Search results

In the era of artificial intelligence (AI), algorithms have gained unprecedented importance. Scientific studies have shown that algorithms are frequently mentioned in papers, making mention frequency a classical indicator of their popularity and influence. However, contemporary methods for evaluating influence tend to focus solely on individual algorithms, disregarding the collective impact resulting from the interconnectedness of these algorithms, which can provide a new way to reveal their roles and importance within algorithm clusters. This paper aims to build the co-occurrence network of algorithms in the natural language processing field based on the full-text content of academic papers and analyze the academic influence of algorithms in the group based on the features of the network.

We use deep learning models to extract algorithm entities from articles and construct the whole, cumulative and annual co-occurrence networks. We first analyze the characteristics of algorithm networks and then use various centrality metrics to obtain the score and ranking of group influence for each algorithm in the whole domain and each year. Finally, we analyze the influence evolution of different representative algorithms.

The results indicate that algorithm networks also have the characteristics of complex networks, with tight connections between nodes developing over approximately four decades. For different algorithms, algorithms that are classic, high-performing and appear at the junctions of different eras can possess high popularity, control, central position and balanced influence in the network. As an algorithm gradually diminishes its sway within the group, it typically loses its core position first, followed by a dwindling association with other algorithms.

To the best of the authors’ knowledge, this paper is the first large-scale analysis of algorithm networks. The extensive temporal coverage, spanning over four decades of academic publications, ensures the depth and integrity of the network. Our results serve as a cornerstone for constructing multifaceted networks interlinking algorithms, scholars and tasks, facilitating future exploration of their scientific roles and semantic relations.

The development of social media has led to large numbers of internet users now producing massive amounts of user-generated content (UGC). UGC, which shows users’ opinions about events directly, is valuable for monitoring public opinion. Current researches have focused on analysing topic evolutions in UGC. However, few researches pay attention to emotion evolutions of sub-topics about popular events. Important details about users’ opinions might be missed, as users’ emotions are ignored. This paper aims to extract sub-topics about a popular event from UGC and investigate the emotion evolutions of each sub-topic.

This paper first collects UGC about a popular event as experimental data and conducts subjectivity classification on the data to get subjective corpus. Second, the subjective corpus is classified into different emotion categories using supervised emotion classification. Meanwhile, a topic model is used to extract sub-topics about the event from the subjective corpora. Finally, the authors use the results of emotion classification and sub-topic extraction to analyze emotion evolutions over time.

Experimental results show that specific primary emotions exist in each sub-topic and undergo evolutions differently. Moreover, the authors find that performance of emotion classifier is optimal with term frequency and relevance frequency as the feature-weighting method.

To the best of the authors’ knowledge, this is the first research to mine emotion evolutions of sub-topics about an event with UGC. It mines users’ opinions about sub-topics of event, which may offer more details that are useful for analysing users’ emotions in preparation for decision-making.

Based on user-generated content from a Chinese social media platform, this paper aims to investigate multiple methods of constructing user profiles and their effectiveness in predicting their gender, age and geographic location.

This investigation collected 331,634 posts from 4,440 users of Sina Weibo. The data were divided into two parts, for training and testing . First, a vector space model and topic models were applied to construct user profiles. A classification model was then learned by a support vector machine according to the training data set. Finally, we used the classification model to predict users’ gender, age and geographic location in the testing data set.

The results revealed that in constructing user profiles, latent semantic analysis performed better on the task of predicting gender and age. By contrast, the method based on a traditional vector space model worked better in making predictions regarding the geographic location. In the process of applying a topic model to construct user profiles, the authors found that different prediction tasks should use different numbers of topics.

This study explores different user profile construction methods to predict Chinese social media network users’ gender, age and geographic location. The results of this paper will help to improve the quality of personal information gathered from social media platforms, and thereby improve personalized recommendation systems and personalized marketing.

To better understand the online reviews and help potential consumers, businessmen and product manufacturers effectively obtain users’ evaluation on product aspects, this paper aims to explore the distribution regularities of users’ attention and sentiment on product aspects from the temporal perspective of online reviews.

Temporal characteristics of online reviews (purchase time, review time and time intervals between purchase time and review time), similar attributes clustering and attribute-level sentiment computing technologies are used based on more than 340k smartphone reviews of three products from JD.COM (a famous online shopping platform in China) to explore the distribution regularities of users’ attention and sentiment on product aspects in this paper.

The empirical results show that a power-law distribution can fit users’ attention on product aspects, and the reviews posted in short time intervals contain more product aspects. Besides, the results show that the values of users’ sentiment on product aspects are significantly higher/lower in short time intervals which contribute to judging the advantages and weaknesses of a product.

