CN106227911B - A Visualization Method of Topic Evolution Based on Circuit Diagram Element Metaphor - Google Patents

Abstract

The invention belongs to the field of data visualization analysis, and in particular relates to a theme evolution visualization method based on circuit diagram element metaphors. Including: data preprocessing, preprocessing text data, performing word segmentation and removing stop words, processing text collections into lexicons; using LDA algorithm to process text collections, extracting topics, and recording words and texts corresponding to topics And the time and place information; the words in the theme are metaphorized by the pad icon, that is, the word plate, and the word plate is represented in the form of a hollow pie chart, indicating the geographical distribution ratio of the word within this time period, etc. This method can mainly display the content of the theme, the change of the intensity of the theme over time, and the evolution relationship between themes to facilitate the user to analyze the evolution process of the theme, and can also display the geographical distribution of the intensity of the same theme in each time period.

Description

A Visualization Method of Topic Evolution Based on Circuit Diagram Element Metaphor

technical field

The invention belongs to the field of data visualization analysis, and in particular relates to a theme evolution visualization method based on circuit diagram element metaphors.

Background technique

Topic evolution refers to the change process of topics over time. It is a research hotspot today and can be widely used in text mining, public opinion analysis, scientific research hotspot analysis and other fields. Its main task is to apply the topic evolution algorithm to discover the trend of topic changes in document collections. The results of topic evolution research are mostly expressed in mathematical forms such as probability, which is difficult to understand, especially the deepening relationship between topics is not easy to be discovered. Therefore, a visual analysis method is needed to display and help analyze the topic evolution process.

Existing visualization methods for theme evolution mainly use Stacked Chart, Alluvial Diagram and other visualization methods such as Themeriver, TestFlow, NEViewer, etc., which are characterized by being able to show the evolution process of the theme over time and the evolution relationship of the theme. However, it can only show the relative intensity of the theme; the stacked graph emphasizes the evolution process of the theme, and uses continuous graphics to express discrete quantities, which is easy to cause misunderstandings; in addition, the above-mentioned methods are not clear enough for the situation where the number of themes is large and the connection is complicated, especially for small-intensity themes. Small-intensity topics are easily covered, and the words that make up the topic are unclear or not represented. In view of this, the present invention proposes a new visualization method of theme evolution based on circuit diagram metaphor to solve the above problems.

Contents of the invention

The purpose of the present invention is to provide a clearer visualization method for theme evolution based on circuit diagram element metaphors.

The purpose of the present invention is achieved like this:

(1) Data preprocessing, preprocessing the text data, performing word segmentation and removing stop words, processing the text collection into a thesaurus; using the LDA algorithm to process the text collection, extracting themes, and recording the words corresponding to the topics, text and time and location information;

(2) The words in the topic are metaphorized by the pad icon, that is, the word plate, which is represented in the form of a hollow pie chart, indicating the geographical distribution ratio of the word within this time period;

(3) Word disks under the same topic extracted within the same period of time are arranged nearby;

(4) The theme is represented by the component icon, that is, the theme frame, which surrounds the word disk in the same theme, indicating that the word disk belongs to the same theme within the selected time period, and the theme intensity is represented by the frame width, that is, the theme is included in this time period The number of documents of , and the box height represents the amount of words contained in the topic;

(5) Use "+" and "-" symbols to indicate the appearance and disappearance of words in the theme, and use the current in the circuit diagram to flow from the positive pole to the negative pole to metaphor the evolution direction of the theme;

(6) The content of the word is displayed in a rounded rectangle on the left side of the logo at the position where each word appears;

(7) Use the circuit traces in the circuit diagram as a metaphor for the evolution of vocabulary between topics in different time periods, that is, use the traces to connect the same vocabulary in different time periods. The word connecting lines are drawn in three sections. The first section is issued from the source word disk. The third paragraph is connected to the target word disk, the middle paragraph is connected to the above-mentioned two paragraphs of word connecting lines, and the word strength is identified on the first and third paragraphs of word connecting lines, that is, the number of occurrences of this word.

The beneficial effects of the present invention are:

This method can mainly display the content of the theme, the change of the intensity of the theme over time, and the evolution relationship between themes to facilitate the user to analyze the evolution process of the theme, and can also display the geographical distribution of the intensity of the same theme in each time period. The invention effectively solves the problems in the existing topic evolution visualization method that the intensity representation is not clear, the relationship between complex topics is not clear enough, the topics with small intensity are easily covered up, and the like.

Description of drawings

Fig. 1 is a step diagram of the present invention;

Fig. 2 is a specific implementation process diagram of the present invention;

Fig. 3 is a geographical distribution diagram of the word disk of the present invention.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings.

