Method for automatically performing conceptual highlighting in electronic text
View Patent ↗User's interests can be explicitly determined via keywords that the user specifies, and/or are implicitly constructed from user browsing and reading activity. User interests may be expressed as an interest profile. Conceptual keywords related to user interests are selected by combining spreading activation and word co-occurrence, by latent semantic analysis, or other methods. One embodiment automatically highlights sentences and other information that contain conceptual keywords related to user interests. Highlights can be activated when the user directly performs a keyword search or index search, or the invention can generate information reflecting user interests, apply it to the text, and generate and display highlights. One embodiment includes an algorithm for computing a conceptual keyword vector through an iterative spreading activation process also employing word co-occurrence. A conceptual index of the text may be created and then combined with conceptual highlighting.
1. A computer-implemented method for automatically performing conceptual highlighting in electronic text, comprising:
selecting a set of conceptual keywords related to a user's interests in an electronic text on a computer with at least one processor,
wherein said conceptual keywords are words with the highest levels of semantic similarity with said interests, wherein said interests comprise a user-generated keyword vector K comprising user-generated keywords, wherein said user-generated keywords are generated from a user's activity, wherein a new keyword vector K new is created by adding new keywords based on said activity to an existing keyword vector K;
creating a list of sentences and other information containing one or more of said conceptual keywords, wherein said set of conceptual keywords are described by a conceptual keyword vector K′, wherein said conceptual keywords are conceptually related to said user-generated keywords; and
automatically highlighting sentences and other information in the electronic text containing said conceptual keywords; and
wherein said keyword vector K and said conceptual keyword vector K′ are described by the following iterative equations after t iterations:
K (1)= K;
K (2)= a*K (1)+ b*WC*K (1);
K ( t )= K′=a*K ( t −1)+ b*WC*K ( t −1),
wherein WC is a word co-occurrence matrix; wherein a and b are constants; and wherein a+b=1.
2. The computer-implemented method of claim 1 , wherein said creating comprises:
for each sentence S in said text, creating a conceptual sentence vector S′; and
comparing S′ and K′.
3. A computer-implemented method, comprising:
selecting conceptual keywords in an electronic text on a computer with at least one processor, wherein said conceptual keywords are conceptually related to user-generated keywords comprised in keyword vector K, and wherein said keyword vector K is described by the following iterative equations after t iterations:
K (1)= K;
K (2)= a*K (1)+ b*WC*K (1);
K ( t )= K′=a*K ( t −1)+ b*WC*K ( t −1),
wherein WC is a word co-occurrence matrix; wherein a and b are constants; and wherein a+b=1; and
highlighting sentences and other information in the electronic text containing said conceptual keywords.
4. The computer-implemented method of claim 3 , wherein a normalized conceptual keyword vector K norm ′ is described by the following equation after t iterations:
K norm ′=K ′/( c*TF ),
wherein term-frequency vector TF specifies the term frequency in the text's language of each keyword comprised in K; and wherein c is a constant.
5. The computer-implemented method of claim 1 , wherein said selecting employs spreading activation and word co-occurrence.
6. The computer-implemented method of claim 1 , wherein said selecting employs latent semantic analysis.
7. The computer-implemented method of claim 1 , wherein different colors are used for highlighting different kinds of information relating to user interests.
8. The computer-implemented method of claim 1 , wherein said highlighting is performed only on sentences and other information complying with a user-specified highlighting preference.
9. The computer-implemented method of claim 8 , wherein said highlighting preference is one or more of:
a threshold formula for determining a threshold relevancy level R;
a number formula for determining the number N of sentences and other information to be highlighted on a page; and
a percentage formula for determining the percentage P of sentences and other information to be highlighted on a page.
10. The computer-implemented method of claim 1 , further comprising:
decreasing term weights of the existing keywords as new keywords are added.
11. The computer-implemented method of claim 1 , wherein said conceptual keywords are comprised in a conceptual index of said electronic text.
12. The computer-implemented method of claim 11 , wherein after a user selects an index entry, a tool tip is generated displaying passages conceptually relevant to said index entry and comprised in said electronic text.