Discovering relationships in tabular data

• US 9,606,978 B2
• Filed: 11/26/2013
• Issued: 03/28/2017
• Est. Priority Date: 07/01/2013
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Byron, Donna K., Gerard, Scott N., Pikovsky, Alexander, Sanchez, Matthew B.
• Primary Examiner(s): Stork, Kyle
• Assistant Examiner(s): Level, Barbara
• Application Number:
  US14/090,184
• Publication Number:
  US 20150007007A1
• Time in Patent Office: 1,218 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 40/177 of tables; using ruled lines
• CPC Class Codes: G06F 40/18 of spreadsheets form-filling G06F40/174
• CPC Class Codes: G06F 40/20 Natural language analysis semantic analysis of natural language G06F40/30

Abstract

A method for discovering relationships in tabular data is provided in the illustrative embodiments. A set of documents is received, a document in the set including the tabular data. A cell in the tabular data is selected whose dependencies are to be determined. A hypothesis to use in conjunction with the cell is selected. Whether the hypothesis applies to a selected portion of the document is tested by determining whether a conclusion in the hypothesis can be computed using a function specified in the hypothesis on the selected portion. The selected portion can be a selected cell-range in the tabular data or content in a non-tabular portion of the document. The hypothesis is utilized to describe the cell relative to the selected portion.

Claims

1. A method for determining relationships in tabular data, the method comprising:

receiving a set of documents, a document in the set including the tabular data;

applying, to the tabular data, a library of hypotheses specific to a subject-matter domain of the tabular data, each hypothesis in the library representing a hypothetical relationship between hypothetical cells of a hypothetical table,

a first hypothesis in the library of hypotheses applying to hypothetical cells in a column of the hypothetical table,

a second hypothesis applying to hypothetical cells in a row of the hypothetical table,

a third hypothesis repeating in different columns of hypothetical cells of the hypothetical table,

a fourth hypothesis repeating in different rows of hypothetical cells of the hypothetical table,

wherein the hypotheses in the library are configured such that an applicability of a particular hypothesis to actual cells of the tabular data boosts an applicability of another particular hypothesis to the tabular data;

identifying a markup in the document, the markup relating to a cell in the tabular data;

identifying, using the markup, a selected cell-range in the tabular data;

selecting the cell to determine a dependency of the cell on the cell-range;

selecting, based on the markup, a hypothesis from the library of hypotheses to use in conjunction with the cell and the cell-range;

applying the hypothesis to the cell-range;

evaluating, based on a confidence value, that the hypothesis does not fit the cell-range;

changing, responsive to the evaluating, the cell-range to form an adjusted cell-range;

applying the hypothesis to the adjusted cell-range; and

narrating according to the hypothesis, responsive to the hypothesis fitting the adjusted cell-range, using Natural Language Processing, a functional dependency between the cell and the adjusted cell-range.

2. The method of claim 1, wherein the applying the hypothesis to the cell-range is responsive to the confidence value exceeding a threshold.

3. The method of claim 2, further comprising:

adjusting the confidence value responsive to the function specified in the hypothesis producing a second result that is supported by the markup related to a second cell.

4. The method of claim 3, further comprising:

selecting, using the markup, a second portion, the second portion being a second cell-range, wherein the second selected cell-range is in a second tabular data included in the set of documents.

5. The method of claim 1, wherein the portion is a cell-range in the tabular data, further comprising:

testing whether the hypothesis applies to the adjusted cell-range.

6. The method of claim 1, wherein the markup comprises a visual indentation in the selected portion.

7. The method of claim 1, wherein the hypothesis that is applied to the cell-range is selected based on a clue associated with the cell, wherein the clue is a semantic clue, the semantic clue being present in a header associated with the cell.

8. The method of claim 1, wherein the markup is a structural component of the tabular data, the structural component being a boundary marked in the tabular data.