English. In this paper, we present a preliminary study on the style of headlines in order to evaluate the correlation between linguistic features and newsworthiness. Our hypothesis is that each particular linguistic form or stylistic variation can be motivated by the purpose of encoding a certain newsworthiness value. To discover the correlations between newsworthiness and linguistic features, we perform an analysis on the basis of characteristics considered indicative of a shared communicative function and of discriminating factors for headlines.
Italiano. Questo contributo descrive uno studio preliminare sullo stile dei titoli nelle notizie, al fine di valutare la correlazione tra gli aspetti linguistici e il valore delle notizie. La nostra ipotesi e` che ogni particolare forma linguistica o variazione stilistica possa essere motivata dall’obiettivo di codificare un certo valore di notiziabilit a`. Al fine di analizzare la correlazione tra il valore delle notizie e gli aspetti linguistici, effettuiamo un‘analisi sulla base delle caratteristiche considerate indicative di una funzione comunicativa condivisa e di fattori discriminanti per i titoli. newsworthy increases with the number of factors it complies with.
The newsworthiness factors reflect a set of values
and provide a certain representation of the world
This paper aims to provide a preliminary analysis of the linguistic features in news headlines and how these relate to specific newsworthiness categories. The analysis rests on the hypothesis that each particular linguistic form or stylistic variation can be motivated by the purpose of encoding a certain newsworthy value. The remainder of the paper is structured as follows. In Section 2, we describe the related work on stylistic analysis of news and headlines. In Section 3, we describe the data set and the classification scheme we use. In Section 4 we introduce our methodology together with the analysis we perform, while in Section 5 we discuss the results. Section 6 concludes the paper. 2
Several works, based on sociolinguistic and
discourse analysis frameworks, have investigated
stylistic features and linguistic variations in both
newspapers and headlines, on the basis of
different parameters and aspects
A considerable amount of research has analyzed
the language of news media from a broader
prospective
Another strand of research focuses on specific linguistic aspects in journalistic style. For instance, Tannenbaum and Brewer (1965) analyze the syntactic structure across different news content areas, while Schneider (2000) analyzes the textual structures in British headlines, revising the traditional distinction among verbal and nominal headlines. 3
In our work, we adopt the data set proposed for
SemEval-2007 task 14
Table 1 lists the news value labels, their counts in the data set, and the inter-annotator agreement Our methodology to define the stylistic variations related to newsworthiness categories relies on a descriptive analysis of different features, i.e., syntactic, lexical and compositional features.
We extracted these using Coh-Metrix,1 a
computational tool that provides a wide range of language
and discourse metrics
We ran Coh-Metrix analysis on headlines from our dataset, grouped according to the 11 newsworthiness labels. We then analyzed these results manually and decided to adopt a subset of Coh-Metrix indices, which, according to our initial hypothesis, we consider to be discriminating factors for newsworthiness, i.e., text easibility principal component and word information indices. Being representative of linguistic characteristics and syntax context, such features are suitable to represent stylistic variations and, therefore, the underlied news value. 2.274 2.239 1.834 2.502 1.923 1.27 1.741 2.585 2.057 1.451 2.671 In our preliminary analysis, we consider two main types of linguistic features: text easability and word information scores. 5.1
Coh-Metrix text easibility indices (“Text easability principal component scores”) are designed to measure text ease that goes beyond traditional readability metrics. We focused specifically on two indices related to the syntactic simplicity (PCSYNz) and word concreteness (PCCNCz) (Table 2).
