Automated system and method for generating reasons that a court case is cited

An automated method of designating text, taken from a set of citing documents, as reasons for citing (RFC) a cited document, said RFC being associated with respective citing instances of a citing document, the method comprising:

• inputting text from the citing documents;
• dividing the citing documents' text to define paragraphs, and dividing the paragraphs to define sentences;
• obtaining contexts of the citing instances in the respective citing documents, each context including:
• a sentence that includes the citing instance and at least one sentence that is near the citing instance;
• generating a content word list contain ing content words that are in at least two of the citing documents' contexts or that are in at least one citing document's context and said cited document;
• calculating, for the sentences in the citing documents' contexts, respective content scores that are based on frequency counts of the content words that are recited in the respective sentences, said step of calculating content scores including the steps of,
• calculating respective initial content scores (ICS) for the sentences in the citing documents, based on the content words in the sentences;
• calculating respective distances of the sentences in the citing documents from respective citing instances of the cited document;
• and
• calculating respective content scores (CS) for the sentences in the citing documents, based on at least the ICS and the distances;
• and
• selecting, from the citing documents' contexts, at least one sentence that constitutes the RFC, based on the calculated content scores.

An apparatus for designating text, taken from a set of citing documents, as reasons for citing (RFC) a cited document that are associated with respective citing instances of a citing document, the apparatus comprising:

• means for obtaining contexts of the citing instances in the respective citing documents, each context including a text unit that includes the citing instance and a text unit that is near the citing instance;
• means for analyzing the content of the contexts, said means for analyzing including means for calculating a content score for each text unit based on text unit content words that are common to at least two of the citing documents' contexts or to at least one citing document's context and said cited document, said means for calculating a content score including,
• means for calculating respective initial content scores (ICS) for the text units in the citing, documents, based on the content words in the text units;
• means for calculating respective distances of the text units in the citing documents from respective citing instances of the cited document;
• and
• means for calculating respective content scores (CS) for the text units in the citing documents, based on at least the ICS and the distances;
• and
• means for selecting, from the citing instances' context, at least one text unit that constitutes the RFC, based on the analyzed content of the contexts.

An apparatus for designating text, taken from a set of citing documents, as reasons for citing (RFC) a cited document, said RFC being associated with respective citing instances of a citing document, the apparatus comprising:

• means for dividing the citing documents' text to define paragraphs, and for dividing the paragraphs to define sentences;
• means for obtaining contexts of the citing instances in the respective citing documents, each context including:
• a sentence that includes the citing instance and at least one sentence that is near the citing instance;
• means for generating a content word list containing content words that are in at least two of the citing documents' contexts or that are in at least one citing document's context and said cited document;
• means for calculating, for the sentences in the citing documents' contexts, respective content scores that are based on frequency counts of the content words that are recited in the respective sentences, said means for calculating content scores including,
• means for calculating respective initial content scores (ICS) for the sentences in the citing documents, based on the content words in the sentences;
• means for calculating respective distances of the sentences in th citing documents from respective citing instances of the cited document;
• and
• means for calculating respective content scores (CS) for the sentences in the citing documents, based on at least the ICS and the distances;
• and
• means for selecting, from the citing documents' contexts, at least one sentence that constitutes the RFC, based on the calculated content scores.

A computer-readable memory that, when used in conjunction with a computer, can carry out a method of designating text, taken from a set of citing documents, as reasons for citing (RFC) a cited document that are associated with respective citing instances of a citing document, the computer-readable memory comprising:

• computer-readable code for obtaining contexts of the citing instances in the respective citing documents, each context including a text unit that includes the citing instance and a text unit that is near the citing instance;
• computer-readable code for analyzing the content of the contexts including computer-readable code for calculating a content score for each text unit based on text unit content words that are common to at least two of the citing documents' contexts or to at least one citing document's context and said cited document, said computer-readable code for calculating a content score including,
• computer-readable code for calculating respective initial content scores (ICS) for the text units in the citing documents, based on the content words in the text units;
• computer-readable code for calculating respective distances of the text units in the citing documents from respective citing instances of the cited document;
• and
• computer-readable code for calculating respective content scores (CS) for the text units in the citing documents, based on at least the ICS and the distances;
• and
• computer-readable code for selecting, from the citing instances' context, at least one text unit that constitutes the RFC, based on the analyzed content of the contexts.

A computer-readable memory that, when used in conjunction with a computer, can carry out a method of designating text, taken from a set of citing documents, as reasons for citing (RFC) a cited document, said RFC being associated with respective citing instances of a citing document, the apparatus comprising:

• computer-readable code for inputting text from the citing documents;
• computer-readable code for dividing the citing documents' text to define paragraphs, and dividing the paragraphs to define sentences;
• computer-readable code for obtaining contexts of the citing instances in the respective citing documents, each context including:
• a sentence that includes the citing instance and at least one sentence that is near the citing instance;
• computer-readable code for generating a content word list containing content words that are in at least two of the citing documents' contexts or that are in at least one citing document's context and said cited document;
• computer-readable code for calculating, for the sentences in the citing documents' contexts, respective content scores that are based on frequency counts of the content words that are recited in the respective sentences said computer-readable code for calculating content scores including,
• computer-readable code for calculating respective initial content scores (ICS) for the sentences in the citing documents, based on the content words in the sentences;
• computer-readable code for calculating respective distances of the sentences in the citing documents from respective citing instances of the cited document;
• and
• computer-readable code for calculating respective content scores (CS) for the sentences in the citing documents, based on at least the ICS and the distances;
• and
• computer-readable code for selecting, from the citing documents' contexts, at least one sentence that constitutes the RFC, based on the calculated content scores.