A visualization method for theme evolution based on circuit diagram element metaphor, its implementation includes the following steps:

Step 1, data preprocessing, to preprocess the text data, first perform operations such as word segmentation and stop word removal, and process the text collection into a lexicon; then use algorithms such as LDA to process the text collection, extract topics, and record and Information such as words, texts, and time and place corresponding to the topic;

Step 2, use the pad icon as a metaphor for the words in the topic, which can be called "word disk" here, and the word disk is represented in the form of a hollow pie chart, as shown by the number 1 in Figure 2, which can represent the time period during this period. The geographical distribution ratio of words is shown in Figure 3.

Step 3, the word trays under the same topic extracted in the same period of time are arranged nearby, as shown in Figure 3;

Step 4, use the component icon (rectangular frame) to represent the theme, which can be called "theme frame" here, and the theme frame surrounds the dictionaries in the same theme, indicating that they belong to the same theme within the selected time period, and is represented by the width of the box Topic strength (the number of documents containing this topic in this period), as shown by the number 2 in Figure 2;

Step 5, use "+" and "-" symbols to indicate the appearance and disappearance of words in the theme, and use the current in the circuit diagram to flow from the positive pole to the negative pole to metaphorize the evolution direction of the theme, as shown in the number 3 in Figure 2 and the number 4 in Figure 2;

Step 6, the content of the word is displayed in a rounded rectangle on the left side of the logo at the position where each word appears, as shown in the label 5 in Figure 2;

Step 7, use the circuit diagram to metaphor the evolution relationship of vocabulary between topics in different time periods, that is, use the wiring to connect the same vocabulary in different time periods, the word connecting lines are drawn in three sections, the first section is sent from the source word disk, The third section is connected to the target word disk, and the middle section is connected to the above-mentioned 2 paragraphs of word connection lines, as shown in the label 6 in Figure 2, and the first paragraph and the third paragraph word connection line identify word strength (promptly appear this word quantity), as shown by number 7 in Figure 2.

Including the words included in the theme of the pad icon metaphor described in step 2, the word disk is represented in the form of a hollow pie chart, which can indicate the geographical distribution ratio of the word within this time period.

Including the theme represented by the component icon (rectangular box) described in step 4, which can be called "theme box" here, and the theme box surrounds the dictionaries in the same theme, indicating that they belong to the same theme within the selected time period, so as to The width of the box indicates the strength of the topic (the number of documents containing the topic in this period), and the height of the box indicates the amount of words contained in the topic.

Including the use of "+" and "-" symbols in step 5 to indicate the appearance and disappearance of words in the theme, and the evolution direction of the theme as a metaphor for the current flowing from the positive pole to the negative pole in the circuit diagram.

Including the evolution relationship of vocabulary between topics in different time periods as described in step 7, using the circuit diagram to connect lines to metaphor, that is, use lines to connect the same vocabulary in different time periods.

The word connection line is drawn in three sections, the first section is sent from the source word disk, the third section is connected to the target word disk, the middle section connects the above two word connection lines, and the first and third paragraph word connection lines Identifies word strength (that is, the number of occurrences of the word).

Claims (1)

1. a kind of subject evolution method for visualizing based on circuit diagram element metaphor, which comprises the steps of:

(1) data prediction pre-processes text data, is segmented, goes to stop word operation, text collection is processed into word Library；Text collection is handled using LDA algorithm, extracts theme, and record corresponding with theme word, text and time and Location information；

(2) with the word in pad icon metaphor theme, i.e. word disk, word disk is indicated using hollow cake diagram form, is indicated in this time Geographical distribution ratio locating for this word in section；

(3) the word disk under the same subject extracted within the same period is arranged nearby；

(4) theme, i.e. subject box are indicated with the symbol's icon, subject box surrounds the word disk in same subject, indicates word disk selected Belong to same subject in period, theme intensity is indicated with width of frame, i.e., includes the number of files of the theme in this period, with frame height Degree indicates the word amount that theme includes；

(5) appearance and extinction of the word in theme are indicated with "+" "-" symbol, is flowed with electric current in circuit diagram from anode to cathode The evolution direction of metaphor theme；

(6) it is identified in the position that each word occurs, left side shows the content of word with round rectangle；

(7) it with the Evolvement of vocabulary between circuit trace metaphor different time sections theme in circuit diagram, i.e., is connected with cabling different Identical vocabulary in period, three sections of draftings of word connecting line point, first segment are issued by source word disk, and third section is connected to purpose word disk, Interlude connects above-mentioned two sections of word connecting lines, and word intensity is identified on first segment and third section word connecting line, that is, this word occurs Quantity.