The syntactic simplicity is evaluated on the basis of the number of words and the complexity of syntactic structures of sentences. As far as the syntactic simplicity is concerned, the variability among the categories is not so high, nevertheless, we can distinguish two groups. The first group, with a higher PCSYNz, consists of headlines labeled with the ‘Power elite’, ‘Bad news’, ‘Shareability’, ‘Drama’, ‘Magnitude’, and ‘Celebrity’ news values. Higher scores here indicate that the sentence presents more words and uses complex syntatic structures, as exemplifed by the following headlines from this group: (1a) China says rich countries should take lead on global warming (Power elite) (1b) Iraqi suicide attack kills two US troops as militants fight purge (Bad news) (1c) Second opinion: girl or boy? as fertility technology advances, so does an ethical debate (Shareability) (1d) Damaged Japanese whaling ship may resume hunting off Antarctica (Drama) (1e) Ready to eat chicken breasts recalled due to suspected listeria (Magnitude) (1f) Jackass’ star marries childhood friend The secrets people reveal (Celebrity)
The second group consists of headlines labled with ‘Entertainment’, ‘Surprise’, ‘Follow up’, ‘Good news’, and ‘Conflict’, which received lower PCSYNz scores, and are thus of less syntactic complexity. Examples of headlines form this group are as follows: (2a) Action games improve eyesight (Entertainment) (2b) Breast cancer drug promises hope (Good news) (2c) Merkel: Stop Iran (Conflict)
The second index, word concreteness, differentiates three groups of headlines: (i) ‘Power elite’, ‘Entertainment’, ‘Shareability’, ‘Celebrity’, and ‘Conflict’, all with a low z-score; (ii) ‘Follow up’, ‘Drama’ and ‘Magnitude’, with a medium z-score; and (iii) ‘Bad news’, ‘Surprise’ and ‘Good news’ with a high z-score. The following headlines exemplify each of the three groups: (1a) Action intensity boosts vision (Shareability) (2a) Ex-suspect slams anti-terror laws (Drama) (3a) Ancient coin shows Cleopatra was no beauty (Surprise)
The word concreteness index measures the concreteness level of content words. Thus, news values with lower scores are characterized by a higer number of abstract words and, for this reason, may be less easy to understand without an appropriate context. Our analysis thus suggests that ‘Bad news‘, ‘Surprise‘, and ‘Good news‘ headlines are typically refering to more concrete events and entities than the other categories of news values. 5.2
This Coh-Metrix index refers to information about syntactic categories and function words, evaluated in the sentence context. To visualize the relations among newsworthiness and word information, we performed a hierarchical cluster analysis. We first represent each headline as a vector of ten word incidence scores (the number of words of a specific part-speech per 1000 words): incidence scores for nouns, verbs, adjectives, adverbs, personal pronouns, pronouns in first, second, and third person, separately for singular and plural. We then use hierarchical agglomerative clustering with complete linkage and one minus Pearsons correlation coefficient as the distance measure to obtain the clusters.
Fig. 1 shows the resulting dendrogram. We can identify three groups of news values on the basis of their syntactic structures.
The first group consists of only news values that can be defined positive contents/sentiments, namely ‘Good news’, ‘Entertainment’, and ‘Shareability’. This group is characterized by a quite high incidence of adjective, low incidence of first person singular and third person plural pronouns. Furthermore, this group presents the highest incidence of second person pronouns. As in the samples below: (1a) Feeding your brain: new benefits found in chocolate (Good news) (1b) Free Will: Now you have it, now you don’t (Entertainment) (3c) Eight years for Damilola killers (Follow up) (1c) Nap your way to a successful career (Shareability) (3d) Bomb kills 18 on military bus in Iran (Bad news)
The second group consists of ‘Celebrity’, ‘Power elite’, and ‘Drama’. This group presents low incidence of adjective and adverbs. The most incident pronouns are the first person plural and the third person singular. (2a) Beyonce new SI bikini cover girl (Celebrity) (2b) Bush vows cooperation on health care (Power elite) (2c) Collision on icy road kills 7 (Drama)
The third group consists of two subsets, the first one formed by ‘Surprise’ and ‘Magnitude’, and the second subset formed by ‘Follow up’, ‘Bad news’, and ‘Conflict’. ‘Surprise’ and ‘Magnitude’ form a different subset due to the presence of the highest score within all categories for the adjective incidence and a low incidence of pronouns. For instance: (3a) In the world of life-saving drugs, a growing epidemic of deadly fakes (Surprise) (3b) Flu Vaccine Appears Safe for Young Children (Magnitude)
The second subset is formed by negative contents/sentiment, characterized by the lowest incidence of adverbs and pronouns: (3e) Venezuela, Iran fight U.S. dominance (Conflict). 6
We described a preliminary study for on style of headlines in order to evaluate the correlation among syntactic features and newsworthiness. Our hypothesis is that each particular linguistic form or stylistic variation can be motivated by the purpose of encoding a certain newsworthy value. We performed a linguistic analysis to discover the correlations among newsworthiness and some stylistic features, on the basis of characteristics considered indicative of a shared communicative function and discriminating factors for headlines.
This preliminary analysis opens up a number of interesting research directions. One is the study of other stylistic variations of headlines, besides the ones examined in this paper. Another research direction is the comparison between style in headlines and full-text stories. It would also be interesting to analyze how communicative functions in headlines correlate with the events described in the pertaining text. We intend to pursue some of this work in the near future.
This work has been funded by the Unity Through Knowledge Fund of the Croatian Science Foundation, under the grant 19/15: EVEnt Retrieval Based on semantically Enriched Structures for Interactive user Tasks (EVERBEST).