# Literature Review of Credibility of web sources

## Reliability of web sources

Source Links:

1.https://scholar.google.co.in/citations?user=mbqkcfsAAAAJ&hl=en
2.https://scholar.google.co.in/citations?user=T4iBN5cAAAAJ&hl=en

### Paper-1
* Kashyap Popat, Subhabrata Mukherjee, Jannik Strötgen, and Gerhard Weikum. 2016. **Credibility Assessment of Textual Claims on the Web.** *In Proceedings of the 25th ACM International on Conference on Information and Knowledge Management (CIKM '16). Association for Computing Machinery, New York, NY, USA, 2173–2178.* https://doi.org/10.1145/2983323.2983661

What is being done?

1. Given a claim in the form of a sentence or paragraph, they first use a search engine to identify documents from multiple web-sources, which refer to the claim. Then, they analyze the interplay between the language (e.g., bias, subjectivity, etc.) of the retrieved articles, and the reliability of the web-sources where the articles appeared. Finally, they propose a Distant Supervision based classifier which uses these factors to assess the credibility of the claim reported by multiple sources.
2. To capture the reliability of the web-source for each web article, they determine the AlexaRank and PageRank of its source and use them as proxies for the source reliability.

Novelty:
1. Language stylistic features for detecting credibility of new article

### Paper-2
Kashyap Popat, Subhabrata Mukherjee, Jannik Strötgen, and Gerhard Weikum. 2017. **Where the Truth Lies: Explaining the Credibility of Emerging Claims on the Web and Social Media.** *In Proceedings of the 26th International Conference on World Wide Web Companion (WWW '17 Companion). International World Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE, 1003–1012.* https://doi.org/10.1145/3041021.3055133

What is being done?

1. The authors assess the credibility of newly emerging and “long-tail” claims with sparse presence on the web by determining the stance, reliability, and trend of retrieved sources of evidence or counter-evidence, and by providing user interpretable explanations for the credibility verdict.


Novelty:
1. Extracted evidence is considered as explanation.
2. Trend aware assesment
3. Stance classification

### Paper-3
Kashyap Popat, Subhabrata Mukherjee, Jannik Strötgen, and Gerhard Weikum. 2018. **CredEye: A Credibility Lens for Analyzing and Explaining Misinformation.** *In Companion Proceedings of the The Web Conference 2018 (WWW '18). International World Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE, 155–158.* https://doi.org/10.1145/3184558.3186967

What is being done?

Given an input claim in arbitrary textual form on an arbitrary
topic, CredEye automatically retrieves relevant articles from the
Web, using a search engine. It analyzes the credibility of each text by language features, the stance of the text, and the trustworthiness of the source, aggregating all these into an overall verdict.


Novelty:
1. Content analysis and stance detection uses linguistic features.
2. DEMO is presented.

### Paper-4
Kashyap Popat, Subhabrata Mukherjee, Andrew Yates, Gerhard Weikum:
**DeClarE: Debunking Fake News and False Claims using Evidence-Aware Deep Learning.** *CoRR abs/1809.06416 (2018)*

What is being done?

Given an input claim, DeClarE searches for web articles related to the claim. It considers the context of the claim via word embeddings and the (language of) web articles captured via a bidirectional LSTM (biLSTM), while using an attention mechanism to focus on parts of the articles according to their relevance to the claim. DeClarE then aggregates all the information about claim source, web article contexts, attention weights, and trustworthiness of the underlying sources to assess the claim. It also derives informative features for interpretability, like source embeddings that capture trustworthiness and salient words captured via attention.


Novelty:
1. Single model for end-end task