Analysis and Research trends using Word Co-occurrence Network

Co-occurrence networks are basically used to provide and give a graphic visualization of potential relationships between people, community, organizations, concepts or other entities represented within the form of written material. The generation and visualization of co-occurrence networks has become practical with the advent of electronically stored and save the text amenable to text mining.

By way of definition, co-occurrence networks are the collective interconnection of terms based on their paired presence within a specified unit of text. Networks are generated by connecting pairs of terms using a set of criteria defining co-occurrence. For example, terms A and B may be said to “co-occur” if they both appear in a particular article. Another article may contain terms B and C. Linking A to B and B to C creates a co-occurrence network of these three terms. Rules to define co-occurrence within a text corpus can be set according to desired criteria. For example, more stringent criteria for co-occurrence may require a pair of terms to appear in the same sentence.

Research & Understanding Trends:

1.      Bibliometric cartography of information:

·        It is important nowadays, for both intellectual and policy reasons, to be able to map the relationship between concepts, ideas and problems in science and social sciences. Bibliometric research is devoted to quantitative studies of literature.

·        Co-word analysis reduces and projects the data into a specific visual representation with the maintenance of essential information containing in the data.

·        Co-word analysis, that counts and analyses the co-occurrences of keywords in the publications on a given subject, on the other hand, has the potential to address precisely this kind of analytic problem (Callon, Courtial & Laville, 1991).

2.      Methodology:

·        Co-word analysis draws upon the assumption that a paper’s keywords constitute an adequate description of its content or, the links the paper established between problems. Two keywords co-occurring within the same paper are an indication of a link between the topics to which they refer (Cambrosio, et. al., 1993).

·        The presence of many co-occurrences around the same word or pair of words points to a locus of strategic alliance within papers that may correspond to a research theme.

·        The main feature of co-word analysis is that it visualizes the intellectual structure of one specific discipline into maps of the conceptual space of this field.

3.      Data collection:

·        Words are the most important research elements in co-word analysis. There are two ways to extract words from journal articles, conference papers, reports or even chapters of books.

·        One of the most significant reservations about this data collection from controlled vocabulary is the possibility of an “indexer effect”.

·        Another method of data collection involves extracting words directly from full-text documents by using some software, such as NPtools (Voutilainen, 1993).

4.      Network:

·        Keywords co-concurrence network is composed of three successive stages of data elicitation, data transformation, and mapping. First, in the data elicitation stage, core keywords are identified from collected literatures. In the data transformation stage, co-word matrix is constructed by measuring the co-occurrence frequency of keywords in the articles.

·        In the keyword network, this represents the importance of a keyword in bridging subsets of keywords. A keywords that lies between two distinctive research themes can have high between’s centrality even though it may have a small number of connections to other keywords in each theme(Freeman, 1979).

5.      Measurement:

·        We examined the characteristics of the keyword networks and identified important keywords from the view point of their frequency of use in the publications and centrality in the keyword network.

·        In order to understand the characteristics of the overall keyword network in ET research, we selectively used between’s centrality measuring. This is the extent to which a node lies on the paths between other nodes.

·        This method enables the researchers to explicitly understand representation of emerging themes.

·        In this  computer software, Net draw, is used to visualize network and then network properties are subsequently calculated and play a important rule.

6.      Result Findings:

·        The basic matrix for the analysis represents the occurrences of words in documents. Documents are considered as the units of analysis.

·        The asymmetrical word–document matrix called a 2-mode matrix. An affiliation network can also be represented as a bipartite graph. In a bipartite graph nodes can be partitioned into two subsets and all lines are between nodes from different subsets.

·        In the bipartite graph for an affiliation network, the lines indicate ties of affiliation.

7.      Keywords co-concurrence network:

·        The links between nodes are symmetric and a link between two keywords is numbered in keyword network. Also this shows how many times the two keywords appear in the network, this number shows the strength of the connection.

·        The size of the node provides a relative indication of the number of times each keyword was mentioned.

·        The width of the links represents the number of times each pair of keywords were mentioned together in paper.

·        Isolates have been removed from the network to aid interpretation.