Method and apparatus for browsing information

    公开(公告)号:US11893337B2

    公开(公告)日:2024-02-06

    申请号:US17333953

    申请日:2021-05-28

    摘要: Disclosed is a method of generating a multi-level summary of an article. The method may comprise generating, by a computing device, a low-level summary from article-matter in an article. The method may also comprise generating, by the computing device, a mid-level summary based on the low-level summary and the article-matter. The method may also comprise generating, by the computing device, an upper-level summary based on the mid-level summary, the low-level summary, and the article-matter.

    Ambiguous date resolution for electronic communication documents

    公开(公告)号:US11580291B2

    公开(公告)日:2023-02-14

    申请号:US16922141

    申请日:2020-07-07

    摘要: A computer-implemented method for resolving date ambiguities in electronic communication documents includes identifying, within the documents, date field values each associated with a different instance of a communication segment. The method also includes resolving a candidate date for each different communication segment instance, with each candidate date being associated with a respective priority level indicative of a level of certainty with which the candidate date was resolved, and determining a final date from among the candidate dates at least by comparing the respective priority levels. The method further includes determining, based on the final date, an ordered relationship between the electronic communication documents, and storing metadata indicating the ordered relationship between the electronic communication documents.

    MULTI-HOP EVIDENCE PURSUIT
    7.
    发明申请

    公开(公告)号:US20230035641A1

    公开(公告)日:2023-02-02

    申请号:US17840987

    申请日:2022-06-15

    发明人: Christopher Malon

    摘要: A method for neural network training is provided. The method inputs a training set of textual claims, lists of evidence including gold evidence chains, and claim labels labelling the evidence with respect to the textual claims. The claim labels include refutes, supports, and not enough information (NEI). The method computes an initial set of document retrievals for each of the textual claims. The method also includes computing an initial set of page element retrievals including sentence retrievals from the initial set of document retrievals for each of the textual claims. The method creates, from the training set of textual claims, a Leave Out Training Set which includes input texts and target texts relating to the labels. The method trains a sequence-to-sequence neural network to generate new target texts from new input texts using the Leave Out Training Set.

    ELECTRONIC HEADER RECOMMENDATION AND APPROVAL

    公开(公告)号:US20230023325A1

    公开(公告)日:2023-01-26

    申请号:US17374075

    申请日:2021-07-13

    摘要: Recommendation and approval of a header for a message includes generating a proposed header based on the name and/or brand of the entity and product and/or content of the message, classifying the proposed header using a machine learning model trained based on historical complaints on previously used headers related to the entity name and brand and product and/or content of the message and recommending the proposed header based on the classification. The training of the machine learning model may include learning a threshold wherein headers having a classification greater than the threshold are not recommended as having a high probability of being wrongly associated with the requesting entity and headers having a classification lower than the threshold are recommended as having a high probability of not being wrongly associated with the requesting entity.

    Identifying section headings in a document

    公开(公告)号:US11494555B2

    公开(公告)日:2022-11-08

    申请号:US16675456

    申请日:2019-11-06

    发明人: Darrell Bellert

    摘要: A method, non-transitory computer readable medium, and system for inferring certain texts as stylized section headings in an electronic document (ED). Stylized section headings are section headings that have unique styling distinct from the body of text below each stylized heading. In particular, the stylized section headings are identified based on styling information in the ED. Identifying stylized section headings includes grouping candidate headings based on identification of dominant styling, locating high level fragments, and repeatedly locating nested fragments from within higher level fragments. The ED may or may not include explicitly identified headings in the document.