The color of the products is one of the most prevalent aspects in many e-commerce domains, and it is one of the decisive purchasing factors. Besides having thousands of color variations and shades, many brands continuously develop proprietary colors and color names to attract more customers. This often leads to color ambiguity (textual and visual), and vocabulary mismatch between buyers and sellers. Therefore, it is crucial for any e-commerce search engine to correctly identify the buyer's color intent and match it to the corresponding product listings. To address this challenge, in this work, we introduce a color query expansion approach using color Knowledge Graphs. We use Knowledge Graphs to unambiguously identify all the colors based on their properties, and the relationships to other colors, which allows us to perform semantic query expansion. Similar expansion concepts could be applied to domains outside of color.

The Fashion domain is one of the most profitable domains in most of the e-commerce shops, shoes being one of the top-selling categories within this domain. When shopping for shoes, one of the most important aspects for the buyers is the shoe size. Shoe size charts differ between different brands, geographical regions, genders and age groups. Not providing some of these details, as a buyer or a seller, could lead to a query intent to inventory mismatch and reduced or wrong search results. Furthermore, buying the wrong shoe size is one of the top reasons for product returns, which causes shipping delays and loss in revenue. To address this issue, we propose an approach for shoe size resolution and normalization in search queries and product listings using Knowledge Graphs.

Recent advances in text representation have shown that training on large amounts of text is crucial for natural language understanding. However, models trained without predefined notions of topical interest typically require careful fine-tuning when transferred to specialized domains. When a sufficient amount of within-domain text may not be available, expanding a seed corpus of relevant documents from large-scale web data poses several challenges. First, corpus expansion requires scoring and ranking each document in the collection, an operation that can quickly become computationally expensive as the web corpora size grows. Relying on dense vector spaces and pairwise similarity adds to the computational expense. Secondly, as the domain concept becomes more nuanced, capturing the long tail of domain-specific rare terms becomes non-trivial, especially under limited seed corpora scenarios.

Entity linking is a fundamental task for a successful use of knowledge graphs in many information systems. It maps textual mentions to their corresponding entities in a given knowledge graph. However, with the rapid evolution of knowledge graphs, a large number of entities is continuously added over time. Performing entity linking on new, or unseen, entities poses a great challenge, as standard entity linking approaches require large amounts of labeled data for all new entities, and the underlying model must be regularly updated. To address this challenge, several zero-shot entity linking approaches have been proposed, which don't require additional labeled data to perform entity linking over unseen entities and new domains. Most of these approaches use large language models, such as BERT, to encode the textual description of the mentions and entities in a common embedding space, which allows linking mentions to unseen entities. While such approaches have shown good performance, one big drawback is that they are not able to exploit the entity symbolic information from the knowledge graph, such as entity types, relations, popularity scores and graph embeddings. In this paper, we present KG-ZESHEL, a knowledge graph-enhanced zero-shot entity linking approach, which extends an existing BERT-based zero-shot entity linking approach with mention and entity auxiliary information. Experiments on two benchmark entity linking datasets, show that our proposed approach outperforms the related BERT-based state-of-the-art entity linking models.

The use of AI in knowledge dense domains, e.g., chemistry, medicine, biology, etc. - is extremely promising, but often suffers from slow deployment and adaptation to different tasks. We propose a methodology to quickly capture the intent and expertise of a domain expert in order to train personalized AI models for specific tasks. Specifically we focus on the domain of polymer materials design and discovery: it often takes 10 years or more to design, synthesize, test, and introduce a new polymer material into the market. One way to accelerate up the design of polymer materials is through the use of computational methods to design the material, such as combinatorial screening, generative models, inverse design, etc. The drawback of these methods is that they generate a large number of candidates for new molecules, which then need to be manually reviewed by subject matter experts who select only a dozen for further investigation. Our solution is a human-in-the-loop methodology where we rank the candidates according to a utility function that is learned via the continued interaction with the subject matter experts, but which is also constrained by specific chemical knowledge. We prove the viability of our proposed methodology in a polymer production lab and we (i) evaluate against datasets of polymers previously produced in the lab as well as (ii) producing several novel materials that are undergoing experimental development, and (iii) quantitatively show that standard synthetic accessibility scores do not inform about patterns of SME decisions.

Many Knowledge Extraction systems rely on semantic resources - dictionaries, ontologies, lexical resources - to extract information from unstructured text. A key for successful information extraction is to consider such resources as evolving artifacts and keep them up-to-date. In this paper, we tackle the problem of dictionary expansion and we propose a human-in-the-loop approach: we couple neural language models with tight human supervision to assist the user in building and maintaining domain-specific dictionaries. The approach works on any given input text corpus and is based on the explore and exploit paradigm: starting from a few seeds (or an existing dictionary) it effectively discovers new instances (explore) from the text corpus as well as predicts new potential instances which are not in the corpus, i.e. “unseen” , using the current dictionary entries (exploit). We evaluate our approach on five real-world dictionaries, achieving high accuracy with a rapid expansion rate.

Entity matching, or record linkage, is the task of identifying records that refer to the same entity.Naive entity matching techniques (i.e., brute-force pairwise comparisons) have quadratic complexity.A typical shortcut to the problem is to employ blocking techniques to reduce the number of comparisons, i.e. to partition the data in several blocks and only compare records within the same block.While classic blocking methods are designed for data from relational databases with clearly defined schemas, they are not applicable to data from Web tables, which are more prone to noise and do not come with an explicit schema.At the same time, Web tables are an interesting data source for many knowledge intensive tasks, which makes record linkage on Web Tables an important challenge.In this work, we propose an unsupervised approach to partition the data, that does not exploit any external knowledge, but only relies on heuristics to select the blocking attributes.We compare different partitioning methods: we use (i) clustering on bagof-words, (ii) binning via Locality-Sensitive Hashing and (iii) clustering using word embeddings.In particular, the clustering methods show good results on a standard dataset of Web Tables, and, when combined with word embeddings, are a robust solution which allows for computing the clusters in a dense, low-dimensional space.

Ontologies are dynamic artifacts that evolve both in structure and content. Keeping them up-to-date is a very expensive and critical operation for any application relying on semantic Web technologies. In this paper we focus on evolving the content of an ontology by extracting relevant instances of ontological concepts from text. We propose a novel technique which is (i) completely language independent, (ii) combines statistical methods with human-in-the-loop and (iii) exploits Linked Data as bootstrapping source. Our experiments on a publicly available medical corpus and on a Twitter dataset show that the proposed solution achieves comparable performances regardless of language, domain and style of text. Given that the method relies on a human-in-the-loop, our results can be safely fed directly back into Linked Data resources.

Taxonomies are an important ingredient of knowledge organization, and serve as a backbone for more sophisticated knowledge representations in intelligent systems, such as formal ontologies. However, building taxonomies manually is a costly endeavor, and hence, automatic methods for taxonomy induction are a good alternative to build large-scale taxonomies. In this paper, we propose TIEmb, an approach for automatic unsupervised class subsumption axiom extraction from knowledge bases using entity and text embeddings. We apply the approach on the WebIsA database, a database of subsumption relations extracted from the large portion of the World Wide Web, to extract class hierarchies in the Person and Place domain.