17492716. AUTOMATIC MEASUREMENT OF SEMANTIC SIMILARITY OF CONVERSATIONS simplified abstract (International Business Machines Corporation)
Contents
AUTOMATIC MEASUREMENT OF SEMANTIC SIMILARITY OF CONVERSATIONS
Organization Name
International Business Machines Corporation
Inventor(s)
Ella Rabinovich of Hod Hasharon (IL)
Ateret Anaby - Tavor of Givat Ada (IL)
AUTOMATIC MEASUREMENT OF SEMANTIC SIMILARITY OF CONVERSATIONS - A simplified explanation of the abstract
This abstract first appeared for US patent application 17492716 titled 'AUTOMATIC MEASUREMENT OF SEMANTIC SIMILARITY OF CONVERSATIONS
Simplified Explanation
The patent application describes a method for automatically measuring the semantic similarity of conversations. Here is a simplified explanation of the abstract:
- The method takes two conversation texts as input, each containing a sequence of utterances.
- The sequences of utterances are encoded into corresponding sequences of semantic representations.
- A minimal edit distance is computed between the sequences of semantic representations.
- Based on the minimal edit distance, the method can quantify the semantic similarity between the two conversation texts or output an alignment of the utterance sequences.
Potential applications of this technology:
- Chatbot development: The method can be used to assess the semantic similarity between user queries and a database of pre-defined conversation texts, helping chatbots provide more accurate responses.
- Customer support analysis: The method can be applied to measure the semantic similarity between customer queries and support agent responses, enabling companies to evaluate the effectiveness of their support interactions.
- Language learning: The method can be used to compare conversations in different languages, helping language learners understand the semantic similarities and differences between utterances.
Problems solved by this technology:
- Manual assessment: The method automates the measurement of semantic similarity, eliminating the need for manual evaluation of conversation texts.
- Time-consuming analysis: By computing the minimal edit distance, the method provides a quick and efficient way to compare and align conversation sequences.
- Subjective evaluations: The method provides an objective measure of semantic similarity, reducing the reliance on subjective judgments.
Benefits of this technology:
- Improved accuracy: By quantifying semantic similarity, the method provides a more precise measure of the similarity between conversation texts.
- Time-saving: The automated process eliminates the need for manual assessment, saving time and resources.
- Objective evaluation: The method offers an objective measure of semantic similarity, reducing bias and subjectivity in the evaluation process.
Original Abstract Submitted
Automatic measurement of semantic textual similarity of conversations, by: receiving two conversation texts, each comprising a sequence of utterances; encoding each of the sequences of utterances into a corresponding sequence of semantic representations; computing a minimal edit distance between the sequences of semantic representations; and, based on the computation of the minimal edit distance, performing at least one of: quantifying a semantic similarity between the two conversation texts, and outputting an alignment of the two sequences of utterances with each other.
