by the SDF, whereby index identifiers are stored in cells of the SDF cell table. The spill area table comprises an index of spill areas where each dynamic array may automatically spill into. The SDF cell table, formula table, and spill area table are shared between a plurality of invocations of the SDF during invocation time.

Systems and methods are directed to providing efficient and fast invocation of concurrent sheet defined functions (SDFs) including dynamic arrays by front-loading the work. At SDF creation time, a SDF cell table, a formula table, and a spill area table are generated. The SDF cell table represents cells from a worksheet that are used for the SDF. The formula table comprises an index of formulas used by the SDF, whereby index identifiers are stored in cells of the SDF cell table. The spill area table comprises an index of spill areas where each dynamic array may automatically spill into. The SDF cell table, formula table, and spill area table are shared between a plurality of invocations of the SDF during invocation time.