This paper cannot acquire online reviews for more products with temporal characteristics to verify the findings because of the restriction on reviews crawling by the shopping platforms.

This work reveals the distribution regularities of users’ attention and sentiment on product aspects, which is of great significance in assisting decision-making, optimizing review presentation and improving the shopping experience.

To understand the meaning of a sentence, humans can focus on important words in the sentence, which reflects our eyes staying on each word in different gaze time or times. Thus, some studies utilize eye-tracking values to optimize the attention mechanism in deep learning models. But these studies lack to explain the rationality of this approach. Whether the attention mechanism possesses this feature of human reading needs to be explored.

The authors conducted experiments on a sentiment classification task. Firstly, they obtained eye-tracking values from two open-source eye-tracking corpora to describe the feature of human reading. Then, the machine attention values of each sentence were learned from a sentiment classification model. Finally, a comparison was conducted to analyze machine attention values and eye-tracking values.

Through experiments, the authors found the attention mechanism can focus on important words, such as adjectives, adverbs and sentiment words, which are valuable for judging the sentiment of sentences on the sentiment classification task. It possesses the feature of human reading, focusing on important words in sentences when reading. Due to the insufficient learning of the attention mechanism, some words are wrongly focused. The eye-tracking values can help the attention mechanism correct this error and improve the model performance.

Our research not only provides a reasonable explanation for the study of using eye-tracking values to optimize the attention mechanism but also provides new inspiration for the interpretability of attention mechanism.

Citation content in academic papers and academic monographs promotes the knowledge flow among different publications. However, existing citation content analysis (CCA) focuses on academic papers and monographs have not received much research attention. We want to know if monographs are appropriate objects of CCA and whether existing methods of analyzing citation in papers are suitable for citation in monographs. Therefore, this paper aims to learn more about features of cited references and citation content in monographs and compare the characteristic of citation pattern between monographs and papers.

The authors manually annotate the references and syntactic citation content in academic monographs published by Morgan & Claypool and automatically extracted the references and citation content from academic papers published by Public Library of Science. Five features in two types citation pattern, namely, pattern of cited reference (including year, source and mention frequency of reference) and pattern of citation content (including location, length of citation content) are used to examine similarities and differences between monographs and papers.

The results indicate that between monographs and papers, differences are shown in location, length of citation content and year, source of reference, whereas frequency of mention of reference is similar.

Previous studies have explored the patter of citation content in academic papers. However, none of the existing literature, as far as the authors know, has considered the citation content in academic monographs and the similarities or differences among academic documents when studying the citation pattern.

The purpose of this paper is to classify Chinese word semantic relations, which are synonyms, antonyms, hyponyms and meronymys.

Basically, four simple methods are applied, ontology-based, dictionary-based, pattern-based and morpho-syntactic method. The authors make good use of search engine to build lexical and semantic resources for dictionary-based and pattern-based methods. To improve classification performance with more external resources, they also classify the given word pairs in Chinese and in English at the same time by using machine translation.

Experimental results show that the approach achieved an average F1 score of 50.87 per cent, an average accuracy of 70.36 per cent and an average recall of 40.05 per cent over all classification tasks. Synonym and antonym classification achieved high accuracy, i.e. above 90 per cent. Moreover, dictionary-based and pattern-based approaches work effectively on final data set.

For many natural language processing (NLP) tasks, the step of distinguishing word semantic relation can help to improve system performance, such as information extraction and knowledge graph generation. Currently, common methods for this task rely on large corpora for training or dictionaries and thesauri for inference, where limitation lies in freely data access and keeping built lexical resources up-date. This paper builds a primary system for classifying Chinese word semantic relations by seeking new ways to obtain the external resources efficiently.

Academic social (question and answer) Q&A sites are now utilised by millions of scholars and researchers for seeking and sharing discipline-specific information. However, little is known about the factors that can affect their votes on the quality of an answer, nor how the discipline might influence these factors. The paper aims to discuss this issue.

Using 1,021 answers collected over three disciplines (library and information services, history of art, and astrophysics) in ResearchGate, statistical analysis is performed to identify the characteristics of high-quality academic answers, and comparisons were made across the three disciplines. In particular, two major categories of characteristics of the answer provider and answer content were extracted and examined.

The results reveal that high-quality answers on academic social Q&A sites tend to possess two characteristics: first, they are provided by scholars with higher academic reputations (e.g. more followers, etc.); and second, they provide objective information (e.g. longer answer with fewer subjective opinions). However, the impact of these factors varies across disciplines, e.g., objectivity is more favourable in physics than in other disciplines.

The study is envisioned to help academic Q&A sites to select and recommend high-quality answers across different disciplines, especially in a cold-start scenario where the answer has not received enough judgements from peers.